diff --git a/.gitattributes b/.gitattributes
index 4d0692647539581025373e89b4c70fa37efb8f56..6d931bd8e4987ea0420efca09b64441bb0dca5e6 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -48,3 +48,10 @@ cpt_devstral_24B/checkpoints/checkpoint-400/tokenizer.json filter=lfs diff=lfs m
 cpt_devstral_24B/checkpoints/checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 cpt_devstral_24B/checkpoints/checkpoint-600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 cpt_devstral_24B/checkpoints/checkpoint-686/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/run-tqihmbvy.wandb filter=lfs diff=lfs merge=lfs -text
+sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/run-62isnb1e.wandb filter=lfs diff=lfs merge=lfs -text
+sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/run-mifdnkpc.wandb filter=lfs diff=lfs merge=lfs -text
+sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/run-sicxj35d.wandb filter=lfs diff=lfs merge=lfs -text
+sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/run-y3kepwdy.wandb filter=lfs diff=lfs merge=lfs -text
+sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/run-9jed4peb.wandb filter=lfs diff=lfs merge=lfs -text
+sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/run-cktxoubm.wandb filter=lfs diff=lfs merge=lfs -text
diff --git a/sft_devstral_24B/best_adapter/README.md b/sft_devstral_24B/best_adapter/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/best_adapter/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/best_adapter/adapter_config.json b/sft_devstral_24B/best_adapter/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/best_adapter/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/best_adapter/adapter_model.safetensors b/sft_devstral_24B/best_adapter/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..0f65b8c1e5dc3454dbcf69498c2037d670c73fe7
--- /dev/null
+++ b/sft_devstral_24B/best_adapter/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2525c1928ca9de033f220a5e840c681a9d5f8ed45f23f11db8475248b70b962f
+size 45690960
diff --git a/sft_devstral_24B/best_adapter/training_args.bin b/sft_devstral_24B/best_adapter/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/best_adapter/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/README.md b/sft_devstral_24B/checkpoints/checkpoint-1000/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-1000/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-1000/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c3e1a54ff30c07a7ae6d547c163482da7439523a
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea4c757c4c0f3bbeeaf2adc988ec942be6365d6898b69b63df419440670b6de8
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-1000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..553a71d2273b6820ede160bf0da558fd613d9704
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3c4d1c020b9578461ce598133e846fa5f23fb37515ccdc96eadc37880ce8b65
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-1000/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5d44c8feb810981e1bb37dcefb8504ff5127fbd8
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9117a0d2e77a257b35af0540b2152c9ac110688113b9e80bfcff71a2f4afd0c
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-1000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e30111c97a457e24af8941912ac7e0fa069426f7
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d143e54fa61abf54a1448fc63239d34fad8fd9690881106ffe349bfaf2479434
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-1000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..c450e17b860a58ee8b79a0289cbb20d57a9b3175
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/trainer_state.json
@@ -0,0 +1,3623 @@
+{
+  "best_global_step": 1000,
+  "best_metric": 0.8266019821166992,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-1000",
+  "epoch": 0.43149946062567424,
+  "eval_steps": 100,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0412690692397507e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1000/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-1000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/README.md b/sft_devstral_24B/checkpoints/checkpoint-1500/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-1500/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-1500/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..1fa090658ca5ca3efc5539971dd1c3ae2382fab4
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20fd579a68196fbf4f469c8f239ce89f433a5edc4747b2e6061e6ba08299082c
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-1500/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..543715749df3a651add05f57582dce1da387b2fa
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0eed4f313d3dc98c58cb77b09f35e4f0561e4ef87cc09b34da341b229d42dc7
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-1500/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..16eed24db1ed1541368923860a78ac7ef06cc63a
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4655532fe7fb62f39c418f0d96c32a9b86cae1ff752e2a984ec0fc714cf3c30
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-1500/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8b4896db19865015080d5a16654f083c5397bfa2
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32e80aa3f12daa32611e34f457685e6959d7a8561803cf45e2d67dae6fd24c81
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-1500/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..573ba7e5540153d54140fe7271291fa7ea6629d5
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/trainer_state.json
@@ -0,0 +1,5413 @@
+{
+  "best_global_step": 1500,
+  "best_metric": 0.778505802154541,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-1500",
+  "epoch": 0.6472491909385113,
+  "eval_steps": 100,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4323624595469256,
+      "grad_norm": 1.5475730895996094,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.7443642616271973,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4332254584681769,
+      "grad_norm": 1.4621447324752808,
+      "learning_rate": 9.011680143755616e-05,
+      "loss": 0.8210510015487671,
+      "step": 1004
+    },
+    {
+      "epoch": 0.43408845738942825,
+      "grad_norm": 1.3651922941207886,
+      "learning_rate": 9.029649595687331e-05,
+      "loss": 0.8130766153335571,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4349514563106796,
+      "grad_norm": 1.3587357997894287,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.79145747423172,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 1.464652180671692,
+      "learning_rate": 9.065588499550764e-05,
+      "loss": 0.8286946415901184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4366774541531823,
+      "grad_norm": 1.3912625312805176,
+      "learning_rate": 9.083557951482479e-05,
+      "loss": 0.717319667339325,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4375404530744337,
+      "grad_norm": 1.2045491933822632,
+      "learning_rate": 9.101527403414196e-05,
+      "loss": 0.7294418215751648,
+      "step": 1014
+    },
+    {
+      "epoch": 0.438403451995685,
+      "grad_norm": 1.3449069261550903,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.7366700768470764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43926645091693634,
+      "grad_norm": 2.901384115219116,
+      "learning_rate": 9.137466307277629e-05,
+      "loss": 0.8551293611526489,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 1.6648688316345215,
+      "learning_rate": 9.155435759209344e-05,
+      "loss": 0.8944112658500671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44099244875943905,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 9.17340521114106e-05,
+      "loss": 0.7644214630126953,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4418554476806904,
+      "grad_norm": 1.4643231630325317,
+      "learning_rate": 9.191374663072777e-05,
+      "loss": 0.8644799590110779,
+      "step": 1024
+    },
+    {
+      "epoch": 0.44271844660194176,
+      "grad_norm": 1.4690274000167847,
+      "learning_rate": 9.209344115004492e-05,
+      "loss": 0.8281022906303406,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4435814455231931,
+      "grad_norm": 1.3321986198425293,
+      "learning_rate": 9.227313566936209e-05,
+      "loss": 0.7432141900062561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.8143638372421265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4453074433656958,
+      "grad_norm": 1.747295618057251,
+      "learning_rate": 9.263252470799642e-05,
+      "loss": 0.8102239370346069,
+      "step": 1032
+    },
+    {
+      "epoch": 0.44617044228694713,
+      "grad_norm": 1.3273073434829712,
+      "learning_rate": 9.281221922731357e-05,
+      "loss": 0.7768088579177856,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4470334412081985,
+      "grad_norm": 1.5019903182983398,
+      "learning_rate": 9.299191374663073e-05,
+      "loss": 0.7886938452720642,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44789644012944985,
+      "grad_norm": 1.6711527109146118,
+      "learning_rate": 9.31716082659479e-05,
+      "loss": 0.8152258396148682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 1.8504793643951416,
+      "learning_rate": 9.335130278526505e-05,
+      "loss": 0.8339349627494812,
+      "step": 1040
+    },
+    {
+      "epoch": 0.44962243797195256,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 9.353099730458221e-05,
+      "loss": 0.7798911929130554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.45048543689320386,
+      "grad_norm": 1.3361700773239136,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.7676728963851929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4513484358144552,
+      "grad_norm": 1.338934063911438,
+      "learning_rate": 9.389038634321654e-05,
+      "loss": 0.7729206085205078,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4522114347357066,
+      "grad_norm": 1.4347442388534546,
+      "learning_rate": 9.40700808625337e-05,
+      "loss": 0.7923640608787537,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 1.4451628923416138,
+      "learning_rate": 9.424977538185086e-05,
+      "loss": 0.8068996071815491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4539374325782093,
+      "grad_norm": 1.1904500722885132,
+      "learning_rate": 9.442946990116803e-05,
+      "loss": 0.7532910704612732,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45480043149946064,
+      "grad_norm": 1.4103976488113403,
+      "learning_rate": 9.460916442048518e-05,
+      "loss": 0.8236854672431946,
+      "step": 1054
+    },
+    {
+      "epoch": 0.455663430420712,
+      "grad_norm": 1.3762754201889038,
+      "learning_rate": 9.478885893980234e-05,
+      "loss": 0.7532716393470764,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4565264293419633,
+      "grad_norm": 1.461421012878418,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.8506031632423401,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 1.7207789421081543,
+      "learning_rate": 9.514824797843667e-05,
+      "loss": 0.7824664115905762,
+      "step": 1060
+    },
+    {
+      "epoch": 0.458252427184466,
+      "grad_norm": 1.5946974754333496,
+      "learning_rate": 9.532794249775382e-05,
+      "loss": 0.8674250245094299,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4591154261057174,
+      "grad_norm": 1.3552565574645996,
+      "learning_rate": 9.550763701707099e-05,
+      "loss": 0.7851845622062683,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45997842502696873,
+      "grad_norm": 1.414617657661438,
+      "learning_rate": 9.568733153638815e-05,
+      "loss": 0.8890700936317444,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4608414239482201,
+      "grad_norm": 1.3322749137878418,
+      "learning_rate": 9.58670260557053e-05,
+      "loss": 0.8226593732833862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 1.41420316696167,
+      "learning_rate": 9.604672057502247e-05,
+      "loss": 0.7663303017616272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.46256742179072274,
+      "grad_norm": 1.324967384338379,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.7122516632080078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4634304207119741,
+      "grad_norm": 1.2046154737472534,
+      "learning_rate": 9.640610961365679e-05,
+      "loss": 0.7177618145942688,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46429341963322546,
+      "grad_norm": 1.5221766233444214,
+      "learning_rate": 9.658580413297395e-05,
+      "loss": 0.7618930339813232,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4651564185544768,
+      "grad_norm": 1.3517179489135742,
+      "learning_rate": 9.67654986522911e-05,
+      "loss": 0.7483780384063721,
+      "step": 1078
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.747865080833435,
+      "learning_rate": 9.694519317160827e-05,
+      "loss": 0.8507670164108276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4668824163969795,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 9.712488769092543e-05,
+      "loss": 0.7893540859222412,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46774541531823083,
+      "grad_norm": 1.6290462017059326,
+      "learning_rate": 9.730458221024258e-05,
+      "loss": 0.7891272306442261,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4686084142394822,
+      "grad_norm": 1.6605802774429321,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.841018795967102,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46947141316073354,
+      "grad_norm": 1.388964295387268,
+      "learning_rate": 9.766397124887691e-05,
+      "loss": 0.8099003434181213,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 1.3901978731155396,
+      "learning_rate": 9.784366576819407e-05,
+      "loss": 0.7177658677101135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.47119741100323626,
+      "grad_norm": 1.3341423273086548,
+      "learning_rate": 9.802336028751123e-05,
+      "loss": 0.8285676836967468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4720604099244876,
+      "grad_norm": 1.5451914072036743,
+      "learning_rate": 9.82030548068284e-05,
+      "loss": 0.7582666873931885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47292340884573897,
+      "grad_norm": 1.5395640134811401,
+      "learning_rate": 9.838274932614556e-05,
+      "loss": 0.7293475270271301,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47378640776699027,
+      "grad_norm": 1.5005141496658325,
+      "learning_rate": 9.856244384546271e-05,
+      "loss": 0.8858493566513062,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 1.3562248945236206,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.7176282405853271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "eval_loss": 0.8152187466621399,
+      "eval_runtime": 835.0394,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.475512405609493,
+      "grad_norm": 1.5226716995239258,
+      "learning_rate": 9.892183288409704e-05,
+      "loss": 0.8048524856567383,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47637540453074434,
+      "grad_norm": 1.3544842004776,
+      "learning_rate": 9.91015274034142e-05,
+      "loss": 0.7954329252243042,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4772384034519957,
+      "grad_norm": 1.5169771909713745,
+      "learning_rate": 9.928122192273136e-05,
+      "loss": 0.804015040397644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47810140237324705,
+      "grad_norm": 1.4011081457138062,
+      "learning_rate": 9.946091644204852e-05,
+      "loss": 0.7652344703674316,
+      "step": 1108
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 1.4163414239883423,
+      "learning_rate": 9.964061096136569e-05,
+      "loss": 0.8501335382461548,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4798274002157497,
+      "grad_norm": 1.7406005859375,
+      "learning_rate": 9.982030548068284e-05,
+      "loss": 0.9024044871330261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.48069039913700107,
+      "grad_norm": 1.1490557193756104,
+      "learning_rate": 0.0001,
+      "loss": 0.7308287024497986,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4815533980582524,
+      "grad_norm": 1.2512775659561157,
+      "learning_rate": 9.99999939713627e-05,
+      "loss": 0.7273974418640137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4824163969795038,
+      "grad_norm": 1.3944242000579834,
+      "learning_rate": 9.999997588545227e-05,
+      "loss": 0.8071790337562561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 1.2209112644195557,
+      "learning_rate": 9.999994574227303e-05,
+      "loss": 0.6586678624153137,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4841423948220065,
+      "grad_norm": 1.7328095436096191,
+      "learning_rate": 9.999990354183228e-05,
+      "loss": 0.8529233336448669,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4850053937432578,
+      "grad_norm": 1.3309327363967896,
+      "learning_rate": 9.999984928414021e-05,
+      "loss": 0.7529556155204773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.48586839266450915,
+      "grad_norm": 1.22034752368927,
+      "learning_rate": 9.999978296920987e-05,
+      "loss": 0.6780388355255127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4867313915857605,
+      "grad_norm": 1.975162386894226,
+      "learning_rate": 9.999970459705727e-05,
+      "loss": 0.7941693067550659,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 1.4939875602722168,
+      "learning_rate": 9.999961416770132e-05,
+      "loss": 0.7895156145095825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4884573894282632,
+      "grad_norm": 1.4237034320831299,
+      "learning_rate": 9.999951168116379e-05,
+      "loss": 0.7662149667739868,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4893203883495146,
+      "grad_norm": 1.24979567527771,
+      "learning_rate": 9.999939713746942e-05,
+      "loss": 0.8079231381416321,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49018338727076594,
+      "grad_norm": 1.2252353429794312,
+      "learning_rate": 9.999927053664585e-05,
+      "loss": 0.7616434693336487,
+      "step": 1136
+    },
+    {
+      "epoch": 0.49104638619201724,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.999913187872357e-05,
+      "loss": 0.8576793074607849,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 1.2108005285263062,
+      "learning_rate": 9.999898116373605e-05,
+      "loss": 0.8060292601585388,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49277238403451995,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 9.99988183917196e-05,
+      "loss": 0.7532875537872314,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4936353829557713,
+      "grad_norm": 1.261606216430664,
+      "learning_rate": 9.99986435627135e-05,
+      "loss": 0.7241997718811035,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49449838187702266,
+      "grad_norm": 1.2678134441375732,
+      "learning_rate": 9.99984566767599e-05,
+      "loss": 0.78656005859375,
+      "step": 1146
+    },
+    {
+      "epoch": 0.495361380798274,
+      "grad_norm": 1.4185810089111328,
+      "learning_rate": 9.999825773390387e-05,
+      "loss": 0.7423107624053955,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 1.3218237161636353,
+      "learning_rate": 9.999804673419338e-05,
+      "loss": 0.8178372383117676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4970873786407767,
+      "grad_norm": 1.9698240756988525,
+      "learning_rate": 9.99978236776793e-05,
+      "loss": 0.8016347885131836,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49795037756202803,
+      "grad_norm": 1.2228590250015259,
+      "learning_rate": 9.999758856441544e-05,
+      "loss": 0.782421886920929,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4988133764832794,
+      "grad_norm": 1.2753938436508179,
+      "learning_rate": 9.999734139445849e-05,
+      "loss": 0.7235615849494934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.49967637540453075,
+      "grad_norm": 1.353155493736267,
+      "learning_rate": 9.999708216786805e-05,
+      "loss": 0.7686187028884888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 1.4242770671844482,
+      "learning_rate": 9.999681088470663e-05,
+      "loss": 0.7842718362808228,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5014023732470334,
+      "grad_norm": 1.501827359199524,
+      "learning_rate": 9.999652754503966e-05,
+      "loss": 0.8324156999588013,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5022653721682848,
+      "grad_norm": 1.5403931140899658,
+      "learning_rate": 9.999623214893545e-05,
+      "loss": 0.7797855138778687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5031283710895361,
+      "grad_norm": 1.4075721502304077,
+      "learning_rate": 9.999592469646525e-05,
+      "loss": 0.7667348384857178,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5039913700107875,
+      "grad_norm": 1.4411870241165161,
+      "learning_rate": 9.99956051877032e-05,
+      "loss": 0.8153020739555359,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.4197871685028076,
+      "learning_rate": 9.999527362272632e-05,
+      "loss": 0.8548861742019653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5057173678532901,
+      "grad_norm": 1.5030468702316284,
+      "learning_rate": 9.999493000161458e-05,
+      "loss": 0.7978097796440125,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5065803667745415,
+      "grad_norm": 1.483534574508667,
+      "learning_rate": 9.999457432445088e-05,
+      "loss": 0.7990087866783142,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5074433656957928,
+      "grad_norm": 1.5424920320510864,
+      "learning_rate": 9.999420659132093e-05,
+      "loss": 0.8072807788848877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5083063646170443,
+      "grad_norm": 1.129503846168518,
+      "learning_rate": 9.999382680231344e-05,
+      "loss": 0.7216097712516785,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 1.3208131790161133,
+      "learning_rate": 9.999343495752001e-05,
+      "loss": 0.8637414574623108,
+      "step": 1180
+    },
+    {
+      "epoch": 0.510032362459547,
+      "grad_norm": 1.5239129066467285,
+      "learning_rate": 9.999303105703508e-05,
+      "loss": 0.8561362028121948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5108953613807983,
+      "grad_norm": 1.3193504810333252,
+      "learning_rate": 9.99926151009561e-05,
+      "loss": 0.7514780759811401,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5117583603020496,
+      "grad_norm": 1.312649130821228,
+      "learning_rate": 9.999218708938336e-05,
+      "loss": 0.8634516596794128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.512621359223301,
+      "grad_norm": 1.2677429914474487,
+      "learning_rate": 9.999174702242006e-05,
+      "loss": 0.7493832111358643,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 1.3901816606521606,
+      "learning_rate": 9.999129490017234e-05,
+      "loss": 0.8981389999389648,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5143473570658037,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 9.999083072274921e-05,
+      "loss": 0.6771019697189331,
+      "step": 1192
+    },
+    {
+      "epoch": 0.515210355987055,
+      "grad_norm": 1.1299076080322266,
+      "learning_rate": 9.999035449026261e-05,
+      "loss": 0.789888858795166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5160733549083064,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 9.998986620282738e-05,
+      "loss": 0.6656325459480286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5169363538295577,
+      "grad_norm": 1.3438260555267334,
+      "learning_rate": 9.998936586056127e-05,
+      "loss": 0.795595645904541,
+      "step": 1198
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 1.45097017288208,
+      "learning_rate": 9.998885346358495e-05,
+      "loss": 0.7769836187362671,
+      "step": 1200
+    },
+    {
+      "epoch": 0.517799352750809,
+      "eval_loss": 0.8070361018180847,
+      "eval_runtime": 823.9296,
+      "eval_samples_per_second": 2.5,
+      "eval_steps_per_second": 2.5,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5186623516720604,
+      "grad_norm": 1.3791370391845703,
+      "learning_rate": 9.998832901202195e-05,
+      "loss": 0.7527318000793457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5195253505933117,
+      "grad_norm": 1.2574659585952759,
+      "learning_rate": 9.998779250599876e-05,
+      "loss": 0.8021673560142517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5203883495145631,
+      "grad_norm": 1.3305809497833252,
+      "learning_rate": 9.998724394564476e-05,
+      "loss": 0.776639461517334,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5212513484358144,
+      "grad_norm": 1.2330507040023804,
+      "learning_rate": 9.998668333109223e-05,
+      "loss": 0.7381181120872498,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 1.0482324361801147,
+      "learning_rate": 9.998611066247635e-05,
+      "loss": 0.6595721244812012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5229773462783172,
+      "grad_norm": 1.5456492900848389,
+      "learning_rate": 9.998552593993522e-05,
+      "loss": 0.8388534784317017,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5238403451995685,
+      "grad_norm": 1.3730167150497437,
+      "learning_rate": 9.998492916360986e-05,
+      "loss": 0.7593658566474915,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5247033441208199,
+      "grad_norm": 1.2668508291244507,
+      "learning_rate": 9.998432033364416e-05,
+      "loss": 0.7796505093574524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5255663430420712,
+      "grad_norm": 1.1175068616867065,
+      "learning_rate": 9.998369945018493e-05,
+      "loss": 0.7493594288825989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 1.121848464012146,
+      "learning_rate": 9.99830665133819e-05,
+      "loss": 0.7791322469711304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5272923408845739,
+      "grad_norm": 1.49788236618042,
+      "learning_rate": 9.998242152338772e-05,
+      "loss": 0.8299469947814941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5281553398058253,
+      "grad_norm": 1.227125644683838,
+      "learning_rate": 9.998176448035791e-05,
+      "loss": 0.7911439538002014,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5290183387270766,
+      "grad_norm": 1.176081895828247,
+      "learning_rate": 9.998109538445091e-05,
+      "loss": 0.6487553119659424,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5298813376483279,
+      "grad_norm": 1.169835090637207,
+      "learning_rate": 9.99804142358281e-05,
+      "loss": 0.7331781983375549,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 1.4085001945495605,
+      "learning_rate": 9.997972103465367e-05,
+      "loss": 0.8198121190071106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5316073354908306,
+      "grad_norm": 1.8634135723114014,
+      "learning_rate": 9.997901578109484e-05,
+      "loss": 0.7099791169166565,
+      "step": 1232
+    },
+    {
+      "epoch": 0.532470334412082,
+      "grad_norm": 1.4719511270523071,
+      "learning_rate": 9.997829847532166e-05,
+      "loss": 0.7416951656341553,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.2430769205093384,
+      "learning_rate": 9.99775691175071e-05,
+      "loss": 0.7583408355712891,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5341963322545846,
+      "grad_norm": 1.2502130270004272,
+      "learning_rate": 9.997682770782706e-05,
+      "loss": 0.7639391422271729,
+      "step": 1238
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 1.3577396869659424,
+      "learning_rate": 9.997607424646032e-05,
+      "loss": 0.8304449319839478,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5359223300970873,
+      "grad_norm": 1.0986218452453613,
+      "learning_rate": 9.997530873358854e-05,
+      "loss": 0.7331952452659607,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5367853290183388,
+      "grad_norm": 1.4607514142990112,
+      "learning_rate": 9.997453116939637e-05,
+      "loss": 0.8173373341560364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.53764832793959,
+      "grad_norm": 1.2897552251815796,
+      "learning_rate": 9.997374155407128e-05,
+      "loss": 0.7677372694015503,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5385113268608415,
+      "grad_norm": 1.4581117630004883,
+      "learning_rate": 9.997293988780372e-05,
+      "loss": 0.8524296283721924,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 1.2029366493225098,
+      "learning_rate": 9.997212617078698e-05,
+      "loss": 0.7920866012573242,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5402373247033441,
+      "grad_norm": 1.109483242034912,
+      "learning_rate": 9.997130040321727e-05,
+      "loss": 0.7889736294746399,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5411003236245955,
+      "grad_norm": 1.1930333375930786,
+      "learning_rate": 9.997046258529377e-05,
+      "loss": 0.6593763828277588,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5419633225458468,
+      "grad_norm": 1.3668162822723389,
+      "learning_rate": 9.996961271721845e-05,
+      "loss": 0.7768212556838989,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5428263214670982,
+      "grad_norm": 1.2901886701583862,
+      "learning_rate": 9.996875079919632e-05,
+      "loss": 0.7392467260360718,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.366720199584961,
+      "learning_rate": 9.996787683143518e-05,
+      "loss": 0.6961479187011719,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5445523193096009,
+      "grad_norm": 1.2925949096679688,
+      "learning_rate": 9.99669908141458e-05,
+      "loss": 0.801517128944397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5454153182308522,
+      "grad_norm": 1.1967244148254395,
+      "learning_rate": 9.996609274754182e-05,
+      "loss": 0.7287030816078186,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5462783171521035,
+      "grad_norm": 1.419996738433838,
+      "learning_rate": 9.996518263183985e-05,
+      "loss": 0.7791255116462708,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5471413160733549,
+      "grad_norm": 1.3544845581054688,
+      "learning_rate": 9.996426046725933e-05,
+      "loss": 0.6665297150611877,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 1.6272345781326294,
+      "learning_rate": 9.996332625402262e-05,
+      "loss": 0.7902728915214539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5488673139158576,
+      "grad_norm": 1.4264246225357056,
+      "learning_rate": 9.996237999235502e-05,
+      "loss": 0.8483795523643494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5497303128371089,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 9.996142168248473e-05,
+      "loss": 0.7432113289833069,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5505933117583603,
+      "grad_norm": 0.9958710074424744,
+      "learning_rate": 9.996045132464283e-05,
+      "loss": 0.6721545457839966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5514563106796116,
+      "grad_norm": 1.371719241142273,
+      "learning_rate": 9.995946891906331e-05,
+      "loss": 0.7770468592643738,
+      "step": 1278
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 1.1906969547271729,
+      "learning_rate": 9.995847446598307e-05,
+      "loss": 0.816495418548584,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5531823085221144,
+      "grad_norm": 1.262184977531433,
+      "learning_rate": 9.995746796564195e-05,
+      "loss": 0.7122345566749573,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5540453074433657,
+      "grad_norm": 1.454692006111145,
+      "learning_rate": 9.995644941828262e-05,
+      "loss": 0.8476479649543762,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5549083063646171,
+      "grad_norm": 1.44277024269104,
+      "learning_rate": 9.995541882415072e-05,
+      "loss": 0.7595248818397522,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5557713052858684,
+      "grad_norm": 1.151465654373169,
+      "learning_rate": 9.995437618349478e-05,
+      "loss": 0.7413201928138733,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 1.3160182237625122,
+      "learning_rate": 9.995332149656621e-05,
+      "loss": 0.7242997884750366,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5574973031283711,
+      "grad_norm": 1.2748576402664185,
+      "learning_rate": 9.995225476361936e-05,
+      "loss": 0.7779543995857239,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5583603020496224,
+      "grad_norm": 1.3639990091323853,
+      "learning_rate": 9.995117598491145e-05,
+      "loss": 0.8075780868530273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5592233009708738,
+      "grad_norm": 1.0247999429702759,
+      "learning_rate": 9.995008516070263e-05,
+      "loss": 0.6940524578094482,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5600862998921251,
+      "grad_norm": 1.3383198976516724,
+      "learning_rate": 9.994898229125597e-05,
+      "loss": 0.799100399017334,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 1.37660551071167,
+      "learning_rate": 9.994786737683738e-05,
+      "loss": 0.8038942217826843,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "eval_loss": 0.7962777018547058,
+      "eval_runtime": 827.8798,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5618122977346278,
+      "grad_norm": 1.5424575805664062,
+      "learning_rate": 9.994674041771574e-05,
+      "loss": 0.8553245067596436,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5626752966558792,
+      "grad_norm": 1.1047813892364502,
+      "learning_rate": 9.994560141416282e-05,
+      "loss": 0.6650274991989136,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5635382955771305,
+      "grad_norm": 1.4532393217086792,
+      "learning_rate": 9.994445036645328e-05,
+      "loss": 0.794471025466919,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5644012944983818,
+      "grad_norm": 1.2588927745819092,
+      "learning_rate": 9.994328727486468e-05,
+      "loss": 0.7241487503051758,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 1.4981781244277954,
+      "learning_rate": 9.994211213967748e-05,
+      "loss": 0.7781044244766235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5661272923408845,
+      "grad_norm": 1.426277995109558,
+      "learning_rate": 9.99409249611751e-05,
+      "loss": 0.7772427201271057,
+      "step": 1312
+    },
+    {
+      "epoch": 0.566990291262136,
+      "grad_norm": 1.1987197399139404,
+      "learning_rate": 9.993972573964379e-05,
+      "loss": 0.7443774342536926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5678532901833873,
+      "grad_norm": 1.1490470170974731,
+      "learning_rate": 9.993851447537275e-05,
+      "loss": 0.7203393578529358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5687162891046387,
+      "grad_norm": 1.3586851358413696,
+      "learning_rate": 9.993729116865406e-05,
+      "loss": 0.8038793206214905,
+      "step": 1318
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 1.4160653352737427,
+      "learning_rate": 9.993605581978274e-05,
+      "loss": 0.827308714389801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5704422869471413,
+      "grad_norm": 1.2725251913070679,
+      "learning_rate": 9.993480842905667e-05,
+      "loss": 0.8183196783065796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5713052858683927,
+      "grad_norm": 1.3789756298065186,
+      "learning_rate": 9.993354899677665e-05,
+      "loss": 0.8268020153045654,
+      "step": 1324
+    },
+    {
+      "epoch": 0.572168284789644,
+      "grad_norm": 1.2641910314559937,
+      "learning_rate": 9.993227752324639e-05,
+      "loss": 0.7434715628623962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5730312837108954,
+      "grad_norm": 1.2429786920547485,
+      "learning_rate": 9.99309940087725e-05,
+      "loss": 0.7636955976486206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 1.3281534910202026,
+      "learning_rate": 9.992969845366449e-05,
+      "loss": 0.794617772102356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.574757281553398,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 9.992839085823477e-05,
+      "loss": 0.7658806443214417,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5756202804746494,
+      "grad_norm": 1.4143259525299072,
+      "learning_rate": 9.99270712227987e-05,
+      "loss": 0.8483052849769592,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5764832793959007,
+      "grad_norm": 1.2387818098068237,
+      "learning_rate": 9.992573954767449e-05,
+      "loss": 0.8610434532165527,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5773462783171521,
+      "grad_norm": 1.3307154178619385,
+      "learning_rate": 9.992439583318322e-05,
+      "loss": 0.8021210432052612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 1.293459415435791,
+      "learning_rate": 9.992304007964896e-05,
+      "loss": 0.7536460757255554,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5790722761596548,
+      "grad_norm": 1.1508948802947998,
+      "learning_rate": 9.992167228739865e-05,
+      "loss": 0.7519120573997498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5799352750809061,
+      "grad_norm": 1.5906150341033936,
+      "learning_rate": 9.992029245676211e-05,
+      "loss": 0.784415066242218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5807982740021574,
+      "grad_norm": 1.1421746015548706,
+      "learning_rate": 9.99189005880721e-05,
+      "loss": 0.8217549324035645,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5816612729234089,
+      "grad_norm": 1.0767277479171753,
+      "learning_rate": 9.991749668166422e-05,
+      "loss": 0.7446820735931396,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 1.147594928741455,
+      "learning_rate": 9.991608073787706e-05,
+      "loss": 0.8426011800765991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5833872707659116,
+      "grad_norm": 1.6618242263793945,
+      "learning_rate": 9.991465275705206e-05,
+      "loss": 0.8185623288154602,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5842502696871629,
+      "grad_norm": 1.414926290512085,
+      "learning_rate": 9.991321273953355e-05,
+      "loss": 0.7521756887435913,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5851132686084143,
+      "grad_norm": 1.298245906829834,
+      "learning_rate": 9.991176068566882e-05,
+      "loss": 0.7517471313476562,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5859762675296656,
+      "grad_norm": 1.3243110179901123,
+      "learning_rate": 9.9910296595808e-05,
+      "loss": 0.749305248260498,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 1.1953799724578857,
+      "learning_rate": 9.990882047030413e-05,
+      "loss": 0.6883649230003357,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5877022653721683,
+      "grad_norm": 1.3934388160705566,
+      "learning_rate": 9.990733230951322e-05,
+      "loss": 0.7859632968902588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5885652642934196,
+      "grad_norm": 1.2212340831756592,
+      "learning_rate": 9.990583211379409e-05,
+      "loss": 0.7313099503517151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.589428263214671,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 9.990431988350853e-05,
+      "loss": 0.8007960319519043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5902912621359223,
+      "grad_norm": 1.1826236248016357,
+      "learning_rate": 9.990279561902118e-05,
+      "loss": 0.773026168346405,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 1.1181561946868896,
+      "learning_rate": 9.990125932069964e-05,
+      "loss": 0.7469213604927063,
+      "step": 1370
+    },
+    {
+      "epoch": 0.592017259978425,
+      "grad_norm": 1.3735432624816895,
+      "learning_rate": 9.989971098891439e-05,
+      "loss": 0.7788950204849243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5928802588996763,
+      "grad_norm": 1.1906198263168335,
+      "learning_rate": 9.989815062403877e-05,
+      "loss": 0.8117855191230774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5937432578209277,
+      "grad_norm": 1.326940894126892,
+      "learning_rate": 9.989657822644907e-05,
+      "loss": 0.7536540627479553,
+      "step": 1376
+    },
+    {
+      "epoch": 0.594606256742179,
+      "grad_norm": 1.4508434534072876,
+      "learning_rate": 9.989499379652447e-05,
+      "loss": 0.860320508480072,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 1.114499568939209,
+      "learning_rate": 9.989339733464704e-05,
+      "loss": 0.749508798122406,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5963322545846818,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 9.989178884120176e-05,
+      "loss": 0.7294815182685852,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5971952535059332,
+      "grad_norm": 1.291913628578186,
+      "learning_rate": 9.989016831657652e-05,
+      "loss": 0.7797688245773315,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5980582524271845,
+      "grad_norm": 1.1245452165603638,
+      "learning_rate": 9.988853576116208e-05,
+      "loss": 0.7824407815933228,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5989212513484358,
+      "grad_norm": 1.2384979724884033,
+      "learning_rate": 9.988689117535215e-05,
+      "loss": 0.714794933795929,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 1.1945903301239014,
+      "learning_rate": 9.988523455954332e-05,
+      "loss": 0.7072597742080688,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6006472491909385,
+      "grad_norm": 1.3414266109466553,
+      "learning_rate": 9.988356591413504e-05,
+      "loss": 0.7075814008712769,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6015102481121899,
+      "grad_norm": 1.4575660228729248,
+      "learning_rate": 9.988188523952973e-05,
+      "loss": 0.8237126469612122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6023732470334412,
+      "grad_norm": 1.3836476802825928,
+      "learning_rate": 9.988019253613265e-05,
+      "loss": 0.7991626858711243,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6032362459546926,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.987848780435201e-05,
+      "loss": 0.6936330199241638,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 1.171294093132019,
+      "learning_rate": 9.98767710445989e-05,
+      "loss": 0.7834921479225159,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "eval_loss": 0.7866972088813782,
+      "eval_runtime": 836.0759,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6049622437971952,
+      "grad_norm": 1.5934456586837769,
+      "learning_rate": 9.987504225728727e-05,
+      "loss": 0.765119194984436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6058252427184466,
+      "grad_norm": 1.231372356414795,
+      "learning_rate": 9.987330144283407e-05,
+      "loss": 0.752061665058136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6066882416396979,
+      "grad_norm": 1.315215826034546,
+      "learning_rate": 9.987154860165903e-05,
+      "loss": 0.7475255727767944,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6075512405609493,
+      "grad_norm": 1.2640653848648071,
+      "learning_rate": 9.986978373418488e-05,
+      "loss": 0.777019202709198,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 1.085493803024292,
+      "learning_rate": 9.98680068408372e-05,
+      "loss": 0.6682190299034119,
+      "step": 1410
+    },
+    {
+      "epoch": 0.609277238403452,
+      "grad_norm": 1.4247251749038696,
+      "learning_rate": 9.986621792204447e-05,
+      "loss": 0.7765674591064453,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6101402373247033,
+      "grad_norm": 1.2993831634521484,
+      "learning_rate": 9.986441697823808e-05,
+      "loss": 0.7429938316345215,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6110032362459547,
+      "grad_norm": 2.614980936050415,
+      "learning_rate": 9.986260400985233e-05,
+      "loss": 0.7838386297225952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6118662351672061,
+      "grad_norm": 1.4305615425109863,
+      "learning_rate": 9.98607790173244e-05,
+      "loss": 0.7679475545883179,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 1.1412850618362427,
+      "learning_rate": 9.98589420010944e-05,
+      "loss": 0.6889669895172119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6135922330097088,
+      "grad_norm": 1.372313141822815,
+      "learning_rate": 9.98570929616053e-05,
+      "loss": 0.7659094333648682,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6144552319309601,
+      "grad_norm": 1.2629581689834595,
+      "learning_rate": 9.985523189930297e-05,
+      "loss": 0.7404800653457642,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6153182308522115,
+      "grad_norm": 1.5652254819869995,
+      "learning_rate": 9.985335881463623e-05,
+      "loss": 0.8082308769226074,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6161812297734628,
+      "grad_norm": 1.2268856763839722,
+      "learning_rate": 9.985147370805675e-05,
+      "loss": 0.7773997187614441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 1.2190957069396973,
+      "learning_rate": 9.984957658001912e-05,
+      "loss": 0.7412725687026978,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6179072276159655,
+      "grad_norm": 1.481990098953247,
+      "learning_rate": 9.984766743098081e-05,
+      "loss": 0.7257611751556396,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6187702265372168,
+      "grad_norm": 1.0297027826309204,
+      "learning_rate": 9.984574626140223e-05,
+      "loss": 0.7151464819908142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6196332254584682,
+      "grad_norm": 1.2404571771621704,
+      "learning_rate": 9.984381307174663e-05,
+      "loss": 0.7855883240699768,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6204962243797195,
+      "grad_norm": 1.2952802181243896,
+      "learning_rate": 9.984186786248023e-05,
+      "loss": 0.6360295414924622,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.5405724048614502,
+      "learning_rate": 9.983991063407206e-05,
+      "loss": 0.7280036807060242,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.303699254989624,
+      "learning_rate": 9.983794138699411e-05,
+      "loss": 0.7553068995475769,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6230852211434735,
+      "grad_norm": 1.1452997922897339,
+      "learning_rate": 9.983596012172127e-05,
+      "loss": 0.7329047918319702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.623948220064725,
+      "grad_norm": 1.2648253440856934,
+      "learning_rate": 9.983396683873133e-05,
+      "loss": 0.753168523311615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6248112189859762,
+      "grad_norm": 1.2095013856887817,
+      "learning_rate": 9.983196153850493e-05,
+      "loss": 0.7378705739974976,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 1.2855647802352905,
+      "learning_rate": 9.982994422152564e-05,
+      "loss": 0.7910448908805847,
+      "step": 1450
+    },
+    {
+      "epoch": 0.626537216828479,
+      "grad_norm": 1.3895999193191528,
+      "learning_rate": 9.982791488827994e-05,
+      "loss": 0.785302996635437,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6274002157497303,
+      "grad_norm": 1.2398446798324585,
+      "learning_rate": 9.982587353925719e-05,
+      "loss": 0.7542029023170471,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6282632146709817,
+      "grad_norm": 1.2903637886047363,
+      "learning_rate": 9.982382017494965e-05,
+      "loss": 0.8081737756729126,
+      "step": 1456
+    },
+    {
+      "epoch": 0.629126213592233,
+      "grad_norm": 1.223611831665039,
+      "learning_rate": 9.98217547958525e-05,
+      "loss": 0.7886892557144165,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 1.3036459684371948,
+      "learning_rate": 9.981967740246377e-05,
+      "loss": 0.844115138053894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6308522114347357,
+      "grad_norm": 1.6421090364456177,
+      "learning_rate": 9.98175879952844e-05,
+      "loss": 0.7784068584442139,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6317152103559871,
+      "grad_norm": 1.244330644607544,
+      "learning_rate": 9.981548657481828e-05,
+      "loss": 0.746979832649231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6325782092772384,
+      "grad_norm": 1.0323492288589478,
+      "learning_rate": 9.981337314157215e-05,
+      "loss": 0.6680153608322144,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6334412081984897,
+      "grad_norm": 1.2238314151763916,
+      "learning_rate": 9.981124769605563e-05,
+      "loss": 0.7103193402290344,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 1.2647801637649536,
+      "learning_rate": 9.98091102387813e-05,
+      "loss": 0.7426751255989075,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6351672060409924,
+      "grad_norm": 1.1923259496688843,
+      "learning_rate": 9.980696077026456e-05,
+      "loss": 0.7664520144462585,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6360302049622438,
+      "grad_norm": 1.0503894090652466,
+      "learning_rate": 9.980479929102377e-05,
+      "loss": 0.7348257899284363,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6368932038834951,
+      "grad_norm": 1.2223190069198608,
+      "learning_rate": 9.980262580158014e-05,
+      "loss": 0.687703013420105,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6377562028047465,
+      "grad_norm": 1.1469025611877441,
+      "learning_rate": 9.980044030245782e-05,
+      "loss": 0.7903051972389221,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 1.2581169605255127,
+      "learning_rate": 9.979824279418381e-05,
+      "loss": 0.7920540571212769,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6394822006472491,
+      "grad_norm": 1.2616536617279053,
+      "learning_rate": 9.979603327728806e-05,
+      "loss": 0.7030006647109985,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6403451995685006,
+      "grad_norm": 1.3970606327056885,
+      "learning_rate": 9.979381175230335e-05,
+      "loss": 0.7489339709281921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6412081984897519,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 9.979157821976542e-05,
+      "loss": 0.6733010411262512,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6420711974110033,
+      "grad_norm": 1.4987449645996094,
+      "learning_rate": 9.978933268021286e-05,
+      "loss": 0.8221960663795471,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 1.2843966484069824,
+      "learning_rate": 9.978707513418719e-05,
+      "loss": 0.8829001188278198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.643797195253506,
+      "grad_norm": 1.175879955291748,
+      "learning_rate": 9.978480558223278e-05,
+      "loss": 0.7605764269828796,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6446601941747573,
+      "grad_norm": 1.1991835832595825,
+      "learning_rate": 9.978252402489692e-05,
+      "loss": 0.7618141174316406,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6455231930960086,
+      "grad_norm": 1.167635202407837,
+      "learning_rate": 9.978023046272983e-05,
+      "loss": 0.7071146368980408,
+      "step": 1496
+    },
+    {
+      "epoch": 0.64638619201726,
+      "grad_norm": 1.263371229171753,
+      "learning_rate": 9.977792489628456e-05,
+      "loss": 0.7559816837310791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 1.2108519077301025,
+      "learning_rate": 9.977560732611713e-05,
+      "loss": 0.7253944873809814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "eval_loss": 0.778505802154541,
+      "eval_runtime": 836.2317,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.564200848328358e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-1500/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-1500/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-1500/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/README.md b/sft_devstral_24B/checkpoints/checkpoint-2000/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-2000/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-2000/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..5d4407a7209a9499be52e3caad6e304d0583bdfe
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:225825667e56783844ed24014e00a49b1af502ff61db12a0b15f467261ca070a
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-2000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6c15512845fc50ee3cc7932f67b1b220b76d2632
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1a5e7f450f3d05374406f39347483741e5ea3c9643beba9b397429a88c41c4c
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-2000/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b36e33acdbb22188977cc3fafee8720487369347
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01e13f651f318fbe600e3a90939a92c9240e80e5bb9954f55c343c0abfb99e48
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-2000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..06b9fe507a5790bb0b121cde0d8434d4f65a1082
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4207b23446848036378cf48f5d44035986145025a9cf3c265ffcfe1a8e01a20d
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-2000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..93fcea0dc34cc36f4e8a08c79295e1e55c067fb0
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/trainer_state.json
@@ -0,0 +1,7203 @@
+{
+  "best_global_step": 2000,
+  "best_metric": 0.7487027645111084,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-2000",
+  "epoch": 0.8629989212513485,
+  "eval_steps": 100,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4323624595469256,
+      "grad_norm": 1.5475730895996094,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.7443642616271973,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4332254584681769,
+      "grad_norm": 1.4621447324752808,
+      "learning_rate": 9.011680143755616e-05,
+      "loss": 0.8210510015487671,
+      "step": 1004
+    },
+    {
+      "epoch": 0.43408845738942825,
+      "grad_norm": 1.3651922941207886,
+      "learning_rate": 9.029649595687331e-05,
+      "loss": 0.8130766153335571,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4349514563106796,
+      "grad_norm": 1.3587357997894287,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.79145747423172,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 1.464652180671692,
+      "learning_rate": 9.065588499550764e-05,
+      "loss": 0.8286946415901184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4366774541531823,
+      "grad_norm": 1.3912625312805176,
+      "learning_rate": 9.083557951482479e-05,
+      "loss": 0.717319667339325,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4375404530744337,
+      "grad_norm": 1.2045491933822632,
+      "learning_rate": 9.101527403414196e-05,
+      "loss": 0.7294418215751648,
+      "step": 1014
+    },
+    {
+      "epoch": 0.438403451995685,
+      "grad_norm": 1.3449069261550903,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.7366700768470764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43926645091693634,
+      "grad_norm": 2.901384115219116,
+      "learning_rate": 9.137466307277629e-05,
+      "loss": 0.8551293611526489,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 1.6648688316345215,
+      "learning_rate": 9.155435759209344e-05,
+      "loss": 0.8944112658500671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44099244875943905,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 9.17340521114106e-05,
+      "loss": 0.7644214630126953,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4418554476806904,
+      "grad_norm": 1.4643231630325317,
+      "learning_rate": 9.191374663072777e-05,
+      "loss": 0.8644799590110779,
+      "step": 1024
+    },
+    {
+      "epoch": 0.44271844660194176,
+      "grad_norm": 1.4690274000167847,
+      "learning_rate": 9.209344115004492e-05,
+      "loss": 0.8281022906303406,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4435814455231931,
+      "grad_norm": 1.3321986198425293,
+      "learning_rate": 9.227313566936209e-05,
+      "loss": 0.7432141900062561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.8143638372421265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4453074433656958,
+      "grad_norm": 1.747295618057251,
+      "learning_rate": 9.263252470799642e-05,
+      "loss": 0.8102239370346069,
+      "step": 1032
+    },
+    {
+      "epoch": 0.44617044228694713,
+      "grad_norm": 1.3273073434829712,
+      "learning_rate": 9.281221922731357e-05,
+      "loss": 0.7768088579177856,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4470334412081985,
+      "grad_norm": 1.5019903182983398,
+      "learning_rate": 9.299191374663073e-05,
+      "loss": 0.7886938452720642,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44789644012944985,
+      "grad_norm": 1.6711527109146118,
+      "learning_rate": 9.31716082659479e-05,
+      "loss": 0.8152258396148682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 1.8504793643951416,
+      "learning_rate": 9.335130278526505e-05,
+      "loss": 0.8339349627494812,
+      "step": 1040
+    },
+    {
+      "epoch": 0.44962243797195256,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 9.353099730458221e-05,
+      "loss": 0.7798911929130554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.45048543689320386,
+      "grad_norm": 1.3361700773239136,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.7676728963851929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4513484358144552,
+      "grad_norm": 1.338934063911438,
+      "learning_rate": 9.389038634321654e-05,
+      "loss": 0.7729206085205078,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4522114347357066,
+      "grad_norm": 1.4347442388534546,
+      "learning_rate": 9.40700808625337e-05,
+      "loss": 0.7923640608787537,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 1.4451628923416138,
+      "learning_rate": 9.424977538185086e-05,
+      "loss": 0.8068996071815491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4539374325782093,
+      "grad_norm": 1.1904500722885132,
+      "learning_rate": 9.442946990116803e-05,
+      "loss": 0.7532910704612732,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45480043149946064,
+      "grad_norm": 1.4103976488113403,
+      "learning_rate": 9.460916442048518e-05,
+      "loss": 0.8236854672431946,
+      "step": 1054
+    },
+    {
+      "epoch": 0.455663430420712,
+      "grad_norm": 1.3762754201889038,
+      "learning_rate": 9.478885893980234e-05,
+      "loss": 0.7532716393470764,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4565264293419633,
+      "grad_norm": 1.461421012878418,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.8506031632423401,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 1.7207789421081543,
+      "learning_rate": 9.514824797843667e-05,
+      "loss": 0.7824664115905762,
+      "step": 1060
+    },
+    {
+      "epoch": 0.458252427184466,
+      "grad_norm": 1.5946974754333496,
+      "learning_rate": 9.532794249775382e-05,
+      "loss": 0.8674250245094299,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4591154261057174,
+      "grad_norm": 1.3552565574645996,
+      "learning_rate": 9.550763701707099e-05,
+      "loss": 0.7851845622062683,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45997842502696873,
+      "grad_norm": 1.414617657661438,
+      "learning_rate": 9.568733153638815e-05,
+      "loss": 0.8890700936317444,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4608414239482201,
+      "grad_norm": 1.3322749137878418,
+      "learning_rate": 9.58670260557053e-05,
+      "loss": 0.8226593732833862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 1.41420316696167,
+      "learning_rate": 9.604672057502247e-05,
+      "loss": 0.7663303017616272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.46256742179072274,
+      "grad_norm": 1.324967384338379,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.7122516632080078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4634304207119741,
+      "grad_norm": 1.2046154737472534,
+      "learning_rate": 9.640610961365679e-05,
+      "loss": 0.7177618145942688,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46429341963322546,
+      "grad_norm": 1.5221766233444214,
+      "learning_rate": 9.658580413297395e-05,
+      "loss": 0.7618930339813232,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4651564185544768,
+      "grad_norm": 1.3517179489135742,
+      "learning_rate": 9.67654986522911e-05,
+      "loss": 0.7483780384063721,
+      "step": 1078
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.747865080833435,
+      "learning_rate": 9.694519317160827e-05,
+      "loss": 0.8507670164108276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4668824163969795,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 9.712488769092543e-05,
+      "loss": 0.7893540859222412,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46774541531823083,
+      "grad_norm": 1.6290462017059326,
+      "learning_rate": 9.730458221024258e-05,
+      "loss": 0.7891272306442261,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4686084142394822,
+      "grad_norm": 1.6605802774429321,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.841018795967102,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46947141316073354,
+      "grad_norm": 1.388964295387268,
+      "learning_rate": 9.766397124887691e-05,
+      "loss": 0.8099003434181213,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 1.3901978731155396,
+      "learning_rate": 9.784366576819407e-05,
+      "loss": 0.7177658677101135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.47119741100323626,
+      "grad_norm": 1.3341423273086548,
+      "learning_rate": 9.802336028751123e-05,
+      "loss": 0.8285676836967468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4720604099244876,
+      "grad_norm": 1.5451914072036743,
+      "learning_rate": 9.82030548068284e-05,
+      "loss": 0.7582666873931885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47292340884573897,
+      "grad_norm": 1.5395640134811401,
+      "learning_rate": 9.838274932614556e-05,
+      "loss": 0.7293475270271301,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47378640776699027,
+      "grad_norm": 1.5005141496658325,
+      "learning_rate": 9.856244384546271e-05,
+      "loss": 0.8858493566513062,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 1.3562248945236206,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.7176282405853271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "eval_loss": 0.8152187466621399,
+      "eval_runtime": 835.0394,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.475512405609493,
+      "grad_norm": 1.5226716995239258,
+      "learning_rate": 9.892183288409704e-05,
+      "loss": 0.8048524856567383,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47637540453074434,
+      "grad_norm": 1.3544842004776,
+      "learning_rate": 9.91015274034142e-05,
+      "loss": 0.7954329252243042,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4772384034519957,
+      "grad_norm": 1.5169771909713745,
+      "learning_rate": 9.928122192273136e-05,
+      "loss": 0.804015040397644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47810140237324705,
+      "grad_norm": 1.4011081457138062,
+      "learning_rate": 9.946091644204852e-05,
+      "loss": 0.7652344703674316,
+      "step": 1108
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 1.4163414239883423,
+      "learning_rate": 9.964061096136569e-05,
+      "loss": 0.8501335382461548,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4798274002157497,
+      "grad_norm": 1.7406005859375,
+      "learning_rate": 9.982030548068284e-05,
+      "loss": 0.9024044871330261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.48069039913700107,
+      "grad_norm": 1.1490557193756104,
+      "learning_rate": 0.0001,
+      "loss": 0.7308287024497986,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4815533980582524,
+      "grad_norm": 1.2512775659561157,
+      "learning_rate": 9.99999939713627e-05,
+      "loss": 0.7273974418640137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4824163969795038,
+      "grad_norm": 1.3944242000579834,
+      "learning_rate": 9.999997588545227e-05,
+      "loss": 0.8071790337562561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 1.2209112644195557,
+      "learning_rate": 9.999994574227303e-05,
+      "loss": 0.6586678624153137,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4841423948220065,
+      "grad_norm": 1.7328095436096191,
+      "learning_rate": 9.999990354183228e-05,
+      "loss": 0.8529233336448669,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4850053937432578,
+      "grad_norm": 1.3309327363967896,
+      "learning_rate": 9.999984928414021e-05,
+      "loss": 0.7529556155204773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.48586839266450915,
+      "grad_norm": 1.22034752368927,
+      "learning_rate": 9.999978296920987e-05,
+      "loss": 0.6780388355255127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4867313915857605,
+      "grad_norm": 1.975162386894226,
+      "learning_rate": 9.999970459705727e-05,
+      "loss": 0.7941693067550659,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 1.4939875602722168,
+      "learning_rate": 9.999961416770132e-05,
+      "loss": 0.7895156145095825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4884573894282632,
+      "grad_norm": 1.4237034320831299,
+      "learning_rate": 9.999951168116379e-05,
+      "loss": 0.7662149667739868,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4893203883495146,
+      "grad_norm": 1.24979567527771,
+      "learning_rate": 9.999939713746942e-05,
+      "loss": 0.8079231381416321,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49018338727076594,
+      "grad_norm": 1.2252353429794312,
+      "learning_rate": 9.999927053664585e-05,
+      "loss": 0.7616434693336487,
+      "step": 1136
+    },
+    {
+      "epoch": 0.49104638619201724,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.999913187872357e-05,
+      "loss": 0.8576793074607849,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 1.2108005285263062,
+      "learning_rate": 9.999898116373605e-05,
+      "loss": 0.8060292601585388,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49277238403451995,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 9.99988183917196e-05,
+      "loss": 0.7532875537872314,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4936353829557713,
+      "grad_norm": 1.261606216430664,
+      "learning_rate": 9.99986435627135e-05,
+      "loss": 0.7241997718811035,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49449838187702266,
+      "grad_norm": 1.2678134441375732,
+      "learning_rate": 9.99984566767599e-05,
+      "loss": 0.78656005859375,
+      "step": 1146
+    },
+    {
+      "epoch": 0.495361380798274,
+      "grad_norm": 1.4185810089111328,
+      "learning_rate": 9.999825773390387e-05,
+      "loss": 0.7423107624053955,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 1.3218237161636353,
+      "learning_rate": 9.999804673419338e-05,
+      "loss": 0.8178372383117676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4970873786407767,
+      "grad_norm": 1.9698240756988525,
+      "learning_rate": 9.99978236776793e-05,
+      "loss": 0.8016347885131836,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49795037756202803,
+      "grad_norm": 1.2228590250015259,
+      "learning_rate": 9.999758856441544e-05,
+      "loss": 0.782421886920929,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4988133764832794,
+      "grad_norm": 1.2753938436508179,
+      "learning_rate": 9.999734139445849e-05,
+      "loss": 0.7235615849494934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.49967637540453075,
+      "grad_norm": 1.353155493736267,
+      "learning_rate": 9.999708216786805e-05,
+      "loss": 0.7686187028884888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 1.4242770671844482,
+      "learning_rate": 9.999681088470663e-05,
+      "loss": 0.7842718362808228,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5014023732470334,
+      "grad_norm": 1.501827359199524,
+      "learning_rate": 9.999652754503966e-05,
+      "loss": 0.8324156999588013,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5022653721682848,
+      "grad_norm": 1.5403931140899658,
+      "learning_rate": 9.999623214893545e-05,
+      "loss": 0.7797855138778687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5031283710895361,
+      "grad_norm": 1.4075721502304077,
+      "learning_rate": 9.999592469646525e-05,
+      "loss": 0.7667348384857178,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5039913700107875,
+      "grad_norm": 1.4411870241165161,
+      "learning_rate": 9.99956051877032e-05,
+      "loss": 0.8153020739555359,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.4197871685028076,
+      "learning_rate": 9.999527362272632e-05,
+      "loss": 0.8548861742019653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5057173678532901,
+      "grad_norm": 1.5030468702316284,
+      "learning_rate": 9.999493000161458e-05,
+      "loss": 0.7978097796440125,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5065803667745415,
+      "grad_norm": 1.483534574508667,
+      "learning_rate": 9.999457432445088e-05,
+      "loss": 0.7990087866783142,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5074433656957928,
+      "grad_norm": 1.5424920320510864,
+      "learning_rate": 9.999420659132093e-05,
+      "loss": 0.8072807788848877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5083063646170443,
+      "grad_norm": 1.129503846168518,
+      "learning_rate": 9.999382680231344e-05,
+      "loss": 0.7216097712516785,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 1.3208131790161133,
+      "learning_rate": 9.999343495752001e-05,
+      "loss": 0.8637414574623108,
+      "step": 1180
+    },
+    {
+      "epoch": 0.510032362459547,
+      "grad_norm": 1.5239129066467285,
+      "learning_rate": 9.999303105703508e-05,
+      "loss": 0.8561362028121948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5108953613807983,
+      "grad_norm": 1.3193504810333252,
+      "learning_rate": 9.99926151009561e-05,
+      "loss": 0.7514780759811401,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5117583603020496,
+      "grad_norm": 1.312649130821228,
+      "learning_rate": 9.999218708938336e-05,
+      "loss": 0.8634516596794128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.512621359223301,
+      "grad_norm": 1.2677429914474487,
+      "learning_rate": 9.999174702242006e-05,
+      "loss": 0.7493832111358643,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 1.3901816606521606,
+      "learning_rate": 9.999129490017234e-05,
+      "loss": 0.8981389999389648,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5143473570658037,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 9.999083072274921e-05,
+      "loss": 0.6771019697189331,
+      "step": 1192
+    },
+    {
+      "epoch": 0.515210355987055,
+      "grad_norm": 1.1299076080322266,
+      "learning_rate": 9.999035449026261e-05,
+      "loss": 0.789888858795166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5160733549083064,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 9.998986620282738e-05,
+      "loss": 0.6656325459480286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5169363538295577,
+      "grad_norm": 1.3438260555267334,
+      "learning_rate": 9.998936586056127e-05,
+      "loss": 0.795595645904541,
+      "step": 1198
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 1.45097017288208,
+      "learning_rate": 9.998885346358495e-05,
+      "loss": 0.7769836187362671,
+      "step": 1200
+    },
+    {
+      "epoch": 0.517799352750809,
+      "eval_loss": 0.8070361018180847,
+      "eval_runtime": 823.9296,
+      "eval_samples_per_second": 2.5,
+      "eval_steps_per_second": 2.5,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5186623516720604,
+      "grad_norm": 1.3791370391845703,
+      "learning_rate": 9.998832901202195e-05,
+      "loss": 0.7527318000793457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5195253505933117,
+      "grad_norm": 1.2574659585952759,
+      "learning_rate": 9.998779250599876e-05,
+      "loss": 0.8021673560142517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5203883495145631,
+      "grad_norm": 1.3305809497833252,
+      "learning_rate": 9.998724394564476e-05,
+      "loss": 0.776639461517334,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5212513484358144,
+      "grad_norm": 1.2330507040023804,
+      "learning_rate": 9.998668333109223e-05,
+      "loss": 0.7381181120872498,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 1.0482324361801147,
+      "learning_rate": 9.998611066247635e-05,
+      "loss": 0.6595721244812012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5229773462783172,
+      "grad_norm": 1.5456492900848389,
+      "learning_rate": 9.998552593993522e-05,
+      "loss": 0.8388534784317017,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5238403451995685,
+      "grad_norm": 1.3730167150497437,
+      "learning_rate": 9.998492916360986e-05,
+      "loss": 0.7593658566474915,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5247033441208199,
+      "grad_norm": 1.2668508291244507,
+      "learning_rate": 9.998432033364416e-05,
+      "loss": 0.7796505093574524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5255663430420712,
+      "grad_norm": 1.1175068616867065,
+      "learning_rate": 9.998369945018493e-05,
+      "loss": 0.7493594288825989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 1.121848464012146,
+      "learning_rate": 9.99830665133819e-05,
+      "loss": 0.7791322469711304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5272923408845739,
+      "grad_norm": 1.49788236618042,
+      "learning_rate": 9.998242152338772e-05,
+      "loss": 0.8299469947814941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5281553398058253,
+      "grad_norm": 1.227125644683838,
+      "learning_rate": 9.998176448035791e-05,
+      "loss": 0.7911439538002014,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5290183387270766,
+      "grad_norm": 1.176081895828247,
+      "learning_rate": 9.998109538445091e-05,
+      "loss": 0.6487553119659424,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5298813376483279,
+      "grad_norm": 1.169835090637207,
+      "learning_rate": 9.99804142358281e-05,
+      "loss": 0.7331781983375549,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 1.4085001945495605,
+      "learning_rate": 9.997972103465367e-05,
+      "loss": 0.8198121190071106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5316073354908306,
+      "grad_norm": 1.8634135723114014,
+      "learning_rate": 9.997901578109484e-05,
+      "loss": 0.7099791169166565,
+      "step": 1232
+    },
+    {
+      "epoch": 0.532470334412082,
+      "grad_norm": 1.4719511270523071,
+      "learning_rate": 9.997829847532166e-05,
+      "loss": 0.7416951656341553,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.2430769205093384,
+      "learning_rate": 9.99775691175071e-05,
+      "loss": 0.7583408355712891,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5341963322545846,
+      "grad_norm": 1.2502130270004272,
+      "learning_rate": 9.997682770782706e-05,
+      "loss": 0.7639391422271729,
+      "step": 1238
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 1.3577396869659424,
+      "learning_rate": 9.997607424646032e-05,
+      "loss": 0.8304449319839478,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5359223300970873,
+      "grad_norm": 1.0986218452453613,
+      "learning_rate": 9.997530873358854e-05,
+      "loss": 0.7331952452659607,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5367853290183388,
+      "grad_norm": 1.4607514142990112,
+      "learning_rate": 9.997453116939637e-05,
+      "loss": 0.8173373341560364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.53764832793959,
+      "grad_norm": 1.2897552251815796,
+      "learning_rate": 9.997374155407128e-05,
+      "loss": 0.7677372694015503,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5385113268608415,
+      "grad_norm": 1.4581117630004883,
+      "learning_rate": 9.997293988780372e-05,
+      "loss": 0.8524296283721924,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 1.2029366493225098,
+      "learning_rate": 9.997212617078698e-05,
+      "loss": 0.7920866012573242,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5402373247033441,
+      "grad_norm": 1.109483242034912,
+      "learning_rate": 9.997130040321727e-05,
+      "loss": 0.7889736294746399,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5411003236245955,
+      "grad_norm": 1.1930333375930786,
+      "learning_rate": 9.997046258529377e-05,
+      "loss": 0.6593763828277588,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5419633225458468,
+      "grad_norm": 1.3668162822723389,
+      "learning_rate": 9.996961271721845e-05,
+      "loss": 0.7768212556838989,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5428263214670982,
+      "grad_norm": 1.2901886701583862,
+      "learning_rate": 9.996875079919632e-05,
+      "loss": 0.7392467260360718,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.366720199584961,
+      "learning_rate": 9.996787683143518e-05,
+      "loss": 0.6961479187011719,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5445523193096009,
+      "grad_norm": 1.2925949096679688,
+      "learning_rate": 9.99669908141458e-05,
+      "loss": 0.801517128944397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5454153182308522,
+      "grad_norm": 1.1967244148254395,
+      "learning_rate": 9.996609274754182e-05,
+      "loss": 0.7287030816078186,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5462783171521035,
+      "grad_norm": 1.419996738433838,
+      "learning_rate": 9.996518263183985e-05,
+      "loss": 0.7791255116462708,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5471413160733549,
+      "grad_norm": 1.3544845581054688,
+      "learning_rate": 9.996426046725933e-05,
+      "loss": 0.6665297150611877,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 1.6272345781326294,
+      "learning_rate": 9.996332625402262e-05,
+      "loss": 0.7902728915214539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5488673139158576,
+      "grad_norm": 1.4264246225357056,
+      "learning_rate": 9.996237999235502e-05,
+      "loss": 0.8483795523643494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5497303128371089,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 9.996142168248473e-05,
+      "loss": 0.7432113289833069,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5505933117583603,
+      "grad_norm": 0.9958710074424744,
+      "learning_rate": 9.996045132464283e-05,
+      "loss": 0.6721545457839966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5514563106796116,
+      "grad_norm": 1.371719241142273,
+      "learning_rate": 9.995946891906331e-05,
+      "loss": 0.7770468592643738,
+      "step": 1278
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 1.1906969547271729,
+      "learning_rate": 9.995847446598307e-05,
+      "loss": 0.816495418548584,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5531823085221144,
+      "grad_norm": 1.262184977531433,
+      "learning_rate": 9.995746796564195e-05,
+      "loss": 0.7122345566749573,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5540453074433657,
+      "grad_norm": 1.454692006111145,
+      "learning_rate": 9.995644941828262e-05,
+      "loss": 0.8476479649543762,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5549083063646171,
+      "grad_norm": 1.44277024269104,
+      "learning_rate": 9.995541882415072e-05,
+      "loss": 0.7595248818397522,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5557713052858684,
+      "grad_norm": 1.151465654373169,
+      "learning_rate": 9.995437618349478e-05,
+      "loss": 0.7413201928138733,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 1.3160182237625122,
+      "learning_rate": 9.995332149656621e-05,
+      "loss": 0.7242997884750366,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5574973031283711,
+      "grad_norm": 1.2748576402664185,
+      "learning_rate": 9.995225476361936e-05,
+      "loss": 0.7779543995857239,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5583603020496224,
+      "grad_norm": 1.3639990091323853,
+      "learning_rate": 9.995117598491145e-05,
+      "loss": 0.8075780868530273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5592233009708738,
+      "grad_norm": 1.0247999429702759,
+      "learning_rate": 9.995008516070263e-05,
+      "loss": 0.6940524578094482,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5600862998921251,
+      "grad_norm": 1.3383198976516724,
+      "learning_rate": 9.994898229125597e-05,
+      "loss": 0.799100399017334,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 1.37660551071167,
+      "learning_rate": 9.994786737683738e-05,
+      "loss": 0.8038942217826843,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "eval_loss": 0.7962777018547058,
+      "eval_runtime": 827.8798,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5618122977346278,
+      "grad_norm": 1.5424575805664062,
+      "learning_rate": 9.994674041771574e-05,
+      "loss": 0.8553245067596436,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5626752966558792,
+      "grad_norm": 1.1047813892364502,
+      "learning_rate": 9.994560141416282e-05,
+      "loss": 0.6650274991989136,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5635382955771305,
+      "grad_norm": 1.4532393217086792,
+      "learning_rate": 9.994445036645328e-05,
+      "loss": 0.794471025466919,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5644012944983818,
+      "grad_norm": 1.2588927745819092,
+      "learning_rate": 9.994328727486468e-05,
+      "loss": 0.7241487503051758,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 1.4981781244277954,
+      "learning_rate": 9.994211213967748e-05,
+      "loss": 0.7781044244766235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5661272923408845,
+      "grad_norm": 1.426277995109558,
+      "learning_rate": 9.99409249611751e-05,
+      "loss": 0.7772427201271057,
+      "step": 1312
+    },
+    {
+      "epoch": 0.566990291262136,
+      "grad_norm": 1.1987197399139404,
+      "learning_rate": 9.993972573964379e-05,
+      "loss": 0.7443774342536926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5678532901833873,
+      "grad_norm": 1.1490470170974731,
+      "learning_rate": 9.993851447537275e-05,
+      "loss": 0.7203393578529358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5687162891046387,
+      "grad_norm": 1.3586851358413696,
+      "learning_rate": 9.993729116865406e-05,
+      "loss": 0.8038793206214905,
+      "step": 1318
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 1.4160653352737427,
+      "learning_rate": 9.993605581978274e-05,
+      "loss": 0.827308714389801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5704422869471413,
+      "grad_norm": 1.2725251913070679,
+      "learning_rate": 9.993480842905667e-05,
+      "loss": 0.8183196783065796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5713052858683927,
+      "grad_norm": 1.3789756298065186,
+      "learning_rate": 9.993354899677665e-05,
+      "loss": 0.8268020153045654,
+      "step": 1324
+    },
+    {
+      "epoch": 0.572168284789644,
+      "grad_norm": 1.2641910314559937,
+      "learning_rate": 9.993227752324639e-05,
+      "loss": 0.7434715628623962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5730312837108954,
+      "grad_norm": 1.2429786920547485,
+      "learning_rate": 9.99309940087725e-05,
+      "loss": 0.7636955976486206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 1.3281534910202026,
+      "learning_rate": 9.992969845366449e-05,
+      "loss": 0.794617772102356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.574757281553398,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 9.992839085823477e-05,
+      "loss": 0.7658806443214417,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5756202804746494,
+      "grad_norm": 1.4143259525299072,
+      "learning_rate": 9.99270712227987e-05,
+      "loss": 0.8483052849769592,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5764832793959007,
+      "grad_norm": 1.2387818098068237,
+      "learning_rate": 9.992573954767449e-05,
+      "loss": 0.8610434532165527,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5773462783171521,
+      "grad_norm": 1.3307154178619385,
+      "learning_rate": 9.992439583318322e-05,
+      "loss": 0.8021210432052612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 1.293459415435791,
+      "learning_rate": 9.992304007964896e-05,
+      "loss": 0.7536460757255554,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5790722761596548,
+      "grad_norm": 1.1508948802947998,
+      "learning_rate": 9.992167228739865e-05,
+      "loss": 0.7519120573997498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5799352750809061,
+      "grad_norm": 1.5906150341033936,
+      "learning_rate": 9.992029245676211e-05,
+      "loss": 0.784415066242218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5807982740021574,
+      "grad_norm": 1.1421746015548706,
+      "learning_rate": 9.99189005880721e-05,
+      "loss": 0.8217549324035645,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5816612729234089,
+      "grad_norm": 1.0767277479171753,
+      "learning_rate": 9.991749668166422e-05,
+      "loss": 0.7446820735931396,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 1.147594928741455,
+      "learning_rate": 9.991608073787706e-05,
+      "loss": 0.8426011800765991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5833872707659116,
+      "grad_norm": 1.6618242263793945,
+      "learning_rate": 9.991465275705206e-05,
+      "loss": 0.8185623288154602,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5842502696871629,
+      "grad_norm": 1.414926290512085,
+      "learning_rate": 9.991321273953355e-05,
+      "loss": 0.7521756887435913,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5851132686084143,
+      "grad_norm": 1.298245906829834,
+      "learning_rate": 9.991176068566882e-05,
+      "loss": 0.7517471313476562,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5859762675296656,
+      "grad_norm": 1.3243110179901123,
+      "learning_rate": 9.9910296595808e-05,
+      "loss": 0.749305248260498,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 1.1953799724578857,
+      "learning_rate": 9.990882047030413e-05,
+      "loss": 0.6883649230003357,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5877022653721683,
+      "grad_norm": 1.3934388160705566,
+      "learning_rate": 9.990733230951322e-05,
+      "loss": 0.7859632968902588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5885652642934196,
+      "grad_norm": 1.2212340831756592,
+      "learning_rate": 9.990583211379409e-05,
+      "loss": 0.7313099503517151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.589428263214671,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 9.990431988350853e-05,
+      "loss": 0.8007960319519043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5902912621359223,
+      "grad_norm": 1.1826236248016357,
+      "learning_rate": 9.990279561902118e-05,
+      "loss": 0.773026168346405,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 1.1181561946868896,
+      "learning_rate": 9.990125932069964e-05,
+      "loss": 0.7469213604927063,
+      "step": 1370
+    },
+    {
+      "epoch": 0.592017259978425,
+      "grad_norm": 1.3735432624816895,
+      "learning_rate": 9.989971098891439e-05,
+      "loss": 0.7788950204849243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5928802588996763,
+      "grad_norm": 1.1906198263168335,
+      "learning_rate": 9.989815062403877e-05,
+      "loss": 0.8117855191230774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5937432578209277,
+      "grad_norm": 1.326940894126892,
+      "learning_rate": 9.989657822644907e-05,
+      "loss": 0.7536540627479553,
+      "step": 1376
+    },
+    {
+      "epoch": 0.594606256742179,
+      "grad_norm": 1.4508434534072876,
+      "learning_rate": 9.989499379652447e-05,
+      "loss": 0.860320508480072,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 1.114499568939209,
+      "learning_rate": 9.989339733464704e-05,
+      "loss": 0.749508798122406,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5963322545846818,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 9.989178884120176e-05,
+      "loss": 0.7294815182685852,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5971952535059332,
+      "grad_norm": 1.291913628578186,
+      "learning_rate": 9.989016831657652e-05,
+      "loss": 0.7797688245773315,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5980582524271845,
+      "grad_norm": 1.1245452165603638,
+      "learning_rate": 9.988853576116208e-05,
+      "loss": 0.7824407815933228,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5989212513484358,
+      "grad_norm": 1.2384979724884033,
+      "learning_rate": 9.988689117535215e-05,
+      "loss": 0.714794933795929,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 1.1945903301239014,
+      "learning_rate": 9.988523455954332e-05,
+      "loss": 0.7072597742080688,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6006472491909385,
+      "grad_norm": 1.3414266109466553,
+      "learning_rate": 9.988356591413504e-05,
+      "loss": 0.7075814008712769,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6015102481121899,
+      "grad_norm": 1.4575660228729248,
+      "learning_rate": 9.988188523952973e-05,
+      "loss": 0.8237126469612122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6023732470334412,
+      "grad_norm": 1.3836476802825928,
+      "learning_rate": 9.988019253613265e-05,
+      "loss": 0.7991626858711243,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6032362459546926,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.987848780435201e-05,
+      "loss": 0.6936330199241638,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 1.171294093132019,
+      "learning_rate": 9.98767710445989e-05,
+      "loss": 0.7834921479225159,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "eval_loss": 0.7866972088813782,
+      "eval_runtime": 836.0759,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6049622437971952,
+      "grad_norm": 1.5934456586837769,
+      "learning_rate": 9.987504225728727e-05,
+      "loss": 0.765119194984436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6058252427184466,
+      "grad_norm": 1.231372356414795,
+      "learning_rate": 9.987330144283407e-05,
+      "loss": 0.752061665058136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6066882416396979,
+      "grad_norm": 1.315215826034546,
+      "learning_rate": 9.987154860165903e-05,
+      "loss": 0.7475255727767944,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6075512405609493,
+      "grad_norm": 1.2640653848648071,
+      "learning_rate": 9.986978373418488e-05,
+      "loss": 0.777019202709198,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 1.085493803024292,
+      "learning_rate": 9.98680068408372e-05,
+      "loss": 0.6682190299034119,
+      "step": 1410
+    },
+    {
+      "epoch": 0.609277238403452,
+      "grad_norm": 1.4247251749038696,
+      "learning_rate": 9.986621792204447e-05,
+      "loss": 0.7765674591064453,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6101402373247033,
+      "grad_norm": 1.2993831634521484,
+      "learning_rate": 9.986441697823808e-05,
+      "loss": 0.7429938316345215,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6110032362459547,
+      "grad_norm": 2.614980936050415,
+      "learning_rate": 9.986260400985233e-05,
+      "loss": 0.7838386297225952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6118662351672061,
+      "grad_norm": 1.4305615425109863,
+      "learning_rate": 9.98607790173244e-05,
+      "loss": 0.7679475545883179,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 1.1412850618362427,
+      "learning_rate": 9.98589420010944e-05,
+      "loss": 0.6889669895172119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6135922330097088,
+      "grad_norm": 1.372313141822815,
+      "learning_rate": 9.98570929616053e-05,
+      "loss": 0.7659094333648682,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6144552319309601,
+      "grad_norm": 1.2629581689834595,
+      "learning_rate": 9.985523189930297e-05,
+      "loss": 0.7404800653457642,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6153182308522115,
+      "grad_norm": 1.5652254819869995,
+      "learning_rate": 9.985335881463623e-05,
+      "loss": 0.8082308769226074,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6161812297734628,
+      "grad_norm": 1.2268856763839722,
+      "learning_rate": 9.985147370805675e-05,
+      "loss": 0.7773997187614441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 1.2190957069396973,
+      "learning_rate": 9.984957658001912e-05,
+      "loss": 0.7412725687026978,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6179072276159655,
+      "grad_norm": 1.481990098953247,
+      "learning_rate": 9.984766743098081e-05,
+      "loss": 0.7257611751556396,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6187702265372168,
+      "grad_norm": 1.0297027826309204,
+      "learning_rate": 9.984574626140223e-05,
+      "loss": 0.7151464819908142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6196332254584682,
+      "grad_norm": 1.2404571771621704,
+      "learning_rate": 9.984381307174663e-05,
+      "loss": 0.7855883240699768,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6204962243797195,
+      "grad_norm": 1.2952802181243896,
+      "learning_rate": 9.984186786248023e-05,
+      "loss": 0.6360295414924622,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.5405724048614502,
+      "learning_rate": 9.983991063407206e-05,
+      "loss": 0.7280036807060242,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.303699254989624,
+      "learning_rate": 9.983794138699411e-05,
+      "loss": 0.7553068995475769,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6230852211434735,
+      "grad_norm": 1.1452997922897339,
+      "learning_rate": 9.983596012172127e-05,
+      "loss": 0.7329047918319702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.623948220064725,
+      "grad_norm": 1.2648253440856934,
+      "learning_rate": 9.983396683873133e-05,
+      "loss": 0.753168523311615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6248112189859762,
+      "grad_norm": 1.2095013856887817,
+      "learning_rate": 9.983196153850493e-05,
+      "loss": 0.7378705739974976,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 1.2855647802352905,
+      "learning_rate": 9.982994422152564e-05,
+      "loss": 0.7910448908805847,
+      "step": 1450
+    },
+    {
+      "epoch": 0.626537216828479,
+      "grad_norm": 1.3895999193191528,
+      "learning_rate": 9.982791488827994e-05,
+      "loss": 0.785302996635437,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6274002157497303,
+      "grad_norm": 1.2398446798324585,
+      "learning_rate": 9.982587353925719e-05,
+      "loss": 0.7542029023170471,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6282632146709817,
+      "grad_norm": 1.2903637886047363,
+      "learning_rate": 9.982382017494965e-05,
+      "loss": 0.8081737756729126,
+      "step": 1456
+    },
+    {
+      "epoch": 0.629126213592233,
+      "grad_norm": 1.223611831665039,
+      "learning_rate": 9.98217547958525e-05,
+      "loss": 0.7886892557144165,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 1.3036459684371948,
+      "learning_rate": 9.981967740246377e-05,
+      "loss": 0.844115138053894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6308522114347357,
+      "grad_norm": 1.6421090364456177,
+      "learning_rate": 9.98175879952844e-05,
+      "loss": 0.7784068584442139,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6317152103559871,
+      "grad_norm": 1.244330644607544,
+      "learning_rate": 9.981548657481828e-05,
+      "loss": 0.746979832649231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6325782092772384,
+      "grad_norm": 1.0323492288589478,
+      "learning_rate": 9.981337314157215e-05,
+      "loss": 0.6680153608322144,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6334412081984897,
+      "grad_norm": 1.2238314151763916,
+      "learning_rate": 9.981124769605563e-05,
+      "loss": 0.7103193402290344,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 1.2647801637649536,
+      "learning_rate": 9.98091102387813e-05,
+      "loss": 0.7426751255989075,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6351672060409924,
+      "grad_norm": 1.1923259496688843,
+      "learning_rate": 9.980696077026456e-05,
+      "loss": 0.7664520144462585,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6360302049622438,
+      "grad_norm": 1.0503894090652466,
+      "learning_rate": 9.980479929102377e-05,
+      "loss": 0.7348257899284363,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6368932038834951,
+      "grad_norm": 1.2223190069198608,
+      "learning_rate": 9.980262580158014e-05,
+      "loss": 0.687703013420105,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6377562028047465,
+      "grad_norm": 1.1469025611877441,
+      "learning_rate": 9.980044030245782e-05,
+      "loss": 0.7903051972389221,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 1.2581169605255127,
+      "learning_rate": 9.979824279418381e-05,
+      "loss": 0.7920540571212769,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6394822006472491,
+      "grad_norm": 1.2616536617279053,
+      "learning_rate": 9.979603327728806e-05,
+      "loss": 0.7030006647109985,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6403451995685006,
+      "grad_norm": 1.3970606327056885,
+      "learning_rate": 9.979381175230335e-05,
+      "loss": 0.7489339709281921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6412081984897519,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 9.979157821976542e-05,
+      "loss": 0.6733010411262512,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6420711974110033,
+      "grad_norm": 1.4987449645996094,
+      "learning_rate": 9.978933268021286e-05,
+      "loss": 0.8221960663795471,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 1.2843966484069824,
+      "learning_rate": 9.978707513418719e-05,
+      "loss": 0.8829001188278198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.643797195253506,
+      "grad_norm": 1.175879955291748,
+      "learning_rate": 9.978480558223278e-05,
+      "loss": 0.7605764269828796,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6446601941747573,
+      "grad_norm": 1.1991835832595825,
+      "learning_rate": 9.978252402489692e-05,
+      "loss": 0.7618141174316406,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6455231930960086,
+      "grad_norm": 1.167635202407837,
+      "learning_rate": 9.978023046272983e-05,
+      "loss": 0.7071146368980408,
+      "step": 1496
+    },
+    {
+      "epoch": 0.64638619201726,
+      "grad_norm": 1.263371229171753,
+      "learning_rate": 9.977792489628456e-05,
+      "loss": 0.7559816837310791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 1.2108519077301025,
+      "learning_rate": 9.977560732611713e-05,
+      "loss": 0.7253944873809814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "eval_loss": 0.778505802154541,
+      "eval_runtime": 836.2317,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6481121898597627,
+      "grad_norm": 1.4794813394546509,
+      "learning_rate": 9.977327775278636e-05,
+      "loss": 0.8212620615959167,
+      "step": 1502
+    },
+    {
+      "epoch": 0.648975188781014,
+      "grad_norm": 1.2120516300201416,
+      "learning_rate": 9.977093617685404e-05,
+      "loss": 0.7220476865768433,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6498381877022654,
+      "grad_norm": 1.3129044771194458,
+      "learning_rate": 9.976858259888484e-05,
+      "loss": 0.7791266441345215,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6507011866235167,
+      "grad_norm": 1.146490454673767,
+      "learning_rate": 9.976621701944628e-05,
+      "loss": 0.7052001357078552,
+      "step": 1508
+    },
+    {
+      "epoch": 0.651564185544768,
+      "grad_norm": 1.1903722286224365,
+      "learning_rate": 9.976383943910885e-05,
+      "loss": 0.7135450839996338,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6524271844660194,
+      "grad_norm": 1.4106701612472534,
+      "learning_rate": 9.976144985844589e-05,
+      "loss": 0.8114889860153198,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6532901833872707,
+      "grad_norm": 1.245661973953247,
+      "learning_rate": 9.97590482780336e-05,
+      "loss": 0.7501562237739563,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6541531823085222,
+      "grad_norm": 1.2918543815612793,
+      "learning_rate": 9.975663469845114e-05,
+      "loss": 0.7697334289550781,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6550161812297735,
+      "grad_norm": 1.2699558734893799,
+      "learning_rate": 9.975420912028053e-05,
+      "loss": 0.7457168102264404,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6558791801510249,
+      "grad_norm": 1.3968263864517212,
+      "learning_rate": 9.975177154410668e-05,
+      "loss": 0.8368204236030579,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6567421790722762,
+      "grad_norm": 1.2161848545074463,
+      "learning_rate": 9.974932197051741e-05,
+      "loss": 0.7355788350105286,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6576051779935275,
+      "grad_norm": 1.129886507987976,
+      "learning_rate": 9.974686040010341e-05,
+      "loss": 0.8023449182510376,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6584681769147789,
+      "grad_norm": 1.1114557981491089,
+      "learning_rate": 9.974438683345829e-05,
+      "loss": 0.7965474724769592,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6593311758360302,
+      "grad_norm": 1.5741950273513794,
+      "learning_rate": 9.974190127117853e-05,
+      "loss": 0.7176838517189026,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 1.1439597606658936,
+      "learning_rate": 9.973940371386352e-05,
+      "loss": 0.7114462852478027,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6610571736785329,
+      "grad_norm": 1.273659348487854,
+      "learning_rate": 9.973689416211553e-05,
+      "loss": 0.754712700843811,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6619201725997842,
+      "grad_norm": 3.9874372482299805,
+      "learning_rate": 9.973437261653973e-05,
+      "loss": 0.6720641255378723,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6627831715210356,
+      "grad_norm": 1.2506502866744995,
+      "learning_rate": 9.973183907774417e-05,
+      "loss": 0.793043851852417,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6636461704422869,
+      "grad_norm": 1.43257737159729,
+      "learning_rate": 9.972929354633982e-05,
+      "loss": 0.7806380987167358,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6645091693635383,
+      "grad_norm": 1.4794938564300537,
+      "learning_rate": 9.97267360229405e-05,
+      "loss": 0.7176088094711304,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6653721682847896,
+      "grad_norm": 1.2398169040679932,
+      "learning_rate": 9.972416650816297e-05,
+      "loss": 0.7145742774009705,
+      "step": 1542
+    },
+    {
+      "epoch": 0.666235167206041,
+      "grad_norm": 1.1309953927993774,
+      "learning_rate": 9.972158500262683e-05,
+      "loss": 0.8025617003440857,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6670981661272923,
+      "grad_norm": 1.2479451894760132,
+      "learning_rate": 9.971899150695463e-05,
+      "loss": 0.7808589339256287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6679611650485436,
+      "grad_norm": 1.2538350820541382,
+      "learning_rate": 9.971638602177175e-05,
+      "loss": 0.7096936106681824,
+      "step": 1548
+    },
+    {
+      "epoch": 0.668824163969795,
+      "grad_norm": 1.1431525945663452,
+      "learning_rate": 9.97137685477065e-05,
+      "loss": 0.6529367566108704,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6696871628910464,
+      "grad_norm": 1.5966559648513794,
+      "learning_rate": 9.971113908539008e-05,
+      "loss": 0.7780346870422363,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6705501618122978,
+      "grad_norm": 1.186475157737732,
+      "learning_rate": 9.970849763545658e-05,
+      "loss": 0.6889166831970215,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6714131607335491,
+      "grad_norm": 1.1922476291656494,
+      "learning_rate": 9.970584419854295e-05,
+      "loss": 0.7154867649078369,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6722761596548005,
+      "grad_norm": 1.30132257938385,
+      "learning_rate": 9.970317877528906e-05,
+      "loss": 0.7419366240501404,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6731391585760518,
+      "grad_norm": 1.1539888381958008,
+      "learning_rate": 9.970050136633769e-05,
+      "loss": 0.7856602072715759,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6740021574973031,
+      "grad_norm": 1.3056161403656006,
+      "learning_rate": 9.969781197233445e-05,
+      "loss": 0.6834115386009216,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6748651564185545,
+      "grad_norm": 1.2899249792099,
+      "learning_rate": 9.96951105939279e-05,
+      "loss": 0.8490727543830872,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6757281553398058,
+      "grad_norm": 1.2182374000549316,
+      "learning_rate": 9.969239723176946e-05,
+      "loss": 0.7462526559829712,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6765911542610572,
+      "grad_norm": 1.1231273412704468,
+      "learning_rate": 9.968967188651344e-05,
+      "loss": 0.7145209908485413,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6774541531823085,
+      "grad_norm": 1.1776760816574097,
+      "learning_rate": 9.968693455881703e-05,
+      "loss": 0.7911707162857056,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6783171521035599,
+      "grad_norm": 1.3542319536209106,
+      "learning_rate": 9.968418524934035e-05,
+      "loss": 0.6946635246276855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6791801510248112,
+      "grad_norm": 1.3073060512542725,
+      "learning_rate": 9.96814239587464e-05,
+      "loss": 0.7597488164901733,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6800431499460625,
+      "grad_norm": 1.3205567598342896,
+      "learning_rate": 9.967865068770099e-05,
+      "loss": 0.6763246655464172,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6809061488673139,
+      "grad_norm": 1.1734344959259033,
+      "learning_rate": 9.967586543687295e-05,
+      "loss": 0.7560655474662781,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6817691477885652,
+      "grad_norm": 1.445770025253296,
+      "learning_rate": 9.967306820693388e-05,
+      "loss": 0.8050205707550049,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6826321467098166,
+      "grad_norm": 1.1185686588287354,
+      "learning_rate": 9.967025899855836e-05,
+      "loss": 0.6628211140632629,
+      "step": 1582
+    },
+    {
+      "epoch": 0.683495145631068,
+      "grad_norm": 1.3831934928894043,
+      "learning_rate": 9.966743781242379e-05,
+      "loss": 0.7684114575386047,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6843581445523194,
+      "grad_norm": 1.2417019605636597,
+      "learning_rate": 9.966460464921047e-05,
+      "loss": 0.7084140777587891,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6852211434735707,
+      "grad_norm": 1.2138086557388306,
+      "learning_rate": 9.966175950960163e-05,
+      "loss": 0.6817399263381958,
+      "step": 1588
+    },
+    {
+      "epoch": 0.686084142394822,
+      "grad_norm": 1.399778962135315,
+      "learning_rate": 9.965890239428339e-05,
+      "loss": 0.8422354459762573,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6869471413160734,
+      "grad_norm": 1.4348714351654053,
+      "learning_rate": 9.965603330394468e-05,
+      "loss": 0.7432681322097778,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6878101402373247,
+      "grad_norm": 1.192425012588501,
+      "learning_rate": 9.965315223927741e-05,
+      "loss": 0.7571184635162354,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6886731391585761,
+      "grad_norm": 1.2834038734436035,
+      "learning_rate": 9.96502592009763e-05,
+      "loss": 0.802443265914917,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6895361380798274,
+      "grad_norm": 1.1650444269180298,
+      "learning_rate": 9.964735418973901e-05,
+      "loss": 0.8113405704498291,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "grad_norm": 1.1714438199996948,
+      "learning_rate": 9.964443720626606e-05,
+      "loss": 0.7507118582725525,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "eval_loss": 0.7706249952316284,
+      "eval_runtime": 821.4894,
+      "eval_samples_per_second": 2.508,
+      "eval_steps_per_second": 2.508,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6912621359223301,
+      "grad_norm": 1.175459861755371,
+      "learning_rate": 9.964150825126088e-05,
+      "loss": 0.7422796487808228,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6921251348435814,
+      "grad_norm": 1.6418036222457886,
+      "learning_rate": 9.963856732542978e-05,
+      "loss": 0.7125853300094604,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6929881337648328,
+      "grad_norm": 1.3256275653839111,
+      "learning_rate": 9.963561442948193e-05,
+      "loss": 0.757859468460083,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6938511326860841,
+      "grad_norm": 1.19981050491333,
+      "learning_rate": 9.963264956412944e-05,
+      "loss": 0.6849292516708374,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6947141316073355,
+      "grad_norm": 1.0392169952392578,
+      "learning_rate": 9.962967273008723e-05,
+      "loss": 0.7636215090751648,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6955771305285868,
+      "grad_norm": 1.1106646060943604,
+      "learning_rate": 9.962668392807318e-05,
+      "loss": 0.7208617329597473,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6964401294498382,
+      "grad_norm": 1.1569621562957764,
+      "learning_rate": 9.962368315880801e-05,
+      "loss": 0.671805739402771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6973031283710895,
+      "grad_norm": 1.2157769203186035,
+      "learning_rate": 9.962067042301537e-05,
+      "loss": 0.7648762464523315,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6981661272923408,
+      "grad_norm": 1.4804017543792725,
+      "learning_rate": 9.961764572142173e-05,
+      "loss": 0.8490509390830994,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 1.0344431400299072,
+      "learning_rate": 9.961460905475651e-05,
+      "loss": 0.5950891971588135,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6998921251348436,
+      "grad_norm": 1.1109060049057007,
+      "learning_rate": 9.9611560423752e-05,
+      "loss": 0.7288329005241394,
+      "step": 1622
+    },
+    {
+      "epoch": 0.700755124056095,
+      "grad_norm": 1.096996545791626,
+      "learning_rate": 9.960849982914331e-05,
+      "loss": 0.7074542045593262,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7016181229773463,
+      "grad_norm": 1.3109341859817505,
+      "learning_rate": 9.960542727166853e-05,
+      "loss": 0.7348865270614624,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7024811218985976,
+      "grad_norm": 1.3402032852172852,
+      "learning_rate": 9.96023427520686e-05,
+      "loss": 0.7658320665359497,
+      "step": 1628
+    },
+    {
+      "epoch": 0.703344120819849,
+      "grad_norm": 1.3712561130523682,
+      "learning_rate": 9.959924627108732e-05,
+      "loss": 0.809951901435852,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7042071197411003,
+      "grad_norm": 1.2278292179107666,
+      "learning_rate": 9.959613782947139e-05,
+      "loss": 0.6834598779678345,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7050701186623517,
+      "grad_norm": 1.0510832071304321,
+      "learning_rate": 9.95930174279704e-05,
+      "loss": 0.7416030764579773,
+      "step": 1634
+    },
+    {
+      "epoch": 0.705933117583603,
+      "grad_norm": 1.130991816520691,
+      "learning_rate": 9.958988506733684e-05,
+      "loss": 0.7304092645645142,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7067961165048544,
+      "grad_norm": 1.3328672647476196,
+      "learning_rate": 9.958674074832603e-05,
+      "loss": 0.8230003118515015,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7076591154261057,
+      "grad_norm": 1.3842110633850098,
+      "learning_rate": 9.958358447169625e-05,
+      "loss": 0.8413107991218567,
+      "step": 1640
+    },
+    {
+      "epoch": 0.708522114347357,
+      "grad_norm": 1.3143812417984009,
+      "learning_rate": 9.958041623820858e-05,
+      "loss": 0.77432781457901,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7093851132686084,
+      "grad_norm": 1.4727009534835815,
+      "learning_rate": 9.957723604862704e-05,
+      "loss": 0.7563844323158264,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7102481121898597,
+      "grad_norm": 1.1892327070236206,
+      "learning_rate": 9.957404390371854e-05,
+      "loss": 0.7085853219032288,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.124180555343628,
+      "learning_rate": 9.957083980425282e-05,
+      "loss": 0.6837730407714844,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 1.2224774360656738,
+      "learning_rate": 9.956762375100255e-05,
+      "loss": 0.7255740165710449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7128371089536139,
+      "grad_norm": 1.360425591468811,
+      "learning_rate": 9.956439574474326e-05,
+      "loss": 0.7849710583686829,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7137001078748652,
+      "grad_norm": 1.2818151712417603,
+      "learning_rate": 9.956115578625338e-05,
+      "loss": 0.7669665813446045,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7145631067961165,
+      "grad_norm": 1.2415164709091187,
+      "learning_rate": 9.95579038763142e-05,
+      "loss": 0.7052890658378601,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7154261057173679,
+      "grad_norm": 1.3983269929885864,
+      "learning_rate": 9.955464001570992e-05,
+      "loss": 0.8300023674964905,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7162891046386192,
+      "grad_norm": 1.1591373682022095,
+      "learning_rate": 9.955136420522759e-05,
+      "loss": 0.7742894887924194,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7171521035598706,
+      "grad_norm": 1.2503128051757812,
+      "learning_rate": 9.954807644565715e-05,
+      "loss": 0.8458141684532166,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7180151024811219,
+      "grad_norm": 1.2954912185668945,
+      "learning_rate": 9.954477673779146e-05,
+      "loss": 0.7044742107391357,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7188781014023733,
+      "grad_norm": 1.1669788360595703,
+      "learning_rate": 9.954146508242619e-05,
+      "loss": 0.7917183041572571,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7197411003236246,
+      "grad_norm": 1.1466871500015259,
+      "learning_rate": 9.953814148035996e-05,
+      "loss": 0.7174562811851501,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7206040992448759,
+      "grad_norm": 1.367224097251892,
+      "learning_rate": 9.953480593239421e-05,
+      "loss": 0.8925880193710327,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7214670981661273,
+      "grad_norm": 1.049497365951538,
+      "learning_rate": 9.953145843933334e-05,
+      "loss": 0.70949387550354,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7223300970873786,
+      "grad_norm": 0.9995084404945374,
+      "learning_rate": 9.952809900198455e-05,
+      "loss": 0.6731935143470764,
+      "step": 1674
+    },
+    {
+      "epoch": 0.72319309600863,
+      "grad_norm": 1.100899577140808,
+      "learning_rate": 9.952472762115794e-05,
+      "loss": 0.7316275835037231,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7240560949298813,
+      "grad_norm": 1.2594135999679565,
+      "learning_rate": 9.952134429766655e-05,
+      "loss": 0.7379520535469055,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7249190938511327,
+      "grad_norm": 1.3124911785125732,
+      "learning_rate": 9.951794903232621e-05,
+      "loss": 0.7072247266769409,
+      "step": 1680
+    },
+    {
+      "epoch": 0.725782092772384,
+      "grad_norm": 1.212144374847412,
+      "learning_rate": 9.95145418259557e-05,
+      "loss": 0.745680034160614,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7266450916936353,
+      "grad_norm": 1.3913586139678955,
+      "learning_rate": 9.951112267937663e-05,
+      "loss": 0.796852707862854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7275080906148867,
+      "grad_norm": 1.3227289915084839,
+      "learning_rate": 9.950769159341353e-05,
+      "loss": 0.700892984867096,
+      "step": 1686
+    },
+    {
+      "epoch": 0.728371089536138,
+      "grad_norm": 1.4656381607055664,
+      "learning_rate": 9.950424856889378e-05,
+      "loss": 0.7973095178604126,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7292340884573895,
+      "grad_norm": 1.1698973178863525,
+      "learning_rate": 9.950079360664766e-05,
+      "loss": 0.7811378836631775,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7300970873786408,
+      "grad_norm": 1.201695203781128,
+      "learning_rate": 9.949732670750831e-05,
+      "loss": 0.753012478351593,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7309600862998922,
+      "grad_norm": 1.3465325832366943,
+      "learning_rate": 9.949384787231175e-05,
+      "loss": 0.7620642781257629,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7318230852211435,
+      "grad_norm": 1.2191137075424194,
+      "learning_rate": 9.94903571018969e-05,
+      "loss": 0.7479873299598694,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7326860841423948,
+      "grad_norm": 1.1412187814712524,
+      "learning_rate": 9.948685439710554e-05,
+      "loss": 0.7327720522880554,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "grad_norm": 1.310327410697937,
+      "learning_rate": 9.948333975878233e-05,
+      "loss": 0.761475682258606,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "eval_loss": 0.7649155259132385,
+      "eval_runtime": 828.3542,
+      "eval_samples_per_second": 2.487,
+      "eval_steps_per_second": 2.487,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7344120819848975,
+      "grad_norm": 1.3031405210494995,
+      "learning_rate": 9.947981318777482e-05,
+      "loss": 0.7564467191696167,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7352750809061489,
+      "grad_norm": 1.2672377824783325,
+      "learning_rate": 9.94762746849334e-05,
+      "loss": 0.7689576745033264,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7361380798274002,
+      "grad_norm": 1.1159595251083374,
+      "learning_rate": 9.947272425111138e-05,
+      "loss": 0.7923810482025146,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7370010787486516,
+      "grad_norm": 1.1190921068191528,
+      "learning_rate": 9.946916188716495e-05,
+      "loss": 0.6540790796279907,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 1.373832106590271,
+      "learning_rate": 9.946558759395312e-05,
+      "loss": 0.7267473936080933,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7387270765911542,
+      "grad_norm": 1.2346374988555908,
+      "learning_rate": 9.946200137233785e-05,
+      "loss": 0.7693060636520386,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7395900755124056,
+      "grad_norm": 1.1629668474197388,
+      "learning_rate": 9.945840322318391e-05,
+      "loss": 0.7751407027244568,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7404530744336569,
+      "grad_norm": 1.2268708944320679,
+      "learning_rate": 9.945479314735899e-05,
+      "loss": 0.7798545360565186,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7413160733549083,
+      "grad_norm": 0.9968738555908203,
+      "learning_rate": 9.945117114573366e-05,
+      "loss": 0.6128867268562317,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7421790722761596,
+      "grad_norm": 1.2135992050170898,
+      "learning_rate": 9.944753721918133e-05,
+      "loss": 0.7536954283714294,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7430420711974111,
+      "grad_norm": 1.1361218690872192,
+      "learning_rate": 9.94438913685783e-05,
+      "loss": 0.7034479379653931,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7439050701186624,
+      "grad_norm": 1.0728967189788818,
+      "learning_rate": 9.944023359480378e-05,
+      "loss": 0.7285382747650146,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7447680690399137,
+      "grad_norm": 1.3613945245742798,
+      "learning_rate": 9.94365638987398e-05,
+      "loss": 0.7865195274353027,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7456310679611651,
+      "grad_norm": 1.0985898971557617,
+      "learning_rate": 9.94328822812713e-05,
+      "loss": 0.7511224746704102,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7464940668824164,
+      "grad_norm": 1.1506208181381226,
+      "learning_rate": 9.942918874328606e-05,
+      "loss": 0.732453465461731,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7473570658036678,
+      "grad_norm": 1.0995100736618042,
+      "learning_rate": 9.94254832856748e-05,
+      "loss": 0.702409029006958,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7482200647249191,
+      "grad_norm": 1.0075604915618896,
+      "learning_rate": 9.942176590933106e-05,
+      "loss": 0.7515703439712524,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7490830636461704,
+      "grad_norm": 1.3360812664031982,
+      "learning_rate": 9.941803661515126e-05,
+      "loss": 0.6910237073898315,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7499460625674218,
+      "grad_norm": 1.0172319412231445,
+      "learning_rate": 9.941429540403471e-05,
+      "loss": 0.7185606956481934,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7508090614886731,
+      "grad_norm": 1.2721658945083618,
+      "learning_rate": 9.941054227688358e-05,
+      "loss": 0.7136536836624146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7516720604099245,
+      "grad_norm": 1.0775413513183594,
+      "learning_rate": 9.940677723460292e-05,
+      "loss": 0.7595619559288025,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7525350593311758,
+      "grad_norm": 1.3436479568481445,
+      "learning_rate": 9.940300027810067e-05,
+      "loss": 0.7602618336677551,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7533980582524272,
+      "grad_norm": 1.2611958980560303,
+      "learning_rate": 9.93992114082876e-05,
+      "loss": 0.7954604029655457,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7542610571736785,
+      "grad_norm": 1.2794841527938843,
+      "learning_rate": 9.93954106260774e-05,
+      "loss": 0.7661190629005432,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7551240560949298,
+      "grad_norm": 1.3042595386505127,
+      "learning_rate": 9.939159793238661e-05,
+      "loss": 0.6855312585830688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7559870550161812,
+      "grad_norm": 1.2799923419952393,
+      "learning_rate": 9.938777332813463e-05,
+      "loss": 0.7970905900001526,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7568500539374325,
+      "grad_norm": 1.169205904006958,
+      "learning_rate": 9.938393681424375e-05,
+      "loss": 0.7072843313217163,
+      "step": 1754
+    },
+    {
+      "epoch": 0.757713052858684,
+      "grad_norm": 1.3833354711532593,
+      "learning_rate": 9.938008839163913e-05,
+      "loss": 0.7134629487991333,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7585760517799353,
+      "grad_norm": 1.223452091217041,
+      "learning_rate": 9.93762280612488e-05,
+      "loss": 0.7873107194900513,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7594390507011867,
+      "grad_norm": 1.2740365266799927,
+      "learning_rate": 9.937235582400367e-05,
+      "loss": 0.7680654525756836,
+      "step": 1760
+    },
+    {
+      "epoch": 0.760302049622438,
+      "grad_norm": 1.2564697265625,
+      "learning_rate": 9.93684716808375e-05,
+      "loss": 0.734126627445221,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7611650485436893,
+      "grad_norm": 1.3858847618103027,
+      "learning_rate": 9.936457563268694e-05,
+      "loss": 0.7972320914268494,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7620280474649407,
+      "grad_norm": 1.2705003023147583,
+      "learning_rate": 9.93606676804915e-05,
+      "loss": 0.7904276847839355,
+      "step": 1766
+    },
+    {
+      "epoch": 0.762891046386192,
+      "grad_norm": 1.210766077041626,
+      "learning_rate": 9.935674782519358e-05,
+      "loss": 0.7542425990104675,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7637540453074434,
+      "grad_norm": 1.4367541074752808,
+      "learning_rate": 9.93528160677384e-05,
+      "loss": 0.7975004315376282,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7646170442286947,
+      "grad_norm": 1.0994079113006592,
+      "learning_rate": 9.934887240907414e-05,
+      "loss": 0.7537807822227478,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7654800431499461,
+      "grad_norm": 1.111660122871399,
+      "learning_rate": 9.934491685015174e-05,
+      "loss": 0.7333855628967285,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7663430420711974,
+      "grad_norm": 1.2879043817520142,
+      "learning_rate": 9.934094939192508e-05,
+      "loss": 0.7472840547561646,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7672060409924487,
+      "grad_norm": 1.3618478775024414,
+      "learning_rate": 9.933697003535094e-05,
+      "loss": 0.7644904255867004,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7680690399137001,
+      "grad_norm": 1.1235666275024414,
+      "learning_rate": 9.933297878138887e-05,
+      "loss": 0.7872052192687988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7689320388349514,
+      "grad_norm": 1.3741775751113892,
+      "learning_rate": 9.932897563100136e-05,
+      "loss": 0.7343124747276306,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7697950377562028,
+      "grad_norm": 1.2115039825439453,
+      "learning_rate": 9.932496058515375e-05,
+      "loss": 0.7486474514007568,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7706580366774541,
+      "grad_norm": 1.183457851409912,
+      "learning_rate": 9.932093364481426e-05,
+      "loss": 0.7333025932312012,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7715210355987056,
+      "grad_norm": 1.2841869592666626,
+      "learning_rate": 9.931689481095394e-05,
+      "loss": 0.7241485118865967,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7723840345199569,
+      "grad_norm": 1.2312977313995361,
+      "learning_rate": 9.931284408454678e-05,
+      "loss": 0.7510396242141724,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7732470334412082,
+      "grad_norm": 1.144895315170288,
+      "learning_rate": 9.930878146656956e-05,
+      "loss": 0.7108346819877625,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7741100323624596,
+      "grad_norm": 1.18603515625,
+      "learning_rate": 9.9304706958002e-05,
+      "loss": 0.7687607407569885,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7749730312837109,
+      "grad_norm": 1.2827935218811035,
+      "learning_rate": 9.93006205598266e-05,
+      "loss": 0.7826682329177856,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7758360302049623,
+      "grad_norm": 1.164852261543274,
+      "learning_rate": 9.929652227302881e-05,
+      "loss": 0.6970428824424744,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 1.0439956188201904,
+      "learning_rate": 9.92924120985969e-05,
+      "loss": 0.7238826751708984,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "eval_loss": 0.7598537802696228,
+      "eval_runtime": 838.9645,
+      "eval_samples_per_second": 2.455,
+      "eval_steps_per_second": 2.455,
+      "step": 1800
+    },
+    {
+      "epoch": 0.777562028047465,
+      "grad_norm": 1.2464938163757324,
+      "learning_rate": 9.928829003752202e-05,
+      "loss": 0.7723641395568848,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7784250269687163,
+      "grad_norm": 1.2445050477981567,
+      "learning_rate": 9.92841560907982e-05,
+      "loss": 0.7154937982559204,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7792880258899676,
+      "grad_norm": 1.2235661745071411,
+      "learning_rate": 9.928001025942231e-05,
+      "loss": 0.8325800895690918,
+      "step": 1806
+    },
+    {
+      "epoch": 0.780151024811219,
+      "grad_norm": 1.268066644668579,
+      "learning_rate": 9.927585254439409e-05,
+      "loss": 0.7093122601509094,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7810140237324703,
+      "grad_norm": 1.1731390953063965,
+      "learning_rate": 9.927168294671618e-05,
+      "loss": 0.7601775527000427,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7818770226537217,
+      "grad_norm": 1.0685608386993408,
+      "learning_rate": 9.926750146739403e-05,
+      "loss": 0.6889236569404602,
+      "step": 1812
+    },
+    {
+      "epoch": 0.782740021574973,
+      "grad_norm": 1.1098639965057373,
+      "learning_rate": 9.926330810743602e-05,
+      "loss": 0.7397196888923645,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7836030204962244,
+      "grad_norm": 1.3606436252593994,
+      "learning_rate": 9.925910286785332e-05,
+      "loss": 0.7180029153823853,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7844660194174757,
+      "grad_norm": 1.143418550491333,
+      "learning_rate": 9.925488574966004e-05,
+      "loss": 0.7039036750793457,
+      "step": 1818
+    },
+    {
+      "epoch": 0.785329018338727,
+      "grad_norm": 1.180968999862671,
+      "learning_rate": 9.92506567538731e-05,
+      "loss": 0.741872251033783,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7861920172599784,
+      "grad_norm": 1.1953548192977905,
+      "learning_rate": 9.924641588151231e-05,
+      "loss": 0.7580198049545288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7870550161812297,
+      "grad_norm": 1.3049067258834839,
+      "learning_rate": 9.924216313360033e-05,
+      "loss": 0.7479216456413269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7879180151024812,
+      "grad_norm": 1.3962669372558594,
+      "learning_rate": 9.92378985111627e-05,
+      "loss": 0.7382357716560364,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7887810140237325,
+      "grad_norm": 1.1519217491149902,
+      "learning_rate": 9.92336220152278e-05,
+      "loss": 0.6356467604637146,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7896440129449838,
+      "grad_norm": 1.2544089555740356,
+      "learning_rate": 9.922933364682689e-05,
+      "loss": 0.7674193978309631,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7905070118662352,
+      "grad_norm": 1.1358295679092407,
+      "learning_rate": 9.922503340699412e-05,
+      "loss": 0.7083898782730103,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7913700107874865,
+      "grad_norm": 1.1744481325149536,
+      "learning_rate": 9.922072129676644e-05,
+      "loss": 0.6273066401481628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7922330097087379,
+      "grad_norm": 1.20509672164917,
+      "learning_rate": 9.92163973171837e-05,
+      "loss": 0.808181643486023,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7930960086299892,
+      "grad_norm": 1.2035597562789917,
+      "learning_rate": 9.921206146928863e-05,
+      "loss": 0.6631553769111633,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7939590075512406,
+      "grad_norm": 1.058695912361145,
+      "learning_rate": 9.920771375412678e-05,
+      "loss": 0.72016841173172,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7948220064724919,
+      "grad_norm": 1.035791039466858,
+      "learning_rate": 9.92033541727466e-05,
+      "loss": 0.7240060567855835,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7956850053937432,
+      "grad_norm": 1.1845753192901611,
+      "learning_rate": 9.919898272619935e-05,
+      "loss": 0.743224561214447,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7965480043149946,
+      "grad_norm": 1.2942636013031006,
+      "learning_rate": 9.919459941553922e-05,
+      "loss": 0.7064419388771057,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7974110032362459,
+      "grad_norm": 1.2736715078353882,
+      "learning_rate": 9.91902042418232e-05,
+      "loss": 0.805932343006134,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7982740021574973,
+      "grad_norm": 1.2803614139556885,
+      "learning_rate": 9.918579720611119e-05,
+      "loss": 0.8050802946090698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7991370010787486,
+      "grad_norm": 1.2225072383880615,
+      "learning_rate": 9.91813783094659e-05,
+      "loss": 0.7013180255889893,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3375571966171265,
+      "learning_rate": 9.917694755295294e-05,
+      "loss": 0.7415283918380737,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8008629989212513,
+      "grad_norm": 1.2209177017211914,
+      "learning_rate": 9.917250493764079e-05,
+      "loss": 0.7264692783355713,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8017259978425026,
+      "grad_norm": 1.1158864498138428,
+      "learning_rate": 9.916805046460074e-05,
+      "loss": 0.7083719372749329,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8025889967637541,
+      "grad_norm": 1.171000361442566,
+      "learning_rate": 9.916358413490696e-05,
+      "loss": 0.820396900177002,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8034519956850054,
+      "grad_norm": 0.9942941069602966,
+      "learning_rate": 9.91591059496365e-05,
+      "loss": 0.6754014492034912,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8043149946062568,
+      "grad_norm": 1.239795207977295,
+      "learning_rate": 9.915461590986926e-05,
+      "loss": 0.813154399394989,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8051779935275081,
+      "grad_norm": 1.076019525527954,
+      "learning_rate": 9.915011401668796e-05,
+      "loss": 0.745580792427063,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8060409924487595,
+      "grad_norm": 1.2134130001068115,
+      "learning_rate": 9.914560027117826e-05,
+      "loss": 0.7564949989318848,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8069039913700108,
+      "grad_norm": 1.1774994134902954,
+      "learning_rate": 9.914107467442859e-05,
+      "loss": 0.665183424949646,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8077669902912621,
+      "grad_norm": 1.0606465339660645,
+      "learning_rate": 9.91365372275303e-05,
+      "loss": 0.7597744464874268,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8086299892125135,
+      "grad_norm": 1.397161841392517,
+      "learning_rate": 9.913198793157757e-05,
+      "loss": 0.7453591227531433,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8094929881337648,
+      "grad_norm": 1.250003695487976,
+      "learning_rate": 9.912742678766742e-05,
+      "loss": 0.715792179107666,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8103559870550162,
+      "grad_norm": 1.2343473434448242,
+      "learning_rate": 9.912285379689979e-05,
+      "loss": 0.7094268798828125,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8112189859762675,
+      "grad_norm": 1.453764796257019,
+      "learning_rate": 9.911826896037739e-05,
+      "loss": 0.6652816534042358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8120819848975189,
+      "grad_norm": 1.030903935432434,
+      "learning_rate": 9.911367227920587e-05,
+      "loss": 0.6764529347419739,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8129449838187702,
+      "grad_norm": 1.2725969552993774,
+      "learning_rate": 9.91090637544937e-05,
+      "loss": 0.7043389678001404,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8138079827400215,
+      "grad_norm": 1.1610769033432007,
+      "learning_rate": 9.910444338735216e-05,
+      "loss": 0.6960558295249939,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8146709816612729,
+      "grad_norm": 1.2049180269241333,
+      "learning_rate": 9.909981117889548e-05,
+      "loss": 0.8029319643974304,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 1.1671873331069946,
+      "learning_rate": 9.909516713024068e-05,
+      "loss": 0.683114767074585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8163969795037757,
+      "grad_norm": 1.247768759727478,
+      "learning_rate": 9.909051124250764e-05,
+      "loss": 0.7193703055381775,
+      "step": 1892
+    },
+    {
+      "epoch": 0.817259978425027,
+      "grad_norm": 1.1376357078552246,
+      "learning_rate": 9.908584351681911e-05,
+      "loss": 0.7957084774971008,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8181229773462784,
+      "grad_norm": 1.1918323040008545,
+      "learning_rate": 9.90811639543007e-05,
+      "loss": 0.7168112993240356,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8189859762675297,
+      "grad_norm": 1.1249650716781616,
+      "learning_rate": 9.907647255608087e-05,
+      "loss": 0.7519245147705078,
+      "step": 1898
+    },
+    {
+      "epoch": 0.819848975188781,
+      "grad_norm": 1.0688576698303223,
+      "learning_rate": 9.907176932329092e-05,
+      "loss": 0.7496176362037659,
+      "step": 1900
+    },
+    {
+      "epoch": 0.819848975188781,
+      "eval_loss": 0.7533769607543945,
+      "eval_runtime": 830.2894,
+      "eval_samples_per_second": 2.481,
+      "eval_steps_per_second": 2.481,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8207119741100324,
+      "grad_norm": 0.9264270663261414,
+      "learning_rate": 9.906705425706499e-05,
+      "loss": 0.639653205871582,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8215749730312837,
+      "grad_norm": 1.2231628894805908,
+      "learning_rate": 9.906232735854016e-05,
+      "loss": 0.7555551528930664,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8224379719525351,
+      "grad_norm": 1.3369156122207642,
+      "learning_rate": 9.905758862885624e-05,
+      "loss": 0.7113592028617859,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8233009708737864,
+      "grad_norm": 1.217238187789917,
+      "learning_rate": 9.905283806915597e-05,
+      "loss": 0.6823481321334839,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8241639697950378,
+      "grad_norm": 1.2462831735610962,
+      "learning_rate": 9.904807568058495e-05,
+      "loss": 0.7347100377082825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8250269687162891,
+      "grad_norm": 1.2507221698760986,
+      "learning_rate": 9.904330146429157e-05,
+      "loss": 0.7981513738632202,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8258899676375404,
+      "grad_norm": 1.2246814966201782,
+      "learning_rate": 9.903851542142715e-05,
+      "loss": 0.759779155254364,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8267529665587918,
+      "grad_norm": 1.1746087074279785,
+      "learning_rate": 9.903371755314579e-05,
+      "loss": 0.741083562374115,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8276159654800431,
+      "grad_norm": 1.1913539171218872,
+      "learning_rate": 9.902890786060448e-05,
+      "loss": 0.7491101622581482,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8284789644012945,
+      "grad_norm": 1.2342501878738403,
+      "learning_rate": 9.902408634496307e-05,
+      "loss": 0.7655978202819824,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8293419633225458,
+      "grad_norm": 1.075892686843872,
+      "learning_rate": 9.901925300738424e-05,
+      "loss": 0.7199848890304565,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8302049622437971,
+      "grad_norm": 1.1517611742019653,
+      "learning_rate": 9.901440784903353e-05,
+      "loss": 0.8325693011283875,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8310679611650486,
+      "grad_norm": 1.2084628343582153,
+      "learning_rate": 9.900955087107933e-05,
+      "loss": 0.7674157023429871,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8319309600862999,
+      "grad_norm": 1.334679365158081,
+      "learning_rate": 9.900468207469287e-05,
+      "loss": 0.7415181398391724,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8327939590075513,
+      "grad_norm": 1.0505887269973755,
+      "learning_rate": 9.899980146104824e-05,
+      "loss": 0.672336220741272,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8336569579288026,
+      "grad_norm": 1.415797233581543,
+      "learning_rate": 9.899490903132238e-05,
+      "loss": 0.8077969551086426,
+      "step": 1932
+    },
+    {
+      "epoch": 0.834519956850054,
+      "grad_norm": 1.1793439388275146,
+      "learning_rate": 9.899000478669507e-05,
+      "loss": 0.8053045272827148,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8353829557713053,
+      "grad_norm": 1.196115493774414,
+      "learning_rate": 9.898508872834896e-05,
+      "loss": 0.7585914731025696,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8362459546925566,
+      "grad_norm": 1.2181763648986816,
+      "learning_rate": 9.898016085746953e-05,
+      "loss": 0.691072940826416,
+      "step": 1938
+    },
+    {
+      "epoch": 0.837108953613808,
+      "grad_norm": 1.2929240465164185,
+      "learning_rate": 9.897522117524512e-05,
+      "loss": 0.7723121643066406,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8379719525350593,
+      "grad_norm": 1.2859764099121094,
+      "learning_rate": 9.89702696828669e-05,
+      "loss": 0.8236626386642456,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8388349514563107,
+      "grad_norm": 1.2043784856796265,
+      "learning_rate": 9.89653063815289e-05,
+      "loss": 0.7315511703491211,
+      "step": 1944
+    },
+    {
+      "epoch": 0.839697950377562,
+      "grad_norm": 1.2063398361206055,
+      "learning_rate": 9.896033127242802e-05,
+      "loss": 0.6707082986831665,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8405609492988134,
+      "grad_norm": 1.1906715631484985,
+      "learning_rate": 9.895534435676393e-05,
+      "loss": 0.6920633912086487,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 1.122567892074585,
+      "learning_rate": 9.895034563573926e-05,
+      "loss": 0.7134768962860107,
+      "step": 1950
+    },
+    {
+      "epoch": 0.842286947141316,
+      "grad_norm": 1.0681787729263306,
+      "learning_rate": 9.894533511055943e-05,
+      "loss": 0.6333967447280884,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8431499460625674,
+      "grad_norm": 1.5216220617294312,
+      "learning_rate": 9.894031278243266e-05,
+      "loss": 0.7744026780128479,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8440129449838187,
+      "grad_norm": 1.0934535264968872,
+      "learning_rate": 9.89352786525701e-05,
+      "loss": 0.7678086757659912,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8448759439050701,
+      "grad_norm": 1.223285436630249,
+      "learning_rate": 9.89302327221857e-05,
+      "loss": 0.7344630360603333,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8457389428263214,
+      "grad_norm": 1.1123322248458862,
+      "learning_rate": 9.892517499249623e-05,
+      "loss": 0.76033616065979,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8466019417475729,
+      "grad_norm": 1.0472193956375122,
+      "learning_rate": 9.89201054647214e-05,
+      "loss": 0.7176017761230469,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8474649406688242,
+      "grad_norm": 1.8538368940353394,
+      "learning_rate": 9.891502414008365e-05,
+      "loss": 0.7220107316970825,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8483279395900755,
+      "grad_norm": 1.2280668020248413,
+      "learning_rate": 9.890993101980834e-05,
+      "loss": 0.7399556636810303,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8491909385113269,
+      "grad_norm": 1.1482552289962769,
+      "learning_rate": 9.890482610512364e-05,
+      "loss": 0.7700466513633728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8500539374325782,
+      "grad_norm": 1.1877986192703247,
+      "learning_rate": 9.889970939726061e-05,
+      "loss": 0.6666419506072998,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8509169363538296,
+      "grad_norm": 1.407792329788208,
+      "learning_rate": 9.88945808974531e-05,
+      "loss": 0.7883278131484985,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8517799352750809,
+      "grad_norm": 1.1426323652267456,
+      "learning_rate": 9.888944060693781e-05,
+      "loss": 0.7441611289978027,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8526429341963323,
+      "grad_norm": 1.2660622596740723,
+      "learning_rate": 9.888428852695432e-05,
+      "loss": 0.6464171409606934,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8535059331175836,
+      "grad_norm": 1.1639022827148438,
+      "learning_rate": 9.887912465874502e-05,
+      "loss": 0.7242478728294373,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 1.2546422481536865,
+      "learning_rate": 9.887394900355516e-05,
+      "loss": 0.7431875467300415,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8552319309600863,
+      "grad_norm": 1.1057130098342896,
+      "learning_rate": 9.886876156263282e-05,
+      "loss": 0.7662370204925537,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8560949298813376,
+      "grad_norm": 1.1017125844955444,
+      "learning_rate": 9.886356233722894e-05,
+      "loss": 0.7120938897132874,
+      "step": 1984
+    },
+    {
+      "epoch": 0.856957928802589,
+      "grad_norm": 1.3548704385757446,
+      "learning_rate": 9.885835132859727e-05,
+      "loss": 0.789219856262207,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8578209277238403,
+      "grad_norm": 1.1920398473739624,
+      "learning_rate": 9.885312853799443e-05,
+      "loss": 0.7599530816078186,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8586839266450917,
+      "grad_norm": 1.170946478843689,
+      "learning_rate": 9.884789396667988e-05,
+      "loss": 0.8315809369087219,
+      "step": 1990
+    },
+    {
+      "epoch": 0.859546925566343,
+      "grad_norm": 1.2530180215835571,
+      "learning_rate": 9.88426476159159e-05,
+      "loss": 0.713994026184082,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8604099244875943,
+      "grad_norm": 1.1056885719299316,
+      "learning_rate": 9.883738948696765e-05,
+      "loss": 0.6656507849693298,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8612729234088458,
+      "grad_norm": 1.169382929801941,
+      "learning_rate": 9.883211958110308e-05,
+      "loss": 0.7164257764816284,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8621359223300971,
+      "grad_norm": 1.1970325708389282,
+      "learning_rate": 9.882683789959299e-05,
+      "loss": 0.6885558366775513,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "grad_norm": 1.1700761318206787,
+      "learning_rate": 9.882154444371107e-05,
+      "loss": 0.7265812754631042,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "eval_loss": 0.7487027645111084,
+      "eval_runtime": 825.141,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.085339919257815e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2000/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-2000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/README.md b/sft_devstral_24B/checkpoints/checkpoint-2500/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-2500/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-2500/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..5b5f29f2336e672cb8594c10b90645d03787f5e9
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6662a9047a1361809c0185e7b56691647d652fb4382248cd93082d3d414cccb4
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-2500/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..99c260c5e5494d79d169d7706371756032cb1f65
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e94c2bbf28c8c1a731dfa40610e905a42511196ee74c7ba0e4172c06c06eb1d2
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-2500/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..76cf5a816c6967a21e7e2f3fd749dcd5525463ec
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07b985c4a415c8b2261a9cb466d19c38f17bfc709c9097c70384d541f09c4622
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-2500/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0f0857d00870ade6773b23202cd43f39303f7737
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e6c35cfc5291bd34cd8a1e5ad2eb46c14099936b3b296d64c77317879166e9d
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-2500/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e443491cb28e62e1a94cd844e086b19a2a3e1be
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/trainer_state.json
@@ -0,0 +1,8993 @@
+{
+  "best_global_step": 2500,
+  "best_metric": 0.7294915914535522,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-2500",
+  "epoch": 1.0785329018338727,
+  "eval_steps": 100,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4323624595469256,
+      "grad_norm": 1.5475730895996094,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.7443642616271973,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4332254584681769,
+      "grad_norm": 1.4621447324752808,
+      "learning_rate": 9.011680143755616e-05,
+      "loss": 0.8210510015487671,
+      "step": 1004
+    },
+    {
+      "epoch": 0.43408845738942825,
+      "grad_norm": 1.3651922941207886,
+      "learning_rate": 9.029649595687331e-05,
+      "loss": 0.8130766153335571,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4349514563106796,
+      "grad_norm": 1.3587357997894287,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.79145747423172,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 1.464652180671692,
+      "learning_rate": 9.065588499550764e-05,
+      "loss": 0.8286946415901184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4366774541531823,
+      "grad_norm": 1.3912625312805176,
+      "learning_rate": 9.083557951482479e-05,
+      "loss": 0.717319667339325,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4375404530744337,
+      "grad_norm": 1.2045491933822632,
+      "learning_rate": 9.101527403414196e-05,
+      "loss": 0.7294418215751648,
+      "step": 1014
+    },
+    {
+      "epoch": 0.438403451995685,
+      "grad_norm": 1.3449069261550903,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.7366700768470764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43926645091693634,
+      "grad_norm": 2.901384115219116,
+      "learning_rate": 9.137466307277629e-05,
+      "loss": 0.8551293611526489,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 1.6648688316345215,
+      "learning_rate": 9.155435759209344e-05,
+      "loss": 0.8944112658500671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44099244875943905,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 9.17340521114106e-05,
+      "loss": 0.7644214630126953,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4418554476806904,
+      "grad_norm": 1.4643231630325317,
+      "learning_rate": 9.191374663072777e-05,
+      "loss": 0.8644799590110779,
+      "step": 1024
+    },
+    {
+      "epoch": 0.44271844660194176,
+      "grad_norm": 1.4690274000167847,
+      "learning_rate": 9.209344115004492e-05,
+      "loss": 0.8281022906303406,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4435814455231931,
+      "grad_norm": 1.3321986198425293,
+      "learning_rate": 9.227313566936209e-05,
+      "loss": 0.7432141900062561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.8143638372421265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4453074433656958,
+      "grad_norm": 1.747295618057251,
+      "learning_rate": 9.263252470799642e-05,
+      "loss": 0.8102239370346069,
+      "step": 1032
+    },
+    {
+      "epoch": 0.44617044228694713,
+      "grad_norm": 1.3273073434829712,
+      "learning_rate": 9.281221922731357e-05,
+      "loss": 0.7768088579177856,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4470334412081985,
+      "grad_norm": 1.5019903182983398,
+      "learning_rate": 9.299191374663073e-05,
+      "loss": 0.7886938452720642,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44789644012944985,
+      "grad_norm": 1.6711527109146118,
+      "learning_rate": 9.31716082659479e-05,
+      "loss": 0.8152258396148682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 1.8504793643951416,
+      "learning_rate": 9.335130278526505e-05,
+      "loss": 0.8339349627494812,
+      "step": 1040
+    },
+    {
+      "epoch": 0.44962243797195256,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 9.353099730458221e-05,
+      "loss": 0.7798911929130554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.45048543689320386,
+      "grad_norm": 1.3361700773239136,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.7676728963851929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4513484358144552,
+      "grad_norm": 1.338934063911438,
+      "learning_rate": 9.389038634321654e-05,
+      "loss": 0.7729206085205078,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4522114347357066,
+      "grad_norm": 1.4347442388534546,
+      "learning_rate": 9.40700808625337e-05,
+      "loss": 0.7923640608787537,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 1.4451628923416138,
+      "learning_rate": 9.424977538185086e-05,
+      "loss": 0.8068996071815491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4539374325782093,
+      "grad_norm": 1.1904500722885132,
+      "learning_rate": 9.442946990116803e-05,
+      "loss": 0.7532910704612732,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45480043149946064,
+      "grad_norm": 1.4103976488113403,
+      "learning_rate": 9.460916442048518e-05,
+      "loss": 0.8236854672431946,
+      "step": 1054
+    },
+    {
+      "epoch": 0.455663430420712,
+      "grad_norm": 1.3762754201889038,
+      "learning_rate": 9.478885893980234e-05,
+      "loss": 0.7532716393470764,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4565264293419633,
+      "grad_norm": 1.461421012878418,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.8506031632423401,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 1.7207789421081543,
+      "learning_rate": 9.514824797843667e-05,
+      "loss": 0.7824664115905762,
+      "step": 1060
+    },
+    {
+      "epoch": 0.458252427184466,
+      "grad_norm": 1.5946974754333496,
+      "learning_rate": 9.532794249775382e-05,
+      "loss": 0.8674250245094299,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4591154261057174,
+      "grad_norm": 1.3552565574645996,
+      "learning_rate": 9.550763701707099e-05,
+      "loss": 0.7851845622062683,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45997842502696873,
+      "grad_norm": 1.414617657661438,
+      "learning_rate": 9.568733153638815e-05,
+      "loss": 0.8890700936317444,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4608414239482201,
+      "grad_norm": 1.3322749137878418,
+      "learning_rate": 9.58670260557053e-05,
+      "loss": 0.8226593732833862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 1.41420316696167,
+      "learning_rate": 9.604672057502247e-05,
+      "loss": 0.7663303017616272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.46256742179072274,
+      "grad_norm": 1.324967384338379,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.7122516632080078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4634304207119741,
+      "grad_norm": 1.2046154737472534,
+      "learning_rate": 9.640610961365679e-05,
+      "loss": 0.7177618145942688,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46429341963322546,
+      "grad_norm": 1.5221766233444214,
+      "learning_rate": 9.658580413297395e-05,
+      "loss": 0.7618930339813232,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4651564185544768,
+      "grad_norm": 1.3517179489135742,
+      "learning_rate": 9.67654986522911e-05,
+      "loss": 0.7483780384063721,
+      "step": 1078
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.747865080833435,
+      "learning_rate": 9.694519317160827e-05,
+      "loss": 0.8507670164108276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4668824163969795,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 9.712488769092543e-05,
+      "loss": 0.7893540859222412,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46774541531823083,
+      "grad_norm": 1.6290462017059326,
+      "learning_rate": 9.730458221024258e-05,
+      "loss": 0.7891272306442261,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4686084142394822,
+      "grad_norm": 1.6605802774429321,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.841018795967102,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46947141316073354,
+      "grad_norm": 1.388964295387268,
+      "learning_rate": 9.766397124887691e-05,
+      "loss": 0.8099003434181213,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 1.3901978731155396,
+      "learning_rate": 9.784366576819407e-05,
+      "loss": 0.7177658677101135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.47119741100323626,
+      "grad_norm": 1.3341423273086548,
+      "learning_rate": 9.802336028751123e-05,
+      "loss": 0.8285676836967468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4720604099244876,
+      "grad_norm": 1.5451914072036743,
+      "learning_rate": 9.82030548068284e-05,
+      "loss": 0.7582666873931885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47292340884573897,
+      "grad_norm": 1.5395640134811401,
+      "learning_rate": 9.838274932614556e-05,
+      "loss": 0.7293475270271301,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47378640776699027,
+      "grad_norm": 1.5005141496658325,
+      "learning_rate": 9.856244384546271e-05,
+      "loss": 0.8858493566513062,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 1.3562248945236206,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.7176282405853271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "eval_loss": 0.8152187466621399,
+      "eval_runtime": 835.0394,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.475512405609493,
+      "grad_norm": 1.5226716995239258,
+      "learning_rate": 9.892183288409704e-05,
+      "loss": 0.8048524856567383,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47637540453074434,
+      "grad_norm": 1.3544842004776,
+      "learning_rate": 9.91015274034142e-05,
+      "loss": 0.7954329252243042,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4772384034519957,
+      "grad_norm": 1.5169771909713745,
+      "learning_rate": 9.928122192273136e-05,
+      "loss": 0.804015040397644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47810140237324705,
+      "grad_norm": 1.4011081457138062,
+      "learning_rate": 9.946091644204852e-05,
+      "loss": 0.7652344703674316,
+      "step": 1108
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 1.4163414239883423,
+      "learning_rate": 9.964061096136569e-05,
+      "loss": 0.8501335382461548,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4798274002157497,
+      "grad_norm": 1.7406005859375,
+      "learning_rate": 9.982030548068284e-05,
+      "loss": 0.9024044871330261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.48069039913700107,
+      "grad_norm": 1.1490557193756104,
+      "learning_rate": 0.0001,
+      "loss": 0.7308287024497986,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4815533980582524,
+      "grad_norm": 1.2512775659561157,
+      "learning_rate": 9.99999939713627e-05,
+      "loss": 0.7273974418640137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4824163969795038,
+      "grad_norm": 1.3944242000579834,
+      "learning_rate": 9.999997588545227e-05,
+      "loss": 0.8071790337562561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 1.2209112644195557,
+      "learning_rate": 9.999994574227303e-05,
+      "loss": 0.6586678624153137,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4841423948220065,
+      "grad_norm": 1.7328095436096191,
+      "learning_rate": 9.999990354183228e-05,
+      "loss": 0.8529233336448669,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4850053937432578,
+      "grad_norm": 1.3309327363967896,
+      "learning_rate": 9.999984928414021e-05,
+      "loss": 0.7529556155204773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.48586839266450915,
+      "grad_norm": 1.22034752368927,
+      "learning_rate": 9.999978296920987e-05,
+      "loss": 0.6780388355255127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4867313915857605,
+      "grad_norm": 1.975162386894226,
+      "learning_rate": 9.999970459705727e-05,
+      "loss": 0.7941693067550659,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 1.4939875602722168,
+      "learning_rate": 9.999961416770132e-05,
+      "loss": 0.7895156145095825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4884573894282632,
+      "grad_norm": 1.4237034320831299,
+      "learning_rate": 9.999951168116379e-05,
+      "loss": 0.7662149667739868,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4893203883495146,
+      "grad_norm": 1.24979567527771,
+      "learning_rate": 9.999939713746942e-05,
+      "loss": 0.8079231381416321,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49018338727076594,
+      "grad_norm": 1.2252353429794312,
+      "learning_rate": 9.999927053664585e-05,
+      "loss": 0.7616434693336487,
+      "step": 1136
+    },
+    {
+      "epoch": 0.49104638619201724,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.999913187872357e-05,
+      "loss": 0.8576793074607849,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 1.2108005285263062,
+      "learning_rate": 9.999898116373605e-05,
+      "loss": 0.8060292601585388,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49277238403451995,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 9.99988183917196e-05,
+      "loss": 0.7532875537872314,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4936353829557713,
+      "grad_norm": 1.261606216430664,
+      "learning_rate": 9.99986435627135e-05,
+      "loss": 0.7241997718811035,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49449838187702266,
+      "grad_norm": 1.2678134441375732,
+      "learning_rate": 9.99984566767599e-05,
+      "loss": 0.78656005859375,
+      "step": 1146
+    },
+    {
+      "epoch": 0.495361380798274,
+      "grad_norm": 1.4185810089111328,
+      "learning_rate": 9.999825773390387e-05,
+      "loss": 0.7423107624053955,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 1.3218237161636353,
+      "learning_rate": 9.999804673419338e-05,
+      "loss": 0.8178372383117676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4970873786407767,
+      "grad_norm": 1.9698240756988525,
+      "learning_rate": 9.99978236776793e-05,
+      "loss": 0.8016347885131836,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49795037756202803,
+      "grad_norm": 1.2228590250015259,
+      "learning_rate": 9.999758856441544e-05,
+      "loss": 0.782421886920929,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4988133764832794,
+      "grad_norm": 1.2753938436508179,
+      "learning_rate": 9.999734139445849e-05,
+      "loss": 0.7235615849494934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.49967637540453075,
+      "grad_norm": 1.353155493736267,
+      "learning_rate": 9.999708216786805e-05,
+      "loss": 0.7686187028884888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 1.4242770671844482,
+      "learning_rate": 9.999681088470663e-05,
+      "loss": 0.7842718362808228,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5014023732470334,
+      "grad_norm": 1.501827359199524,
+      "learning_rate": 9.999652754503966e-05,
+      "loss": 0.8324156999588013,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5022653721682848,
+      "grad_norm": 1.5403931140899658,
+      "learning_rate": 9.999623214893545e-05,
+      "loss": 0.7797855138778687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5031283710895361,
+      "grad_norm": 1.4075721502304077,
+      "learning_rate": 9.999592469646525e-05,
+      "loss": 0.7667348384857178,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5039913700107875,
+      "grad_norm": 1.4411870241165161,
+      "learning_rate": 9.99956051877032e-05,
+      "loss": 0.8153020739555359,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.4197871685028076,
+      "learning_rate": 9.999527362272632e-05,
+      "loss": 0.8548861742019653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5057173678532901,
+      "grad_norm": 1.5030468702316284,
+      "learning_rate": 9.999493000161458e-05,
+      "loss": 0.7978097796440125,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5065803667745415,
+      "grad_norm": 1.483534574508667,
+      "learning_rate": 9.999457432445088e-05,
+      "loss": 0.7990087866783142,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5074433656957928,
+      "grad_norm": 1.5424920320510864,
+      "learning_rate": 9.999420659132093e-05,
+      "loss": 0.8072807788848877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5083063646170443,
+      "grad_norm": 1.129503846168518,
+      "learning_rate": 9.999382680231344e-05,
+      "loss": 0.7216097712516785,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 1.3208131790161133,
+      "learning_rate": 9.999343495752001e-05,
+      "loss": 0.8637414574623108,
+      "step": 1180
+    },
+    {
+      "epoch": 0.510032362459547,
+      "grad_norm": 1.5239129066467285,
+      "learning_rate": 9.999303105703508e-05,
+      "loss": 0.8561362028121948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5108953613807983,
+      "grad_norm": 1.3193504810333252,
+      "learning_rate": 9.99926151009561e-05,
+      "loss": 0.7514780759811401,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5117583603020496,
+      "grad_norm": 1.312649130821228,
+      "learning_rate": 9.999218708938336e-05,
+      "loss": 0.8634516596794128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.512621359223301,
+      "grad_norm": 1.2677429914474487,
+      "learning_rate": 9.999174702242006e-05,
+      "loss": 0.7493832111358643,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 1.3901816606521606,
+      "learning_rate": 9.999129490017234e-05,
+      "loss": 0.8981389999389648,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5143473570658037,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 9.999083072274921e-05,
+      "loss": 0.6771019697189331,
+      "step": 1192
+    },
+    {
+      "epoch": 0.515210355987055,
+      "grad_norm": 1.1299076080322266,
+      "learning_rate": 9.999035449026261e-05,
+      "loss": 0.789888858795166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5160733549083064,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 9.998986620282738e-05,
+      "loss": 0.6656325459480286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5169363538295577,
+      "grad_norm": 1.3438260555267334,
+      "learning_rate": 9.998936586056127e-05,
+      "loss": 0.795595645904541,
+      "step": 1198
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 1.45097017288208,
+      "learning_rate": 9.998885346358495e-05,
+      "loss": 0.7769836187362671,
+      "step": 1200
+    },
+    {
+      "epoch": 0.517799352750809,
+      "eval_loss": 0.8070361018180847,
+      "eval_runtime": 823.9296,
+      "eval_samples_per_second": 2.5,
+      "eval_steps_per_second": 2.5,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5186623516720604,
+      "grad_norm": 1.3791370391845703,
+      "learning_rate": 9.998832901202195e-05,
+      "loss": 0.7527318000793457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5195253505933117,
+      "grad_norm": 1.2574659585952759,
+      "learning_rate": 9.998779250599876e-05,
+      "loss": 0.8021673560142517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5203883495145631,
+      "grad_norm": 1.3305809497833252,
+      "learning_rate": 9.998724394564476e-05,
+      "loss": 0.776639461517334,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5212513484358144,
+      "grad_norm": 1.2330507040023804,
+      "learning_rate": 9.998668333109223e-05,
+      "loss": 0.7381181120872498,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 1.0482324361801147,
+      "learning_rate": 9.998611066247635e-05,
+      "loss": 0.6595721244812012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5229773462783172,
+      "grad_norm": 1.5456492900848389,
+      "learning_rate": 9.998552593993522e-05,
+      "loss": 0.8388534784317017,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5238403451995685,
+      "grad_norm": 1.3730167150497437,
+      "learning_rate": 9.998492916360986e-05,
+      "loss": 0.7593658566474915,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5247033441208199,
+      "grad_norm": 1.2668508291244507,
+      "learning_rate": 9.998432033364416e-05,
+      "loss": 0.7796505093574524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5255663430420712,
+      "grad_norm": 1.1175068616867065,
+      "learning_rate": 9.998369945018493e-05,
+      "loss": 0.7493594288825989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 1.121848464012146,
+      "learning_rate": 9.99830665133819e-05,
+      "loss": 0.7791322469711304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5272923408845739,
+      "grad_norm": 1.49788236618042,
+      "learning_rate": 9.998242152338772e-05,
+      "loss": 0.8299469947814941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5281553398058253,
+      "grad_norm": 1.227125644683838,
+      "learning_rate": 9.998176448035791e-05,
+      "loss": 0.7911439538002014,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5290183387270766,
+      "grad_norm": 1.176081895828247,
+      "learning_rate": 9.998109538445091e-05,
+      "loss": 0.6487553119659424,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5298813376483279,
+      "grad_norm": 1.169835090637207,
+      "learning_rate": 9.99804142358281e-05,
+      "loss": 0.7331781983375549,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 1.4085001945495605,
+      "learning_rate": 9.997972103465367e-05,
+      "loss": 0.8198121190071106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5316073354908306,
+      "grad_norm": 1.8634135723114014,
+      "learning_rate": 9.997901578109484e-05,
+      "loss": 0.7099791169166565,
+      "step": 1232
+    },
+    {
+      "epoch": 0.532470334412082,
+      "grad_norm": 1.4719511270523071,
+      "learning_rate": 9.997829847532166e-05,
+      "loss": 0.7416951656341553,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.2430769205093384,
+      "learning_rate": 9.99775691175071e-05,
+      "loss": 0.7583408355712891,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5341963322545846,
+      "grad_norm": 1.2502130270004272,
+      "learning_rate": 9.997682770782706e-05,
+      "loss": 0.7639391422271729,
+      "step": 1238
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 1.3577396869659424,
+      "learning_rate": 9.997607424646032e-05,
+      "loss": 0.8304449319839478,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5359223300970873,
+      "grad_norm": 1.0986218452453613,
+      "learning_rate": 9.997530873358854e-05,
+      "loss": 0.7331952452659607,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5367853290183388,
+      "grad_norm": 1.4607514142990112,
+      "learning_rate": 9.997453116939637e-05,
+      "loss": 0.8173373341560364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.53764832793959,
+      "grad_norm": 1.2897552251815796,
+      "learning_rate": 9.997374155407128e-05,
+      "loss": 0.7677372694015503,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5385113268608415,
+      "grad_norm": 1.4581117630004883,
+      "learning_rate": 9.997293988780372e-05,
+      "loss": 0.8524296283721924,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 1.2029366493225098,
+      "learning_rate": 9.997212617078698e-05,
+      "loss": 0.7920866012573242,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5402373247033441,
+      "grad_norm": 1.109483242034912,
+      "learning_rate": 9.997130040321727e-05,
+      "loss": 0.7889736294746399,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5411003236245955,
+      "grad_norm": 1.1930333375930786,
+      "learning_rate": 9.997046258529377e-05,
+      "loss": 0.6593763828277588,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5419633225458468,
+      "grad_norm": 1.3668162822723389,
+      "learning_rate": 9.996961271721845e-05,
+      "loss": 0.7768212556838989,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5428263214670982,
+      "grad_norm": 1.2901886701583862,
+      "learning_rate": 9.996875079919632e-05,
+      "loss": 0.7392467260360718,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.366720199584961,
+      "learning_rate": 9.996787683143518e-05,
+      "loss": 0.6961479187011719,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5445523193096009,
+      "grad_norm": 1.2925949096679688,
+      "learning_rate": 9.99669908141458e-05,
+      "loss": 0.801517128944397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5454153182308522,
+      "grad_norm": 1.1967244148254395,
+      "learning_rate": 9.996609274754182e-05,
+      "loss": 0.7287030816078186,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5462783171521035,
+      "grad_norm": 1.419996738433838,
+      "learning_rate": 9.996518263183985e-05,
+      "loss": 0.7791255116462708,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5471413160733549,
+      "grad_norm": 1.3544845581054688,
+      "learning_rate": 9.996426046725933e-05,
+      "loss": 0.6665297150611877,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 1.6272345781326294,
+      "learning_rate": 9.996332625402262e-05,
+      "loss": 0.7902728915214539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5488673139158576,
+      "grad_norm": 1.4264246225357056,
+      "learning_rate": 9.996237999235502e-05,
+      "loss": 0.8483795523643494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5497303128371089,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 9.996142168248473e-05,
+      "loss": 0.7432113289833069,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5505933117583603,
+      "grad_norm": 0.9958710074424744,
+      "learning_rate": 9.996045132464283e-05,
+      "loss": 0.6721545457839966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5514563106796116,
+      "grad_norm": 1.371719241142273,
+      "learning_rate": 9.995946891906331e-05,
+      "loss": 0.7770468592643738,
+      "step": 1278
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 1.1906969547271729,
+      "learning_rate": 9.995847446598307e-05,
+      "loss": 0.816495418548584,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5531823085221144,
+      "grad_norm": 1.262184977531433,
+      "learning_rate": 9.995746796564195e-05,
+      "loss": 0.7122345566749573,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5540453074433657,
+      "grad_norm": 1.454692006111145,
+      "learning_rate": 9.995644941828262e-05,
+      "loss": 0.8476479649543762,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5549083063646171,
+      "grad_norm": 1.44277024269104,
+      "learning_rate": 9.995541882415072e-05,
+      "loss": 0.7595248818397522,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5557713052858684,
+      "grad_norm": 1.151465654373169,
+      "learning_rate": 9.995437618349478e-05,
+      "loss": 0.7413201928138733,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 1.3160182237625122,
+      "learning_rate": 9.995332149656621e-05,
+      "loss": 0.7242997884750366,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5574973031283711,
+      "grad_norm": 1.2748576402664185,
+      "learning_rate": 9.995225476361936e-05,
+      "loss": 0.7779543995857239,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5583603020496224,
+      "grad_norm": 1.3639990091323853,
+      "learning_rate": 9.995117598491145e-05,
+      "loss": 0.8075780868530273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5592233009708738,
+      "grad_norm": 1.0247999429702759,
+      "learning_rate": 9.995008516070263e-05,
+      "loss": 0.6940524578094482,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5600862998921251,
+      "grad_norm": 1.3383198976516724,
+      "learning_rate": 9.994898229125597e-05,
+      "loss": 0.799100399017334,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 1.37660551071167,
+      "learning_rate": 9.994786737683738e-05,
+      "loss": 0.8038942217826843,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "eval_loss": 0.7962777018547058,
+      "eval_runtime": 827.8798,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5618122977346278,
+      "grad_norm": 1.5424575805664062,
+      "learning_rate": 9.994674041771574e-05,
+      "loss": 0.8553245067596436,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5626752966558792,
+      "grad_norm": 1.1047813892364502,
+      "learning_rate": 9.994560141416282e-05,
+      "loss": 0.6650274991989136,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5635382955771305,
+      "grad_norm": 1.4532393217086792,
+      "learning_rate": 9.994445036645328e-05,
+      "loss": 0.794471025466919,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5644012944983818,
+      "grad_norm": 1.2588927745819092,
+      "learning_rate": 9.994328727486468e-05,
+      "loss": 0.7241487503051758,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 1.4981781244277954,
+      "learning_rate": 9.994211213967748e-05,
+      "loss": 0.7781044244766235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5661272923408845,
+      "grad_norm": 1.426277995109558,
+      "learning_rate": 9.99409249611751e-05,
+      "loss": 0.7772427201271057,
+      "step": 1312
+    },
+    {
+      "epoch": 0.566990291262136,
+      "grad_norm": 1.1987197399139404,
+      "learning_rate": 9.993972573964379e-05,
+      "loss": 0.7443774342536926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5678532901833873,
+      "grad_norm": 1.1490470170974731,
+      "learning_rate": 9.993851447537275e-05,
+      "loss": 0.7203393578529358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5687162891046387,
+      "grad_norm": 1.3586851358413696,
+      "learning_rate": 9.993729116865406e-05,
+      "loss": 0.8038793206214905,
+      "step": 1318
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 1.4160653352737427,
+      "learning_rate": 9.993605581978274e-05,
+      "loss": 0.827308714389801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5704422869471413,
+      "grad_norm": 1.2725251913070679,
+      "learning_rate": 9.993480842905667e-05,
+      "loss": 0.8183196783065796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5713052858683927,
+      "grad_norm": 1.3789756298065186,
+      "learning_rate": 9.993354899677665e-05,
+      "loss": 0.8268020153045654,
+      "step": 1324
+    },
+    {
+      "epoch": 0.572168284789644,
+      "grad_norm": 1.2641910314559937,
+      "learning_rate": 9.993227752324639e-05,
+      "loss": 0.7434715628623962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5730312837108954,
+      "grad_norm": 1.2429786920547485,
+      "learning_rate": 9.99309940087725e-05,
+      "loss": 0.7636955976486206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 1.3281534910202026,
+      "learning_rate": 9.992969845366449e-05,
+      "loss": 0.794617772102356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.574757281553398,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 9.992839085823477e-05,
+      "loss": 0.7658806443214417,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5756202804746494,
+      "grad_norm": 1.4143259525299072,
+      "learning_rate": 9.99270712227987e-05,
+      "loss": 0.8483052849769592,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5764832793959007,
+      "grad_norm": 1.2387818098068237,
+      "learning_rate": 9.992573954767449e-05,
+      "loss": 0.8610434532165527,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5773462783171521,
+      "grad_norm": 1.3307154178619385,
+      "learning_rate": 9.992439583318322e-05,
+      "loss": 0.8021210432052612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 1.293459415435791,
+      "learning_rate": 9.992304007964896e-05,
+      "loss": 0.7536460757255554,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5790722761596548,
+      "grad_norm": 1.1508948802947998,
+      "learning_rate": 9.992167228739865e-05,
+      "loss": 0.7519120573997498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5799352750809061,
+      "grad_norm": 1.5906150341033936,
+      "learning_rate": 9.992029245676211e-05,
+      "loss": 0.784415066242218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5807982740021574,
+      "grad_norm": 1.1421746015548706,
+      "learning_rate": 9.99189005880721e-05,
+      "loss": 0.8217549324035645,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5816612729234089,
+      "grad_norm": 1.0767277479171753,
+      "learning_rate": 9.991749668166422e-05,
+      "loss": 0.7446820735931396,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 1.147594928741455,
+      "learning_rate": 9.991608073787706e-05,
+      "loss": 0.8426011800765991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5833872707659116,
+      "grad_norm": 1.6618242263793945,
+      "learning_rate": 9.991465275705206e-05,
+      "loss": 0.8185623288154602,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5842502696871629,
+      "grad_norm": 1.414926290512085,
+      "learning_rate": 9.991321273953355e-05,
+      "loss": 0.7521756887435913,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5851132686084143,
+      "grad_norm": 1.298245906829834,
+      "learning_rate": 9.991176068566882e-05,
+      "loss": 0.7517471313476562,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5859762675296656,
+      "grad_norm": 1.3243110179901123,
+      "learning_rate": 9.9910296595808e-05,
+      "loss": 0.749305248260498,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 1.1953799724578857,
+      "learning_rate": 9.990882047030413e-05,
+      "loss": 0.6883649230003357,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5877022653721683,
+      "grad_norm": 1.3934388160705566,
+      "learning_rate": 9.990733230951322e-05,
+      "loss": 0.7859632968902588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5885652642934196,
+      "grad_norm": 1.2212340831756592,
+      "learning_rate": 9.990583211379409e-05,
+      "loss": 0.7313099503517151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.589428263214671,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 9.990431988350853e-05,
+      "loss": 0.8007960319519043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5902912621359223,
+      "grad_norm": 1.1826236248016357,
+      "learning_rate": 9.990279561902118e-05,
+      "loss": 0.773026168346405,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 1.1181561946868896,
+      "learning_rate": 9.990125932069964e-05,
+      "loss": 0.7469213604927063,
+      "step": 1370
+    },
+    {
+      "epoch": 0.592017259978425,
+      "grad_norm": 1.3735432624816895,
+      "learning_rate": 9.989971098891439e-05,
+      "loss": 0.7788950204849243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5928802588996763,
+      "grad_norm": 1.1906198263168335,
+      "learning_rate": 9.989815062403877e-05,
+      "loss": 0.8117855191230774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5937432578209277,
+      "grad_norm": 1.326940894126892,
+      "learning_rate": 9.989657822644907e-05,
+      "loss": 0.7536540627479553,
+      "step": 1376
+    },
+    {
+      "epoch": 0.594606256742179,
+      "grad_norm": 1.4508434534072876,
+      "learning_rate": 9.989499379652447e-05,
+      "loss": 0.860320508480072,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 1.114499568939209,
+      "learning_rate": 9.989339733464704e-05,
+      "loss": 0.749508798122406,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5963322545846818,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 9.989178884120176e-05,
+      "loss": 0.7294815182685852,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5971952535059332,
+      "grad_norm": 1.291913628578186,
+      "learning_rate": 9.989016831657652e-05,
+      "loss": 0.7797688245773315,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5980582524271845,
+      "grad_norm": 1.1245452165603638,
+      "learning_rate": 9.988853576116208e-05,
+      "loss": 0.7824407815933228,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5989212513484358,
+      "grad_norm": 1.2384979724884033,
+      "learning_rate": 9.988689117535215e-05,
+      "loss": 0.714794933795929,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 1.1945903301239014,
+      "learning_rate": 9.988523455954332e-05,
+      "loss": 0.7072597742080688,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6006472491909385,
+      "grad_norm": 1.3414266109466553,
+      "learning_rate": 9.988356591413504e-05,
+      "loss": 0.7075814008712769,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6015102481121899,
+      "grad_norm": 1.4575660228729248,
+      "learning_rate": 9.988188523952973e-05,
+      "loss": 0.8237126469612122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6023732470334412,
+      "grad_norm": 1.3836476802825928,
+      "learning_rate": 9.988019253613265e-05,
+      "loss": 0.7991626858711243,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6032362459546926,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.987848780435201e-05,
+      "loss": 0.6936330199241638,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 1.171294093132019,
+      "learning_rate": 9.98767710445989e-05,
+      "loss": 0.7834921479225159,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "eval_loss": 0.7866972088813782,
+      "eval_runtime": 836.0759,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6049622437971952,
+      "grad_norm": 1.5934456586837769,
+      "learning_rate": 9.987504225728727e-05,
+      "loss": 0.765119194984436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6058252427184466,
+      "grad_norm": 1.231372356414795,
+      "learning_rate": 9.987330144283407e-05,
+      "loss": 0.752061665058136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6066882416396979,
+      "grad_norm": 1.315215826034546,
+      "learning_rate": 9.987154860165903e-05,
+      "loss": 0.7475255727767944,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6075512405609493,
+      "grad_norm": 1.2640653848648071,
+      "learning_rate": 9.986978373418488e-05,
+      "loss": 0.777019202709198,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 1.085493803024292,
+      "learning_rate": 9.98680068408372e-05,
+      "loss": 0.6682190299034119,
+      "step": 1410
+    },
+    {
+      "epoch": 0.609277238403452,
+      "grad_norm": 1.4247251749038696,
+      "learning_rate": 9.986621792204447e-05,
+      "loss": 0.7765674591064453,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6101402373247033,
+      "grad_norm": 1.2993831634521484,
+      "learning_rate": 9.986441697823808e-05,
+      "loss": 0.7429938316345215,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6110032362459547,
+      "grad_norm": 2.614980936050415,
+      "learning_rate": 9.986260400985233e-05,
+      "loss": 0.7838386297225952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6118662351672061,
+      "grad_norm": 1.4305615425109863,
+      "learning_rate": 9.98607790173244e-05,
+      "loss": 0.7679475545883179,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 1.1412850618362427,
+      "learning_rate": 9.98589420010944e-05,
+      "loss": 0.6889669895172119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6135922330097088,
+      "grad_norm": 1.372313141822815,
+      "learning_rate": 9.98570929616053e-05,
+      "loss": 0.7659094333648682,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6144552319309601,
+      "grad_norm": 1.2629581689834595,
+      "learning_rate": 9.985523189930297e-05,
+      "loss": 0.7404800653457642,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6153182308522115,
+      "grad_norm": 1.5652254819869995,
+      "learning_rate": 9.985335881463623e-05,
+      "loss": 0.8082308769226074,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6161812297734628,
+      "grad_norm": 1.2268856763839722,
+      "learning_rate": 9.985147370805675e-05,
+      "loss": 0.7773997187614441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 1.2190957069396973,
+      "learning_rate": 9.984957658001912e-05,
+      "loss": 0.7412725687026978,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6179072276159655,
+      "grad_norm": 1.481990098953247,
+      "learning_rate": 9.984766743098081e-05,
+      "loss": 0.7257611751556396,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6187702265372168,
+      "grad_norm": 1.0297027826309204,
+      "learning_rate": 9.984574626140223e-05,
+      "loss": 0.7151464819908142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6196332254584682,
+      "grad_norm": 1.2404571771621704,
+      "learning_rate": 9.984381307174663e-05,
+      "loss": 0.7855883240699768,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6204962243797195,
+      "grad_norm": 1.2952802181243896,
+      "learning_rate": 9.984186786248023e-05,
+      "loss": 0.6360295414924622,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.5405724048614502,
+      "learning_rate": 9.983991063407206e-05,
+      "loss": 0.7280036807060242,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.303699254989624,
+      "learning_rate": 9.983794138699411e-05,
+      "loss": 0.7553068995475769,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6230852211434735,
+      "grad_norm": 1.1452997922897339,
+      "learning_rate": 9.983596012172127e-05,
+      "loss": 0.7329047918319702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.623948220064725,
+      "grad_norm": 1.2648253440856934,
+      "learning_rate": 9.983396683873133e-05,
+      "loss": 0.753168523311615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6248112189859762,
+      "grad_norm": 1.2095013856887817,
+      "learning_rate": 9.983196153850493e-05,
+      "loss": 0.7378705739974976,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 1.2855647802352905,
+      "learning_rate": 9.982994422152564e-05,
+      "loss": 0.7910448908805847,
+      "step": 1450
+    },
+    {
+      "epoch": 0.626537216828479,
+      "grad_norm": 1.3895999193191528,
+      "learning_rate": 9.982791488827994e-05,
+      "loss": 0.785302996635437,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6274002157497303,
+      "grad_norm": 1.2398446798324585,
+      "learning_rate": 9.982587353925719e-05,
+      "loss": 0.7542029023170471,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6282632146709817,
+      "grad_norm": 1.2903637886047363,
+      "learning_rate": 9.982382017494965e-05,
+      "loss": 0.8081737756729126,
+      "step": 1456
+    },
+    {
+      "epoch": 0.629126213592233,
+      "grad_norm": 1.223611831665039,
+      "learning_rate": 9.98217547958525e-05,
+      "loss": 0.7886892557144165,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 1.3036459684371948,
+      "learning_rate": 9.981967740246377e-05,
+      "loss": 0.844115138053894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6308522114347357,
+      "grad_norm": 1.6421090364456177,
+      "learning_rate": 9.98175879952844e-05,
+      "loss": 0.7784068584442139,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6317152103559871,
+      "grad_norm": 1.244330644607544,
+      "learning_rate": 9.981548657481828e-05,
+      "loss": 0.746979832649231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6325782092772384,
+      "grad_norm": 1.0323492288589478,
+      "learning_rate": 9.981337314157215e-05,
+      "loss": 0.6680153608322144,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6334412081984897,
+      "grad_norm": 1.2238314151763916,
+      "learning_rate": 9.981124769605563e-05,
+      "loss": 0.7103193402290344,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 1.2647801637649536,
+      "learning_rate": 9.98091102387813e-05,
+      "loss": 0.7426751255989075,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6351672060409924,
+      "grad_norm": 1.1923259496688843,
+      "learning_rate": 9.980696077026456e-05,
+      "loss": 0.7664520144462585,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6360302049622438,
+      "grad_norm": 1.0503894090652466,
+      "learning_rate": 9.980479929102377e-05,
+      "loss": 0.7348257899284363,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6368932038834951,
+      "grad_norm": 1.2223190069198608,
+      "learning_rate": 9.980262580158014e-05,
+      "loss": 0.687703013420105,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6377562028047465,
+      "grad_norm": 1.1469025611877441,
+      "learning_rate": 9.980044030245782e-05,
+      "loss": 0.7903051972389221,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 1.2581169605255127,
+      "learning_rate": 9.979824279418381e-05,
+      "loss": 0.7920540571212769,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6394822006472491,
+      "grad_norm": 1.2616536617279053,
+      "learning_rate": 9.979603327728806e-05,
+      "loss": 0.7030006647109985,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6403451995685006,
+      "grad_norm": 1.3970606327056885,
+      "learning_rate": 9.979381175230335e-05,
+      "loss": 0.7489339709281921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6412081984897519,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 9.979157821976542e-05,
+      "loss": 0.6733010411262512,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6420711974110033,
+      "grad_norm": 1.4987449645996094,
+      "learning_rate": 9.978933268021286e-05,
+      "loss": 0.8221960663795471,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 1.2843966484069824,
+      "learning_rate": 9.978707513418719e-05,
+      "loss": 0.8829001188278198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.643797195253506,
+      "grad_norm": 1.175879955291748,
+      "learning_rate": 9.978480558223278e-05,
+      "loss": 0.7605764269828796,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6446601941747573,
+      "grad_norm": 1.1991835832595825,
+      "learning_rate": 9.978252402489692e-05,
+      "loss": 0.7618141174316406,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6455231930960086,
+      "grad_norm": 1.167635202407837,
+      "learning_rate": 9.978023046272983e-05,
+      "loss": 0.7071146368980408,
+      "step": 1496
+    },
+    {
+      "epoch": 0.64638619201726,
+      "grad_norm": 1.263371229171753,
+      "learning_rate": 9.977792489628456e-05,
+      "loss": 0.7559816837310791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 1.2108519077301025,
+      "learning_rate": 9.977560732611713e-05,
+      "loss": 0.7253944873809814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "eval_loss": 0.778505802154541,
+      "eval_runtime": 836.2317,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6481121898597627,
+      "grad_norm": 1.4794813394546509,
+      "learning_rate": 9.977327775278636e-05,
+      "loss": 0.8212620615959167,
+      "step": 1502
+    },
+    {
+      "epoch": 0.648975188781014,
+      "grad_norm": 1.2120516300201416,
+      "learning_rate": 9.977093617685404e-05,
+      "loss": 0.7220476865768433,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6498381877022654,
+      "grad_norm": 1.3129044771194458,
+      "learning_rate": 9.976858259888484e-05,
+      "loss": 0.7791266441345215,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6507011866235167,
+      "grad_norm": 1.146490454673767,
+      "learning_rate": 9.976621701944628e-05,
+      "loss": 0.7052001357078552,
+      "step": 1508
+    },
+    {
+      "epoch": 0.651564185544768,
+      "grad_norm": 1.1903722286224365,
+      "learning_rate": 9.976383943910885e-05,
+      "loss": 0.7135450839996338,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6524271844660194,
+      "grad_norm": 1.4106701612472534,
+      "learning_rate": 9.976144985844589e-05,
+      "loss": 0.8114889860153198,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6532901833872707,
+      "grad_norm": 1.245661973953247,
+      "learning_rate": 9.97590482780336e-05,
+      "loss": 0.7501562237739563,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6541531823085222,
+      "grad_norm": 1.2918543815612793,
+      "learning_rate": 9.975663469845114e-05,
+      "loss": 0.7697334289550781,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6550161812297735,
+      "grad_norm": 1.2699558734893799,
+      "learning_rate": 9.975420912028053e-05,
+      "loss": 0.7457168102264404,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6558791801510249,
+      "grad_norm": 1.3968263864517212,
+      "learning_rate": 9.975177154410668e-05,
+      "loss": 0.8368204236030579,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6567421790722762,
+      "grad_norm": 1.2161848545074463,
+      "learning_rate": 9.974932197051741e-05,
+      "loss": 0.7355788350105286,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6576051779935275,
+      "grad_norm": 1.129886507987976,
+      "learning_rate": 9.974686040010341e-05,
+      "loss": 0.8023449182510376,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6584681769147789,
+      "grad_norm": 1.1114557981491089,
+      "learning_rate": 9.974438683345829e-05,
+      "loss": 0.7965474724769592,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6593311758360302,
+      "grad_norm": 1.5741950273513794,
+      "learning_rate": 9.974190127117853e-05,
+      "loss": 0.7176838517189026,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 1.1439597606658936,
+      "learning_rate": 9.973940371386352e-05,
+      "loss": 0.7114462852478027,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6610571736785329,
+      "grad_norm": 1.273659348487854,
+      "learning_rate": 9.973689416211553e-05,
+      "loss": 0.754712700843811,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6619201725997842,
+      "grad_norm": 3.9874372482299805,
+      "learning_rate": 9.973437261653973e-05,
+      "loss": 0.6720641255378723,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6627831715210356,
+      "grad_norm": 1.2506502866744995,
+      "learning_rate": 9.973183907774417e-05,
+      "loss": 0.793043851852417,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6636461704422869,
+      "grad_norm": 1.43257737159729,
+      "learning_rate": 9.972929354633982e-05,
+      "loss": 0.7806380987167358,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6645091693635383,
+      "grad_norm": 1.4794938564300537,
+      "learning_rate": 9.97267360229405e-05,
+      "loss": 0.7176088094711304,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6653721682847896,
+      "grad_norm": 1.2398169040679932,
+      "learning_rate": 9.972416650816297e-05,
+      "loss": 0.7145742774009705,
+      "step": 1542
+    },
+    {
+      "epoch": 0.666235167206041,
+      "grad_norm": 1.1309953927993774,
+      "learning_rate": 9.972158500262683e-05,
+      "loss": 0.8025617003440857,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6670981661272923,
+      "grad_norm": 1.2479451894760132,
+      "learning_rate": 9.971899150695463e-05,
+      "loss": 0.7808589339256287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6679611650485436,
+      "grad_norm": 1.2538350820541382,
+      "learning_rate": 9.971638602177175e-05,
+      "loss": 0.7096936106681824,
+      "step": 1548
+    },
+    {
+      "epoch": 0.668824163969795,
+      "grad_norm": 1.1431525945663452,
+      "learning_rate": 9.97137685477065e-05,
+      "loss": 0.6529367566108704,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6696871628910464,
+      "grad_norm": 1.5966559648513794,
+      "learning_rate": 9.971113908539008e-05,
+      "loss": 0.7780346870422363,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6705501618122978,
+      "grad_norm": 1.186475157737732,
+      "learning_rate": 9.970849763545658e-05,
+      "loss": 0.6889166831970215,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6714131607335491,
+      "grad_norm": 1.1922476291656494,
+      "learning_rate": 9.970584419854295e-05,
+      "loss": 0.7154867649078369,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6722761596548005,
+      "grad_norm": 1.30132257938385,
+      "learning_rate": 9.970317877528906e-05,
+      "loss": 0.7419366240501404,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6731391585760518,
+      "grad_norm": 1.1539888381958008,
+      "learning_rate": 9.970050136633769e-05,
+      "loss": 0.7856602072715759,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6740021574973031,
+      "grad_norm": 1.3056161403656006,
+      "learning_rate": 9.969781197233445e-05,
+      "loss": 0.6834115386009216,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6748651564185545,
+      "grad_norm": 1.2899249792099,
+      "learning_rate": 9.96951105939279e-05,
+      "loss": 0.8490727543830872,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6757281553398058,
+      "grad_norm": 1.2182374000549316,
+      "learning_rate": 9.969239723176946e-05,
+      "loss": 0.7462526559829712,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6765911542610572,
+      "grad_norm": 1.1231273412704468,
+      "learning_rate": 9.968967188651344e-05,
+      "loss": 0.7145209908485413,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6774541531823085,
+      "grad_norm": 1.1776760816574097,
+      "learning_rate": 9.968693455881703e-05,
+      "loss": 0.7911707162857056,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6783171521035599,
+      "grad_norm": 1.3542319536209106,
+      "learning_rate": 9.968418524934035e-05,
+      "loss": 0.6946635246276855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6791801510248112,
+      "grad_norm": 1.3073060512542725,
+      "learning_rate": 9.96814239587464e-05,
+      "loss": 0.7597488164901733,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6800431499460625,
+      "grad_norm": 1.3205567598342896,
+      "learning_rate": 9.967865068770099e-05,
+      "loss": 0.6763246655464172,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6809061488673139,
+      "grad_norm": 1.1734344959259033,
+      "learning_rate": 9.967586543687295e-05,
+      "loss": 0.7560655474662781,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6817691477885652,
+      "grad_norm": 1.445770025253296,
+      "learning_rate": 9.967306820693388e-05,
+      "loss": 0.8050205707550049,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6826321467098166,
+      "grad_norm": 1.1185686588287354,
+      "learning_rate": 9.967025899855836e-05,
+      "loss": 0.6628211140632629,
+      "step": 1582
+    },
+    {
+      "epoch": 0.683495145631068,
+      "grad_norm": 1.3831934928894043,
+      "learning_rate": 9.966743781242379e-05,
+      "loss": 0.7684114575386047,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6843581445523194,
+      "grad_norm": 1.2417019605636597,
+      "learning_rate": 9.966460464921047e-05,
+      "loss": 0.7084140777587891,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6852211434735707,
+      "grad_norm": 1.2138086557388306,
+      "learning_rate": 9.966175950960163e-05,
+      "loss": 0.6817399263381958,
+      "step": 1588
+    },
+    {
+      "epoch": 0.686084142394822,
+      "grad_norm": 1.399778962135315,
+      "learning_rate": 9.965890239428339e-05,
+      "loss": 0.8422354459762573,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6869471413160734,
+      "grad_norm": 1.4348714351654053,
+      "learning_rate": 9.965603330394468e-05,
+      "loss": 0.7432681322097778,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6878101402373247,
+      "grad_norm": 1.192425012588501,
+      "learning_rate": 9.965315223927741e-05,
+      "loss": 0.7571184635162354,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6886731391585761,
+      "grad_norm": 1.2834038734436035,
+      "learning_rate": 9.96502592009763e-05,
+      "loss": 0.802443265914917,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6895361380798274,
+      "grad_norm": 1.1650444269180298,
+      "learning_rate": 9.964735418973901e-05,
+      "loss": 0.8113405704498291,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "grad_norm": 1.1714438199996948,
+      "learning_rate": 9.964443720626606e-05,
+      "loss": 0.7507118582725525,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "eval_loss": 0.7706249952316284,
+      "eval_runtime": 821.4894,
+      "eval_samples_per_second": 2.508,
+      "eval_steps_per_second": 2.508,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6912621359223301,
+      "grad_norm": 1.175459861755371,
+      "learning_rate": 9.964150825126088e-05,
+      "loss": 0.7422796487808228,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6921251348435814,
+      "grad_norm": 1.6418036222457886,
+      "learning_rate": 9.963856732542978e-05,
+      "loss": 0.7125853300094604,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6929881337648328,
+      "grad_norm": 1.3256275653839111,
+      "learning_rate": 9.963561442948193e-05,
+      "loss": 0.757859468460083,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6938511326860841,
+      "grad_norm": 1.19981050491333,
+      "learning_rate": 9.963264956412944e-05,
+      "loss": 0.6849292516708374,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6947141316073355,
+      "grad_norm": 1.0392169952392578,
+      "learning_rate": 9.962967273008723e-05,
+      "loss": 0.7636215090751648,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6955771305285868,
+      "grad_norm": 1.1106646060943604,
+      "learning_rate": 9.962668392807318e-05,
+      "loss": 0.7208617329597473,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6964401294498382,
+      "grad_norm": 1.1569621562957764,
+      "learning_rate": 9.962368315880801e-05,
+      "loss": 0.671805739402771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6973031283710895,
+      "grad_norm": 1.2157769203186035,
+      "learning_rate": 9.962067042301537e-05,
+      "loss": 0.7648762464523315,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6981661272923408,
+      "grad_norm": 1.4804017543792725,
+      "learning_rate": 9.961764572142173e-05,
+      "loss": 0.8490509390830994,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 1.0344431400299072,
+      "learning_rate": 9.961460905475651e-05,
+      "loss": 0.5950891971588135,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6998921251348436,
+      "grad_norm": 1.1109060049057007,
+      "learning_rate": 9.9611560423752e-05,
+      "loss": 0.7288329005241394,
+      "step": 1622
+    },
+    {
+      "epoch": 0.700755124056095,
+      "grad_norm": 1.096996545791626,
+      "learning_rate": 9.960849982914331e-05,
+      "loss": 0.7074542045593262,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7016181229773463,
+      "grad_norm": 1.3109341859817505,
+      "learning_rate": 9.960542727166853e-05,
+      "loss": 0.7348865270614624,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7024811218985976,
+      "grad_norm": 1.3402032852172852,
+      "learning_rate": 9.96023427520686e-05,
+      "loss": 0.7658320665359497,
+      "step": 1628
+    },
+    {
+      "epoch": 0.703344120819849,
+      "grad_norm": 1.3712561130523682,
+      "learning_rate": 9.959924627108732e-05,
+      "loss": 0.809951901435852,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7042071197411003,
+      "grad_norm": 1.2278292179107666,
+      "learning_rate": 9.959613782947139e-05,
+      "loss": 0.6834598779678345,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7050701186623517,
+      "grad_norm": 1.0510832071304321,
+      "learning_rate": 9.95930174279704e-05,
+      "loss": 0.7416030764579773,
+      "step": 1634
+    },
+    {
+      "epoch": 0.705933117583603,
+      "grad_norm": 1.130991816520691,
+      "learning_rate": 9.958988506733684e-05,
+      "loss": 0.7304092645645142,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7067961165048544,
+      "grad_norm": 1.3328672647476196,
+      "learning_rate": 9.958674074832603e-05,
+      "loss": 0.8230003118515015,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7076591154261057,
+      "grad_norm": 1.3842110633850098,
+      "learning_rate": 9.958358447169625e-05,
+      "loss": 0.8413107991218567,
+      "step": 1640
+    },
+    {
+      "epoch": 0.708522114347357,
+      "grad_norm": 1.3143812417984009,
+      "learning_rate": 9.958041623820858e-05,
+      "loss": 0.77432781457901,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7093851132686084,
+      "grad_norm": 1.4727009534835815,
+      "learning_rate": 9.957723604862704e-05,
+      "loss": 0.7563844323158264,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7102481121898597,
+      "grad_norm": 1.1892327070236206,
+      "learning_rate": 9.957404390371854e-05,
+      "loss": 0.7085853219032288,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.124180555343628,
+      "learning_rate": 9.957083980425282e-05,
+      "loss": 0.6837730407714844,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 1.2224774360656738,
+      "learning_rate": 9.956762375100255e-05,
+      "loss": 0.7255740165710449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7128371089536139,
+      "grad_norm": 1.360425591468811,
+      "learning_rate": 9.956439574474326e-05,
+      "loss": 0.7849710583686829,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7137001078748652,
+      "grad_norm": 1.2818151712417603,
+      "learning_rate": 9.956115578625338e-05,
+      "loss": 0.7669665813446045,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7145631067961165,
+      "grad_norm": 1.2415164709091187,
+      "learning_rate": 9.95579038763142e-05,
+      "loss": 0.7052890658378601,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7154261057173679,
+      "grad_norm": 1.3983269929885864,
+      "learning_rate": 9.955464001570992e-05,
+      "loss": 0.8300023674964905,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7162891046386192,
+      "grad_norm": 1.1591373682022095,
+      "learning_rate": 9.955136420522759e-05,
+      "loss": 0.7742894887924194,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7171521035598706,
+      "grad_norm": 1.2503128051757812,
+      "learning_rate": 9.954807644565715e-05,
+      "loss": 0.8458141684532166,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7180151024811219,
+      "grad_norm": 1.2954912185668945,
+      "learning_rate": 9.954477673779146e-05,
+      "loss": 0.7044742107391357,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7188781014023733,
+      "grad_norm": 1.1669788360595703,
+      "learning_rate": 9.954146508242619e-05,
+      "loss": 0.7917183041572571,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7197411003236246,
+      "grad_norm": 1.1466871500015259,
+      "learning_rate": 9.953814148035996e-05,
+      "loss": 0.7174562811851501,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7206040992448759,
+      "grad_norm": 1.367224097251892,
+      "learning_rate": 9.953480593239421e-05,
+      "loss": 0.8925880193710327,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7214670981661273,
+      "grad_norm": 1.049497365951538,
+      "learning_rate": 9.953145843933334e-05,
+      "loss": 0.70949387550354,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7223300970873786,
+      "grad_norm": 0.9995084404945374,
+      "learning_rate": 9.952809900198455e-05,
+      "loss": 0.6731935143470764,
+      "step": 1674
+    },
+    {
+      "epoch": 0.72319309600863,
+      "grad_norm": 1.100899577140808,
+      "learning_rate": 9.952472762115794e-05,
+      "loss": 0.7316275835037231,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7240560949298813,
+      "grad_norm": 1.2594135999679565,
+      "learning_rate": 9.952134429766655e-05,
+      "loss": 0.7379520535469055,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7249190938511327,
+      "grad_norm": 1.3124911785125732,
+      "learning_rate": 9.951794903232621e-05,
+      "loss": 0.7072247266769409,
+      "step": 1680
+    },
+    {
+      "epoch": 0.725782092772384,
+      "grad_norm": 1.212144374847412,
+      "learning_rate": 9.95145418259557e-05,
+      "loss": 0.745680034160614,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7266450916936353,
+      "grad_norm": 1.3913586139678955,
+      "learning_rate": 9.951112267937663e-05,
+      "loss": 0.796852707862854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7275080906148867,
+      "grad_norm": 1.3227289915084839,
+      "learning_rate": 9.950769159341353e-05,
+      "loss": 0.700892984867096,
+      "step": 1686
+    },
+    {
+      "epoch": 0.728371089536138,
+      "grad_norm": 1.4656381607055664,
+      "learning_rate": 9.950424856889378e-05,
+      "loss": 0.7973095178604126,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7292340884573895,
+      "grad_norm": 1.1698973178863525,
+      "learning_rate": 9.950079360664766e-05,
+      "loss": 0.7811378836631775,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7300970873786408,
+      "grad_norm": 1.201695203781128,
+      "learning_rate": 9.949732670750831e-05,
+      "loss": 0.753012478351593,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7309600862998922,
+      "grad_norm": 1.3465325832366943,
+      "learning_rate": 9.949384787231175e-05,
+      "loss": 0.7620642781257629,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7318230852211435,
+      "grad_norm": 1.2191137075424194,
+      "learning_rate": 9.94903571018969e-05,
+      "loss": 0.7479873299598694,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7326860841423948,
+      "grad_norm": 1.1412187814712524,
+      "learning_rate": 9.948685439710554e-05,
+      "loss": 0.7327720522880554,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "grad_norm": 1.310327410697937,
+      "learning_rate": 9.948333975878233e-05,
+      "loss": 0.761475682258606,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "eval_loss": 0.7649155259132385,
+      "eval_runtime": 828.3542,
+      "eval_samples_per_second": 2.487,
+      "eval_steps_per_second": 2.487,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7344120819848975,
+      "grad_norm": 1.3031405210494995,
+      "learning_rate": 9.947981318777482e-05,
+      "loss": 0.7564467191696167,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7352750809061489,
+      "grad_norm": 1.2672377824783325,
+      "learning_rate": 9.94762746849334e-05,
+      "loss": 0.7689576745033264,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7361380798274002,
+      "grad_norm": 1.1159595251083374,
+      "learning_rate": 9.947272425111138e-05,
+      "loss": 0.7923810482025146,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7370010787486516,
+      "grad_norm": 1.1190921068191528,
+      "learning_rate": 9.946916188716495e-05,
+      "loss": 0.6540790796279907,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 1.373832106590271,
+      "learning_rate": 9.946558759395312e-05,
+      "loss": 0.7267473936080933,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7387270765911542,
+      "grad_norm": 1.2346374988555908,
+      "learning_rate": 9.946200137233785e-05,
+      "loss": 0.7693060636520386,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7395900755124056,
+      "grad_norm": 1.1629668474197388,
+      "learning_rate": 9.945840322318391e-05,
+      "loss": 0.7751407027244568,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7404530744336569,
+      "grad_norm": 1.2268708944320679,
+      "learning_rate": 9.945479314735899e-05,
+      "loss": 0.7798545360565186,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7413160733549083,
+      "grad_norm": 0.9968738555908203,
+      "learning_rate": 9.945117114573366e-05,
+      "loss": 0.6128867268562317,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7421790722761596,
+      "grad_norm": 1.2135992050170898,
+      "learning_rate": 9.944753721918133e-05,
+      "loss": 0.7536954283714294,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7430420711974111,
+      "grad_norm": 1.1361218690872192,
+      "learning_rate": 9.94438913685783e-05,
+      "loss": 0.7034479379653931,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7439050701186624,
+      "grad_norm": 1.0728967189788818,
+      "learning_rate": 9.944023359480378e-05,
+      "loss": 0.7285382747650146,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7447680690399137,
+      "grad_norm": 1.3613945245742798,
+      "learning_rate": 9.94365638987398e-05,
+      "loss": 0.7865195274353027,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7456310679611651,
+      "grad_norm": 1.0985898971557617,
+      "learning_rate": 9.94328822812713e-05,
+      "loss": 0.7511224746704102,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7464940668824164,
+      "grad_norm": 1.1506208181381226,
+      "learning_rate": 9.942918874328606e-05,
+      "loss": 0.732453465461731,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7473570658036678,
+      "grad_norm": 1.0995100736618042,
+      "learning_rate": 9.94254832856748e-05,
+      "loss": 0.702409029006958,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7482200647249191,
+      "grad_norm": 1.0075604915618896,
+      "learning_rate": 9.942176590933106e-05,
+      "loss": 0.7515703439712524,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7490830636461704,
+      "grad_norm": 1.3360812664031982,
+      "learning_rate": 9.941803661515126e-05,
+      "loss": 0.6910237073898315,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7499460625674218,
+      "grad_norm": 1.0172319412231445,
+      "learning_rate": 9.941429540403471e-05,
+      "loss": 0.7185606956481934,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7508090614886731,
+      "grad_norm": 1.2721658945083618,
+      "learning_rate": 9.941054227688358e-05,
+      "loss": 0.7136536836624146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7516720604099245,
+      "grad_norm": 1.0775413513183594,
+      "learning_rate": 9.940677723460292e-05,
+      "loss": 0.7595619559288025,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7525350593311758,
+      "grad_norm": 1.3436479568481445,
+      "learning_rate": 9.940300027810067e-05,
+      "loss": 0.7602618336677551,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7533980582524272,
+      "grad_norm": 1.2611958980560303,
+      "learning_rate": 9.93992114082876e-05,
+      "loss": 0.7954604029655457,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7542610571736785,
+      "grad_norm": 1.2794841527938843,
+      "learning_rate": 9.93954106260774e-05,
+      "loss": 0.7661190629005432,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7551240560949298,
+      "grad_norm": 1.3042595386505127,
+      "learning_rate": 9.939159793238661e-05,
+      "loss": 0.6855312585830688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7559870550161812,
+      "grad_norm": 1.2799923419952393,
+      "learning_rate": 9.938777332813463e-05,
+      "loss": 0.7970905900001526,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7568500539374325,
+      "grad_norm": 1.169205904006958,
+      "learning_rate": 9.938393681424375e-05,
+      "loss": 0.7072843313217163,
+      "step": 1754
+    },
+    {
+      "epoch": 0.757713052858684,
+      "grad_norm": 1.3833354711532593,
+      "learning_rate": 9.938008839163913e-05,
+      "loss": 0.7134629487991333,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7585760517799353,
+      "grad_norm": 1.223452091217041,
+      "learning_rate": 9.93762280612488e-05,
+      "loss": 0.7873107194900513,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7594390507011867,
+      "grad_norm": 1.2740365266799927,
+      "learning_rate": 9.937235582400367e-05,
+      "loss": 0.7680654525756836,
+      "step": 1760
+    },
+    {
+      "epoch": 0.760302049622438,
+      "grad_norm": 1.2564697265625,
+      "learning_rate": 9.93684716808375e-05,
+      "loss": 0.734126627445221,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7611650485436893,
+      "grad_norm": 1.3858847618103027,
+      "learning_rate": 9.936457563268694e-05,
+      "loss": 0.7972320914268494,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7620280474649407,
+      "grad_norm": 1.2705003023147583,
+      "learning_rate": 9.93606676804915e-05,
+      "loss": 0.7904276847839355,
+      "step": 1766
+    },
+    {
+      "epoch": 0.762891046386192,
+      "grad_norm": 1.210766077041626,
+      "learning_rate": 9.935674782519358e-05,
+      "loss": 0.7542425990104675,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7637540453074434,
+      "grad_norm": 1.4367541074752808,
+      "learning_rate": 9.93528160677384e-05,
+      "loss": 0.7975004315376282,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7646170442286947,
+      "grad_norm": 1.0994079113006592,
+      "learning_rate": 9.934887240907414e-05,
+      "loss": 0.7537807822227478,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7654800431499461,
+      "grad_norm": 1.111660122871399,
+      "learning_rate": 9.934491685015174e-05,
+      "loss": 0.7333855628967285,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7663430420711974,
+      "grad_norm": 1.2879043817520142,
+      "learning_rate": 9.934094939192508e-05,
+      "loss": 0.7472840547561646,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7672060409924487,
+      "grad_norm": 1.3618478775024414,
+      "learning_rate": 9.933697003535094e-05,
+      "loss": 0.7644904255867004,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7680690399137001,
+      "grad_norm": 1.1235666275024414,
+      "learning_rate": 9.933297878138887e-05,
+      "loss": 0.7872052192687988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7689320388349514,
+      "grad_norm": 1.3741775751113892,
+      "learning_rate": 9.932897563100136e-05,
+      "loss": 0.7343124747276306,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7697950377562028,
+      "grad_norm": 1.2115039825439453,
+      "learning_rate": 9.932496058515375e-05,
+      "loss": 0.7486474514007568,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7706580366774541,
+      "grad_norm": 1.183457851409912,
+      "learning_rate": 9.932093364481426e-05,
+      "loss": 0.7333025932312012,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7715210355987056,
+      "grad_norm": 1.2841869592666626,
+      "learning_rate": 9.931689481095394e-05,
+      "loss": 0.7241485118865967,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7723840345199569,
+      "grad_norm": 1.2312977313995361,
+      "learning_rate": 9.931284408454678e-05,
+      "loss": 0.7510396242141724,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7732470334412082,
+      "grad_norm": 1.144895315170288,
+      "learning_rate": 9.930878146656956e-05,
+      "loss": 0.7108346819877625,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7741100323624596,
+      "grad_norm": 1.18603515625,
+      "learning_rate": 9.9304706958002e-05,
+      "loss": 0.7687607407569885,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7749730312837109,
+      "grad_norm": 1.2827935218811035,
+      "learning_rate": 9.93006205598266e-05,
+      "loss": 0.7826682329177856,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7758360302049623,
+      "grad_norm": 1.164852261543274,
+      "learning_rate": 9.929652227302881e-05,
+      "loss": 0.6970428824424744,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 1.0439956188201904,
+      "learning_rate": 9.92924120985969e-05,
+      "loss": 0.7238826751708984,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "eval_loss": 0.7598537802696228,
+      "eval_runtime": 838.9645,
+      "eval_samples_per_second": 2.455,
+      "eval_steps_per_second": 2.455,
+      "step": 1800
+    },
+    {
+      "epoch": 0.777562028047465,
+      "grad_norm": 1.2464938163757324,
+      "learning_rate": 9.928829003752202e-05,
+      "loss": 0.7723641395568848,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7784250269687163,
+      "grad_norm": 1.2445050477981567,
+      "learning_rate": 9.92841560907982e-05,
+      "loss": 0.7154937982559204,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7792880258899676,
+      "grad_norm": 1.2235661745071411,
+      "learning_rate": 9.928001025942231e-05,
+      "loss": 0.8325800895690918,
+      "step": 1806
+    },
+    {
+      "epoch": 0.780151024811219,
+      "grad_norm": 1.268066644668579,
+      "learning_rate": 9.927585254439409e-05,
+      "loss": 0.7093122601509094,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7810140237324703,
+      "grad_norm": 1.1731390953063965,
+      "learning_rate": 9.927168294671618e-05,
+      "loss": 0.7601775527000427,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7818770226537217,
+      "grad_norm": 1.0685608386993408,
+      "learning_rate": 9.926750146739403e-05,
+      "loss": 0.6889236569404602,
+      "step": 1812
+    },
+    {
+      "epoch": 0.782740021574973,
+      "grad_norm": 1.1098639965057373,
+      "learning_rate": 9.926330810743602e-05,
+      "loss": 0.7397196888923645,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7836030204962244,
+      "grad_norm": 1.3606436252593994,
+      "learning_rate": 9.925910286785332e-05,
+      "loss": 0.7180029153823853,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7844660194174757,
+      "grad_norm": 1.143418550491333,
+      "learning_rate": 9.925488574966004e-05,
+      "loss": 0.7039036750793457,
+      "step": 1818
+    },
+    {
+      "epoch": 0.785329018338727,
+      "grad_norm": 1.180968999862671,
+      "learning_rate": 9.92506567538731e-05,
+      "loss": 0.741872251033783,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7861920172599784,
+      "grad_norm": 1.1953548192977905,
+      "learning_rate": 9.924641588151231e-05,
+      "loss": 0.7580198049545288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7870550161812297,
+      "grad_norm": 1.3049067258834839,
+      "learning_rate": 9.924216313360033e-05,
+      "loss": 0.7479216456413269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7879180151024812,
+      "grad_norm": 1.3962669372558594,
+      "learning_rate": 9.92378985111627e-05,
+      "loss": 0.7382357716560364,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7887810140237325,
+      "grad_norm": 1.1519217491149902,
+      "learning_rate": 9.92336220152278e-05,
+      "loss": 0.6356467604637146,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7896440129449838,
+      "grad_norm": 1.2544089555740356,
+      "learning_rate": 9.922933364682689e-05,
+      "loss": 0.7674193978309631,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7905070118662352,
+      "grad_norm": 1.1358295679092407,
+      "learning_rate": 9.922503340699412e-05,
+      "loss": 0.7083898782730103,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7913700107874865,
+      "grad_norm": 1.1744481325149536,
+      "learning_rate": 9.922072129676644e-05,
+      "loss": 0.6273066401481628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7922330097087379,
+      "grad_norm": 1.20509672164917,
+      "learning_rate": 9.92163973171837e-05,
+      "loss": 0.808181643486023,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7930960086299892,
+      "grad_norm": 1.2035597562789917,
+      "learning_rate": 9.921206146928863e-05,
+      "loss": 0.6631553769111633,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7939590075512406,
+      "grad_norm": 1.058695912361145,
+      "learning_rate": 9.920771375412678e-05,
+      "loss": 0.72016841173172,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7948220064724919,
+      "grad_norm": 1.035791039466858,
+      "learning_rate": 9.92033541727466e-05,
+      "loss": 0.7240060567855835,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7956850053937432,
+      "grad_norm": 1.1845753192901611,
+      "learning_rate": 9.919898272619935e-05,
+      "loss": 0.743224561214447,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7965480043149946,
+      "grad_norm": 1.2942636013031006,
+      "learning_rate": 9.919459941553922e-05,
+      "loss": 0.7064419388771057,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7974110032362459,
+      "grad_norm": 1.2736715078353882,
+      "learning_rate": 9.91902042418232e-05,
+      "loss": 0.805932343006134,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7982740021574973,
+      "grad_norm": 1.2803614139556885,
+      "learning_rate": 9.918579720611119e-05,
+      "loss": 0.8050802946090698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7991370010787486,
+      "grad_norm": 1.2225072383880615,
+      "learning_rate": 9.91813783094659e-05,
+      "loss": 0.7013180255889893,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3375571966171265,
+      "learning_rate": 9.917694755295294e-05,
+      "loss": 0.7415283918380737,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8008629989212513,
+      "grad_norm": 1.2209177017211914,
+      "learning_rate": 9.917250493764079e-05,
+      "loss": 0.7264692783355713,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8017259978425026,
+      "grad_norm": 1.1158864498138428,
+      "learning_rate": 9.916805046460074e-05,
+      "loss": 0.7083719372749329,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8025889967637541,
+      "grad_norm": 1.171000361442566,
+      "learning_rate": 9.916358413490696e-05,
+      "loss": 0.820396900177002,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8034519956850054,
+      "grad_norm": 0.9942941069602966,
+      "learning_rate": 9.91591059496365e-05,
+      "loss": 0.6754014492034912,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8043149946062568,
+      "grad_norm": 1.239795207977295,
+      "learning_rate": 9.915461590986926e-05,
+      "loss": 0.813154399394989,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8051779935275081,
+      "grad_norm": 1.076019525527954,
+      "learning_rate": 9.915011401668796e-05,
+      "loss": 0.745580792427063,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8060409924487595,
+      "grad_norm": 1.2134130001068115,
+      "learning_rate": 9.914560027117826e-05,
+      "loss": 0.7564949989318848,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8069039913700108,
+      "grad_norm": 1.1774994134902954,
+      "learning_rate": 9.914107467442859e-05,
+      "loss": 0.665183424949646,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8077669902912621,
+      "grad_norm": 1.0606465339660645,
+      "learning_rate": 9.91365372275303e-05,
+      "loss": 0.7597744464874268,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8086299892125135,
+      "grad_norm": 1.397161841392517,
+      "learning_rate": 9.913198793157757e-05,
+      "loss": 0.7453591227531433,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8094929881337648,
+      "grad_norm": 1.250003695487976,
+      "learning_rate": 9.912742678766742e-05,
+      "loss": 0.715792179107666,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8103559870550162,
+      "grad_norm": 1.2343473434448242,
+      "learning_rate": 9.912285379689979e-05,
+      "loss": 0.7094268798828125,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8112189859762675,
+      "grad_norm": 1.453764796257019,
+      "learning_rate": 9.911826896037739e-05,
+      "loss": 0.6652816534042358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8120819848975189,
+      "grad_norm": 1.030903935432434,
+      "learning_rate": 9.911367227920587e-05,
+      "loss": 0.6764529347419739,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8129449838187702,
+      "grad_norm": 1.2725969552993774,
+      "learning_rate": 9.91090637544937e-05,
+      "loss": 0.7043389678001404,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8138079827400215,
+      "grad_norm": 1.1610769033432007,
+      "learning_rate": 9.910444338735216e-05,
+      "loss": 0.6960558295249939,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8146709816612729,
+      "grad_norm": 1.2049180269241333,
+      "learning_rate": 9.909981117889548e-05,
+      "loss": 0.8029319643974304,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 1.1671873331069946,
+      "learning_rate": 9.909516713024068e-05,
+      "loss": 0.683114767074585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8163969795037757,
+      "grad_norm": 1.247768759727478,
+      "learning_rate": 9.909051124250764e-05,
+      "loss": 0.7193703055381775,
+      "step": 1892
+    },
+    {
+      "epoch": 0.817259978425027,
+      "grad_norm": 1.1376357078552246,
+      "learning_rate": 9.908584351681911e-05,
+      "loss": 0.7957084774971008,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8181229773462784,
+      "grad_norm": 1.1918323040008545,
+      "learning_rate": 9.90811639543007e-05,
+      "loss": 0.7168112993240356,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8189859762675297,
+      "grad_norm": 1.1249650716781616,
+      "learning_rate": 9.907647255608087e-05,
+      "loss": 0.7519245147705078,
+      "step": 1898
+    },
+    {
+      "epoch": 0.819848975188781,
+      "grad_norm": 1.0688576698303223,
+      "learning_rate": 9.907176932329092e-05,
+      "loss": 0.7496176362037659,
+      "step": 1900
+    },
+    {
+      "epoch": 0.819848975188781,
+      "eval_loss": 0.7533769607543945,
+      "eval_runtime": 830.2894,
+      "eval_samples_per_second": 2.481,
+      "eval_steps_per_second": 2.481,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8207119741100324,
+      "grad_norm": 0.9264270663261414,
+      "learning_rate": 9.906705425706499e-05,
+      "loss": 0.639653205871582,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8215749730312837,
+      "grad_norm": 1.2231628894805908,
+      "learning_rate": 9.906232735854016e-05,
+      "loss": 0.7555551528930664,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8224379719525351,
+      "grad_norm": 1.3369156122207642,
+      "learning_rate": 9.905758862885624e-05,
+      "loss": 0.7113592028617859,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8233009708737864,
+      "grad_norm": 1.217238187789917,
+      "learning_rate": 9.905283806915597e-05,
+      "loss": 0.6823481321334839,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8241639697950378,
+      "grad_norm": 1.2462831735610962,
+      "learning_rate": 9.904807568058495e-05,
+      "loss": 0.7347100377082825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8250269687162891,
+      "grad_norm": 1.2507221698760986,
+      "learning_rate": 9.904330146429157e-05,
+      "loss": 0.7981513738632202,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8258899676375404,
+      "grad_norm": 1.2246814966201782,
+      "learning_rate": 9.903851542142715e-05,
+      "loss": 0.759779155254364,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8267529665587918,
+      "grad_norm": 1.1746087074279785,
+      "learning_rate": 9.903371755314579e-05,
+      "loss": 0.741083562374115,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8276159654800431,
+      "grad_norm": 1.1913539171218872,
+      "learning_rate": 9.902890786060448e-05,
+      "loss": 0.7491101622581482,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8284789644012945,
+      "grad_norm": 1.2342501878738403,
+      "learning_rate": 9.902408634496307e-05,
+      "loss": 0.7655978202819824,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8293419633225458,
+      "grad_norm": 1.075892686843872,
+      "learning_rate": 9.901925300738424e-05,
+      "loss": 0.7199848890304565,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8302049622437971,
+      "grad_norm": 1.1517611742019653,
+      "learning_rate": 9.901440784903353e-05,
+      "loss": 0.8325693011283875,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8310679611650486,
+      "grad_norm": 1.2084628343582153,
+      "learning_rate": 9.900955087107933e-05,
+      "loss": 0.7674157023429871,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8319309600862999,
+      "grad_norm": 1.334679365158081,
+      "learning_rate": 9.900468207469287e-05,
+      "loss": 0.7415181398391724,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8327939590075513,
+      "grad_norm": 1.0505887269973755,
+      "learning_rate": 9.899980146104824e-05,
+      "loss": 0.672336220741272,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8336569579288026,
+      "grad_norm": 1.415797233581543,
+      "learning_rate": 9.899490903132238e-05,
+      "loss": 0.8077969551086426,
+      "step": 1932
+    },
+    {
+      "epoch": 0.834519956850054,
+      "grad_norm": 1.1793439388275146,
+      "learning_rate": 9.899000478669507e-05,
+      "loss": 0.8053045272827148,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8353829557713053,
+      "grad_norm": 1.196115493774414,
+      "learning_rate": 9.898508872834896e-05,
+      "loss": 0.7585914731025696,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8362459546925566,
+      "grad_norm": 1.2181763648986816,
+      "learning_rate": 9.898016085746953e-05,
+      "loss": 0.691072940826416,
+      "step": 1938
+    },
+    {
+      "epoch": 0.837108953613808,
+      "grad_norm": 1.2929240465164185,
+      "learning_rate": 9.897522117524512e-05,
+      "loss": 0.7723121643066406,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8379719525350593,
+      "grad_norm": 1.2859764099121094,
+      "learning_rate": 9.89702696828669e-05,
+      "loss": 0.8236626386642456,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8388349514563107,
+      "grad_norm": 1.2043784856796265,
+      "learning_rate": 9.89653063815289e-05,
+      "loss": 0.7315511703491211,
+      "step": 1944
+    },
+    {
+      "epoch": 0.839697950377562,
+      "grad_norm": 1.2063398361206055,
+      "learning_rate": 9.896033127242802e-05,
+      "loss": 0.6707082986831665,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8405609492988134,
+      "grad_norm": 1.1906715631484985,
+      "learning_rate": 9.895534435676393e-05,
+      "loss": 0.6920633912086487,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 1.122567892074585,
+      "learning_rate": 9.895034563573926e-05,
+      "loss": 0.7134768962860107,
+      "step": 1950
+    },
+    {
+      "epoch": 0.842286947141316,
+      "grad_norm": 1.0681787729263306,
+      "learning_rate": 9.894533511055943e-05,
+      "loss": 0.6333967447280884,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8431499460625674,
+      "grad_norm": 1.5216220617294312,
+      "learning_rate": 9.894031278243266e-05,
+      "loss": 0.7744026780128479,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8440129449838187,
+      "grad_norm": 1.0934535264968872,
+      "learning_rate": 9.89352786525701e-05,
+      "loss": 0.7678086757659912,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8448759439050701,
+      "grad_norm": 1.223285436630249,
+      "learning_rate": 9.89302327221857e-05,
+      "loss": 0.7344630360603333,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8457389428263214,
+      "grad_norm": 1.1123322248458862,
+      "learning_rate": 9.892517499249623e-05,
+      "loss": 0.76033616065979,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8466019417475729,
+      "grad_norm": 1.0472193956375122,
+      "learning_rate": 9.89201054647214e-05,
+      "loss": 0.7176017761230469,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8474649406688242,
+      "grad_norm": 1.8538368940353394,
+      "learning_rate": 9.891502414008365e-05,
+      "loss": 0.7220107316970825,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8483279395900755,
+      "grad_norm": 1.2280668020248413,
+      "learning_rate": 9.890993101980834e-05,
+      "loss": 0.7399556636810303,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8491909385113269,
+      "grad_norm": 1.1482552289962769,
+      "learning_rate": 9.890482610512364e-05,
+      "loss": 0.7700466513633728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8500539374325782,
+      "grad_norm": 1.1877986192703247,
+      "learning_rate": 9.889970939726061e-05,
+      "loss": 0.6666419506072998,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8509169363538296,
+      "grad_norm": 1.407792329788208,
+      "learning_rate": 9.88945808974531e-05,
+      "loss": 0.7883278131484985,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8517799352750809,
+      "grad_norm": 1.1426323652267456,
+      "learning_rate": 9.888944060693781e-05,
+      "loss": 0.7441611289978027,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8526429341963323,
+      "grad_norm": 1.2660622596740723,
+      "learning_rate": 9.888428852695432e-05,
+      "loss": 0.6464171409606934,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8535059331175836,
+      "grad_norm": 1.1639022827148438,
+      "learning_rate": 9.887912465874502e-05,
+      "loss": 0.7242478728294373,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 1.2546422481536865,
+      "learning_rate": 9.887394900355516e-05,
+      "loss": 0.7431875467300415,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8552319309600863,
+      "grad_norm": 1.1057130098342896,
+      "learning_rate": 9.886876156263282e-05,
+      "loss": 0.7662370204925537,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8560949298813376,
+      "grad_norm": 1.1017125844955444,
+      "learning_rate": 9.886356233722894e-05,
+      "loss": 0.7120938897132874,
+      "step": 1984
+    },
+    {
+      "epoch": 0.856957928802589,
+      "grad_norm": 1.3548704385757446,
+      "learning_rate": 9.885835132859727e-05,
+      "loss": 0.789219856262207,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8578209277238403,
+      "grad_norm": 1.1920398473739624,
+      "learning_rate": 9.885312853799443e-05,
+      "loss": 0.7599530816078186,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8586839266450917,
+      "grad_norm": 1.170946478843689,
+      "learning_rate": 9.884789396667988e-05,
+      "loss": 0.8315809369087219,
+      "step": 1990
+    },
+    {
+      "epoch": 0.859546925566343,
+      "grad_norm": 1.2530180215835571,
+      "learning_rate": 9.88426476159159e-05,
+      "loss": 0.713994026184082,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8604099244875943,
+      "grad_norm": 1.1056885719299316,
+      "learning_rate": 9.883738948696765e-05,
+      "loss": 0.6656507849693298,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8612729234088458,
+      "grad_norm": 1.169382929801941,
+      "learning_rate": 9.883211958110308e-05,
+      "loss": 0.7164257764816284,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8621359223300971,
+      "grad_norm": 1.1970325708389282,
+      "learning_rate": 9.882683789959299e-05,
+      "loss": 0.6885558366775513,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "grad_norm": 1.1700761318206787,
+      "learning_rate": 9.882154444371107e-05,
+      "loss": 0.7265812754631042,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "eval_loss": 0.7487027645111084,
+      "eval_runtime": 825.141,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8638619201725998,
+      "grad_norm": 1.3120243549346924,
+      "learning_rate": 9.88162392147338e-05,
+      "loss": 0.7371920347213745,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8647249190938512,
+      "grad_norm": 1.098697304725647,
+      "learning_rate": 9.881092221394049e-05,
+      "loss": 0.7453275918960571,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8655879180151025,
+      "grad_norm": 1.1961677074432373,
+      "learning_rate": 9.880559344261334e-05,
+      "loss": 0.7562432289123535,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8664509169363538,
+      "grad_norm": 1.1655291318893433,
+      "learning_rate": 9.880025290203735e-05,
+      "loss": 0.7654905319213867,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8673139158576052,
+      "grad_norm": 1.0979351997375488,
+      "learning_rate": 9.879490059350036e-05,
+      "loss": 0.6807888746261597,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8681769147788565,
+      "grad_norm": 1.0378475189208984,
+      "learning_rate": 9.878953651829306e-05,
+      "loss": 0.6887364387512207,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8690399137001079,
+      "grad_norm": 2.0228428840637207,
+      "learning_rate": 9.878416067770898e-05,
+      "loss": 0.718578577041626,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8699029126213592,
+      "grad_norm": 1.1610580682754517,
+      "learning_rate": 9.877877307304446e-05,
+      "loss": 0.7424600720405579,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8707659115426106,
+      "grad_norm": 1.1415833234786987,
+      "learning_rate": 9.877337370559873e-05,
+      "loss": 0.7450630068778992,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8716289104638619,
+      "grad_norm": 1.1008672714233398,
+      "learning_rate": 9.876796257667379e-05,
+      "loss": 0.7398725748062134,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8724919093851132,
+      "grad_norm": 1.1662763357162476,
+      "learning_rate": 9.876253968757452e-05,
+      "loss": 0.7122271060943604,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8733549083063646,
+      "grad_norm": 1.1835166215896606,
+      "learning_rate": 9.875710503960862e-05,
+      "loss": 0.7026485800743103,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8742179072276159,
+      "grad_norm": 1.1602214574813843,
+      "learning_rate": 9.875165863408666e-05,
+      "loss": 0.7211011052131653,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8750809061488674,
+      "grad_norm": 1.222589373588562,
+      "learning_rate": 9.874620047232198e-05,
+      "loss": 0.7565322518348694,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8759439050701187,
+      "grad_norm": 1.203661561012268,
+      "learning_rate": 9.87407305556308e-05,
+      "loss": 0.7173075675964355,
+      "step": 2030
+    },
+    {
+      "epoch": 0.87680690399137,
+      "grad_norm": 1.3316307067871094,
+      "learning_rate": 9.87352488853322e-05,
+      "loss": 0.820809006690979,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8776699029126214,
+      "grad_norm": 1.3071057796478271,
+      "learning_rate": 9.872975546274799e-05,
+      "loss": 0.8689352869987488,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8785329018338727,
+      "grad_norm": 1.3557701110839844,
+      "learning_rate": 9.872425028920294e-05,
+      "loss": 0.7090175151824951,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8793959007551241,
+      "grad_norm": 1.2234264612197876,
+      "learning_rate": 9.871873336602459e-05,
+      "loss": 0.6883202195167542,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8802588996763754,
+      "grad_norm": 1.3064714670181274,
+      "learning_rate": 9.87132046945433e-05,
+      "loss": 0.7134175896644592,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8811218985976268,
+      "grad_norm": 1.2502511739730835,
+      "learning_rate": 9.870766427609232e-05,
+      "loss": 0.7127227783203125,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8819848975188781,
+      "grad_norm": 1.1354259252548218,
+      "learning_rate": 9.870211211200766e-05,
+      "loss": 0.719437301158905,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8828478964401294,
+      "grad_norm": 1.1706442832946777,
+      "learning_rate": 9.869654820362821e-05,
+      "loss": 0.6933091282844543,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8837108953613808,
+      "grad_norm": 1.0227864980697632,
+      "learning_rate": 9.86909725522957e-05,
+      "loss": 0.6825079321861267,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8845738942826321,
+      "grad_norm": 1.103418231010437,
+      "learning_rate": 9.868538515935465e-05,
+      "loss": 0.7306133508682251,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8854368932038835,
+      "grad_norm": 1.0938395261764526,
+      "learning_rate": 9.867978602615245e-05,
+      "loss": 0.6956495046615601,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8862998921251348,
+      "grad_norm": 1.115692377090454,
+      "learning_rate": 9.86741751540393e-05,
+      "loss": 0.7740456461906433,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8871628910463862,
+      "grad_norm": 1.0264382362365723,
+      "learning_rate": 9.866855254436824e-05,
+      "loss": 0.6996002197265625,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8880258899676375,
+      "grad_norm": 1.2523163557052612,
+      "learning_rate": 9.866291819849511e-05,
+      "loss": 0.760357677936554,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.0595446825027466,
+      "learning_rate": 9.865727211777864e-05,
+      "loss": 0.6389813423156738,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8897518878101403,
+      "grad_norm": 1.198530912399292,
+      "learning_rate": 9.865161430358036e-05,
+      "loss": 0.7859060168266296,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8906148867313916,
+      "grad_norm": 1.0271193981170654,
+      "learning_rate": 9.864594475726461e-05,
+      "loss": 0.7120063900947571,
+      "step": 2064
+    },
+    {
+      "epoch": 0.891477885652643,
+      "grad_norm": 1.1619046926498413,
+      "learning_rate": 9.864026348019857e-05,
+      "loss": 0.6751878261566162,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8923408845738943,
+      "grad_norm": 1.2812464237213135,
+      "learning_rate": 9.863457047375227e-05,
+      "loss": 0.610360860824585,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 1.112628698348999,
+      "learning_rate": 9.862886573929855e-05,
+      "loss": 0.6540104746818542,
+      "step": 2070
+    },
+    {
+      "epoch": 0.894066882416397,
+      "grad_norm": 1.6257048845291138,
+      "learning_rate": 9.862314927821306e-05,
+      "loss": 0.726277232170105,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8949298813376483,
+      "grad_norm": 1.227334976196289,
+      "learning_rate": 9.861742109187432e-05,
+      "loss": 0.7798765301704407,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8957928802588997,
+      "grad_norm": 1.0728148221969604,
+      "learning_rate": 9.861168118166366e-05,
+      "loss": 0.637127161026001,
+      "step": 2076
+    },
+    {
+      "epoch": 0.896655879180151,
+      "grad_norm": 1.220282793045044,
+      "learning_rate": 9.860592954896523e-05,
+      "loss": 0.7332831025123596,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8975188781014024,
+      "grad_norm": 1.2287596464157104,
+      "learning_rate": 9.8600166195166e-05,
+      "loss": 0.7337175607681274,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8983818770226537,
+      "grad_norm": 1.0701712369918823,
+      "learning_rate": 9.859439112165579e-05,
+      "loss": 0.7422571182250977,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8992448759439051,
+      "grad_norm": 1.1583118438720703,
+      "learning_rate": 9.858860432982721e-05,
+      "loss": 0.7010003328323364,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9001078748651564,
+      "grad_norm": 0.9644797444343567,
+      "learning_rate": 9.858280582107575e-05,
+      "loss": 0.6211312413215637,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9009708737864077,
+      "grad_norm": 1.2448554039001465,
+      "learning_rate": 9.857699559679968e-05,
+      "loss": 0.7299860715866089,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9018338727076591,
+      "grad_norm": 1.2735580205917358,
+      "learning_rate": 9.85711736584001e-05,
+      "loss": 0.7461655735969543,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9026968716289104,
+      "grad_norm": 1.36841881275177,
+      "learning_rate": 9.856534000728093e-05,
+      "loss": 0.7218145132064819,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9035598705501618,
+      "grad_norm": 1.1471478939056396,
+      "learning_rate": 9.855949464484898e-05,
+      "loss": 0.7045709490776062,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9044228694714131,
+      "grad_norm": 1.058254361152649,
+      "learning_rate": 9.85536375725138e-05,
+      "loss": 0.7189744710922241,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9052858683926646,
+      "grad_norm": 1.4070709943771362,
+      "learning_rate": 9.854776879168779e-05,
+      "loss": 0.740775465965271,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 1.2182649374008179,
+      "learning_rate": 9.854188830378622e-05,
+      "loss": 0.8191262483596802,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "eval_loss": 0.7449111342430115,
+      "eval_runtime": 832.5135,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9070118662351672,
+      "grad_norm": 1.139978289604187,
+      "learning_rate": 9.853599611022708e-05,
+      "loss": 0.7295873761177063,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9078748651564186,
+      "grad_norm": 1.1802685260772705,
+      "learning_rate": 9.853009221243127e-05,
+      "loss": 0.7141753435134888,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9087378640776699,
+      "grad_norm": 1.076759934425354,
+      "learning_rate": 9.852417661182253e-05,
+      "loss": 0.7557607293128967,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9096008629989213,
+      "grad_norm": 1.1000943183898926,
+      "learning_rate": 9.851824930982733e-05,
+      "loss": 0.7025325298309326,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9104638619201726,
+      "grad_norm": 1.0235389471054077,
+      "learning_rate": 9.851231030787504e-05,
+      "loss": 0.6967638731002808,
+      "step": 2110
+    },
+    {
+      "epoch": 0.911326860841424,
+      "grad_norm": 1.2311428785324097,
+      "learning_rate": 9.850635960739782e-05,
+      "loss": 0.7890311479568481,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9121898597626753,
+      "grad_norm": 1.173770785331726,
+      "learning_rate": 9.850039720983064e-05,
+      "loss": 0.6824729442596436,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9130528586839266,
+      "grad_norm": 1.2165251970291138,
+      "learning_rate": 9.849442311661131e-05,
+      "loss": 0.7157722115516663,
+      "step": 2116
+    },
+    {
+      "epoch": 0.913915857605178,
+      "grad_norm": 1.251520037651062,
+      "learning_rate": 9.848843732918047e-05,
+      "loss": 0.7164111733436584,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9147788565264293,
+      "grad_norm": 1.1330616474151611,
+      "learning_rate": 9.848243984898153e-05,
+      "loss": 0.6830796003341675,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9156418554476807,
+      "grad_norm": 1.1052451133728027,
+      "learning_rate": 9.847643067746081e-05,
+      "loss": 0.7149171829223633,
+      "step": 2122
+    },
+    {
+      "epoch": 0.916504854368932,
+      "grad_norm": 1.0389703512191772,
+      "learning_rate": 9.847040981606735e-05,
+      "loss": 0.7213287949562073,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9173678532901833,
+      "grad_norm": 1.151895523071289,
+      "learning_rate": 9.846437726625308e-05,
+      "loss": 0.703312873840332,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9182308522114347,
+      "grad_norm": 1.1624029874801636,
+      "learning_rate": 9.845833302947269e-05,
+      "loss": 0.8059951066970825,
+      "step": 2128
+    },
+    {
+      "epoch": 0.919093851132686,
+      "grad_norm": 1.2247511148452759,
+      "learning_rate": 9.845227710718377e-05,
+      "loss": 0.735476553440094,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9199568500539375,
+      "grad_norm": 1.1649967432022095,
+      "learning_rate": 9.844620950084661e-05,
+      "loss": 0.8130571842193604,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9208198489751888,
+      "grad_norm": 1.1915717124938965,
+      "learning_rate": 9.844013021192447e-05,
+      "loss": 0.6908681392669678,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9216828478964402,
+      "grad_norm": 1.211199164390564,
+      "learning_rate": 9.843403924188328e-05,
+      "loss": 0.652173638343811,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9225458468176915,
+      "grad_norm": 1.267242193222046,
+      "learning_rate": 9.842793659219186e-05,
+      "loss": 0.6672233939170837,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9234088457389428,
+      "grad_norm": 1.7038761377334595,
+      "learning_rate": 9.842182226432186e-05,
+      "loss": 0.7274143695831299,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9242718446601942,
+      "grad_norm": 1.151426076889038,
+      "learning_rate": 9.841569625974769e-05,
+      "loss": 0.707998514175415,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9251348435814455,
+      "grad_norm": 1.0414749383926392,
+      "learning_rate": 9.840955857994663e-05,
+      "loss": 0.6639942526817322,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9259978425026969,
+      "grad_norm": 1.1045551300048828,
+      "learning_rate": 9.840340922639875e-05,
+      "loss": 0.7096070647239685,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9268608414239482,
+      "grad_norm": 1.1193996667861938,
+      "learning_rate": 9.839724820058695e-05,
+      "loss": 0.71246737241745,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9277238403451996,
+      "grad_norm": 1.2609117031097412,
+      "learning_rate": 9.839107550399692e-05,
+      "loss": 0.7351052761077881,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9285868392664509,
+      "grad_norm": 1.1127749681472778,
+      "learning_rate": 9.838489113811717e-05,
+      "loss": 0.7125008702278137,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9294498381877022,
+      "grad_norm": 1.4771665334701538,
+      "learning_rate": 9.837869510443905e-05,
+      "loss": 0.7891085743904114,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9303128371089536,
+      "grad_norm": 1.0701470375061035,
+      "learning_rate": 9.83724874044567e-05,
+      "loss": 0.7370918393135071,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9311758360302049,
+      "grad_norm": 1.2045820951461792,
+      "learning_rate": 9.836626803966708e-05,
+      "loss": 0.7543193697929382,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.1944687366485596,
+      "learning_rate": 9.836003701156996e-05,
+      "loss": 0.6903626322746277,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9329018338727076,
+      "grad_norm": 1.1035394668579102,
+      "learning_rate": 9.83537943216679e-05,
+      "loss": 0.6923050880432129,
+      "step": 2162
+    },
+    {
+      "epoch": 0.933764832793959,
+      "grad_norm": 1.3766103982925415,
+      "learning_rate": 9.834753997146633e-05,
+      "loss": 0.8099993467330933,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9346278317152104,
+      "grad_norm": 1.0874576568603516,
+      "learning_rate": 9.834127396247346e-05,
+      "loss": 0.7896219491958618,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9354908306364617,
+      "grad_norm": 1.1633459329605103,
+      "learning_rate": 9.833499629620028e-05,
+      "loss": 0.7659367918968201,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9363538295577131,
+      "grad_norm": 1.405423641204834,
+      "learning_rate": 9.832870697416066e-05,
+      "loss": 0.7474715709686279,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9372168284789644,
+      "grad_norm": 1.1794270277023315,
+      "learning_rate": 9.83224059978712e-05,
+      "loss": 0.7018750905990601,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9380798274002158,
+      "grad_norm": 1.082466959953308,
+      "learning_rate": 9.831609336885139e-05,
+      "loss": 0.6477128267288208,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9389428263214671,
+      "grad_norm": 0.9830297827720642,
+      "learning_rate": 9.830976908862346e-05,
+      "loss": 0.6516029238700867,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9398058252427185,
+      "grad_norm": 1.112289309501648,
+      "learning_rate": 9.83034331587125e-05,
+      "loss": 0.6793080568313599,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9406688241639698,
+      "grad_norm": 1.1212154626846313,
+      "learning_rate": 9.829708558064638e-05,
+      "loss": 0.7220560312271118,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9415318230852211,
+      "grad_norm": 1.246471643447876,
+      "learning_rate": 9.829072635595582e-05,
+      "loss": 0.7621577978134155,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9423948220064725,
+      "grad_norm": 1.1732288599014282,
+      "learning_rate": 9.828435548617428e-05,
+      "loss": 0.6988050937652588,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9432578209277238,
+      "grad_norm": 1.1767359972000122,
+      "learning_rate": 9.827797297283808e-05,
+      "loss": 0.7909345626831055,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9441208198489752,
+      "grad_norm": 1.007215976715088,
+      "learning_rate": 9.827157881748633e-05,
+      "loss": 0.6651493310928345,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9449838187702265,
+      "grad_norm": 1.184118628501892,
+      "learning_rate": 9.826517302166096e-05,
+      "loss": 0.7742198705673218,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9458468176914779,
+      "grad_norm": 1.1227030754089355,
+      "learning_rate": 9.82587555869067e-05,
+      "loss": 0.7614272236824036,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9467098166127292,
+      "grad_norm": 1.1030738353729248,
+      "learning_rate": 9.825232651477109e-05,
+      "loss": 0.7302108407020569,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9475728155339805,
+      "grad_norm": 1.5209563970565796,
+      "learning_rate": 9.824588580680445e-05,
+      "loss": 0.7408936023712158,
+      "step": 2196
+    },
+    {
+      "epoch": 0.948435814455232,
+      "grad_norm": 1.1853431463241577,
+      "learning_rate": 9.823943346455995e-05,
+      "loss": 0.8050196766853333,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "grad_norm": 1.285551905632019,
+      "learning_rate": 9.823296948959354e-05,
+      "loss": 0.726731538772583,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "eval_loss": 0.7400083541870117,
+      "eval_runtime": 838.7633,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9501618122977347,
+      "grad_norm": 1.2159804105758667,
+      "learning_rate": 9.822649388346396e-05,
+      "loss": 0.7061861753463745,
+      "step": 2202
+    },
+    {
+      "epoch": 0.951024811218986,
+      "grad_norm": 1.1189815998077393,
+      "learning_rate": 9.822000664773278e-05,
+      "loss": 0.824293851852417,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9518878101402374,
+      "grad_norm": 0.9968506097793579,
+      "learning_rate": 9.821350778396439e-05,
+      "loss": 0.7230689525604248,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9527508090614887,
+      "grad_norm": 1.101393222808838,
+      "learning_rate": 9.820699729372593e-05,
+      "loss": 0.6633790135383606,
+      "step": 2208
+    },
+    {
+      "epoch": 0.95361380798274,
+      "grad_norm": 1.084307312965393,
+      "learning_rate": 9.820047517858741e-05,
+      "loss": 0.6548227071762085,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9544768069039914,
+      "grad_norm": 1.1291496753692627,
+      "learning_rate": 9.819394144012158e-05,
+      "loss": 0.6434295177459717,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9553398058252427,
+      "grad_norm": 2.303483724594116,
+      "learning_rate": 9.818739607990404e-05,
+      "loss": 0.8122013211250305,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9562028047464941,
+      "grad_norm": 1.244541883468628,
+      "learning_rate": 9.818083909951315e-05,
+      "loss": 0.6120955348014832,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9570658036677454,
+      "grad_norm": 1.1051502227783203,
+      "learning_rate": 9.817427050053012e-05,
+      "loss": 0.7569628953933716,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9579288025889967,
+      "grad_norm": 1.3161911964416504,
+      "learning_rate": 9.816769028453893e-05,
+      "loss": 0.7588980197906494,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9587918015102481,
+      "grad_norm": 1.2526928186416626,
+      "learning_rate": 9.816109845312636e-05,
+      "loss": 0.743370771408081,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9596548004314994,
+      "grad_norm": 1.196215033531189,
+      "learning_rate": 9.815449500788202e-05,
+      "loss": 0.80516517162323,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9605177993527508,
+      "grad_norm": 1.3204832077026367,
+      "learning_rate": 9.81478799503983e-05,
+      "loss": 0.7331632971763611,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9613807982740021,
+      "grad_norm": 1.2442790269851685,
+      "learning_rate": 9.814125328227035e-05,
+      "loss": 0.6810339093208313,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9622437971952535,
+      "grad_norm": 1.1739158630371094,
+      "learning_rate": 9.813461500509623e-05,
+      "loss": 0.7269608378410339,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9631067961165048,
+      "grad_norm": 1.1016241312026978,
+      "learning_rate": 9.812796512047666e-05,
+      "loss": 0.7159227132797241,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9639697950377561,
+      "grad_norm": 1.1654549837112427,
+      "learning_rate": 9.812130363001529e-05,
+      "loss": 0.7145758867263794,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9648327939590076,
+      "grad_norm": 1.0521938800811768,
+      "learning_rate": 9.811463053531846e-05,
+      "loss": 0.6873835325241089,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9656957928802589,
+      "grad_norm": 1.2285776138305664,
+      "learning_rate": 9.810794583799538e-05,
+      "loss": 0.6683027148246765,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9665587918015103,
+      "grad_norm": 0.9826368689537048,
+      "learning_rate": 9.810124953965805e-05,
+      "loss": 0.663141131401062,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9674217907227616,
+      "grad_norm": 1.0102336406707764,
+      "learning_rate": 9.809454164192123e-05,
+      "loss": 0.6688778400421143,
+      "step": 2242
+    },
+    {
+      "epoch": 0.968284789644013,
+      "grad_norm": 0.9781575202941895,
+      "learning_rate": 9.808782214640251e-05,
+      "loss": 0.671240508556366,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9691477885652643,
+      "grad_norm": 1.137385368347168,
+      "learning_rate": 9.808109105472225e-05,
+      "loss": 0.7673252820968628,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9700107874865156,
+      "grad_norm": 1.3359450101852417,
+      "learning_rate": 9.807434836850362e-05,
+      "loss": 0.8085554838180542,
+      "step": 2248
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 1.150694727897644,
+      "learning_rate": 9.806759408937262e-05,
+      "loss": 0.6750202178955078,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9717367853290183,
+      "grad_norm": 1.080316424369812,
+      "learning_rate": 9.8060828218958e-05,
+      "loss": 0.5928418040275574,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9725997842502697,
+      "grad_norm": 1.0489610433578491,
+      "learning_rate": 9.805405075889129e-05,
+      "loss": 0.6947471499443054,
+      "step": 2254
+    },
+    {
+      "epoch": 0.973462783171521,
+      "grad_norm": 1.199119210243225,
+      "learning_rate": 9.80472617108069e-05,
+      "loss": 0.7273924350738525,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9743257820927724,
+      "grad_norm": 0.9830896258354187,
+      "learning_rate": 9.804046107634193e-05,
+      "loss": 0.668613612651825,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9751887810140237,
+      "grad_norm": 1.1655877828598022,
+      "learning_rate": 9.803364885713633e-05,
+      "loss": 0.7255772352218628,
+      "step": 2260
+    },
+    {
+      "epoch": 0.976051779935275,
+      "grad_norm": 1.0289462804794312,
+      "learning_rate": 9.802682505483284e-05,
+      "loss": 0.6865858435630798,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9769147788565264,
+      "grad_norm": 1.1020126342773438,
+      "learning_rate": 9.801998967107702e-05,
+      "loss": 0.6680663228034973,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 1.0846027135849,
+      "learning_rate": 9.801314270751716e-05,
+      "loss": 0.7476468086242676,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9786407766990292,
+      "grad_norm": 1.1834278106689453,
+      "learning_rate": 9.800628416580437e-05,
+      "loss": 0.7173228859901428,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9795037756202805,
+      "grad_norm": 1.0082321166992188,
+      "learning_rate": 9.799941404759257e-05,
+      "loss": 0.6768618226051331,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9803667745415319,
+      "grad_norm": 1.1573102474212646,
+      "learning_rate": 9.799253235453846e-05,
+      "loss": 0.7877311110496521,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9812297734627832,
+      "grad_norm": 1.2030092477798462,
+      "learning_rate": 9.798563908830152e-05,
+      "loss": 0.6843966841697693,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9820927723840345,
+      "grad_norm": 1.080560564994812,
+      "learning_rate": 9.797873425054404e-05,
+      "loss": 0.6084349155426025,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9829557713052859,
+      "grad_norm": 1.1488033533096313,
+      "learning_rate": 9.797181784293107e-05,
+      "loss": 0.6784868836402893,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9838187702265372,
+      "grad_norm": 1.144773006439209,
+      "learning_rate": 9.796488986713049e-05,
+      "loss": 0.6111500263214111,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9846817691477886,
+      "grad_norm": 1.1356254816055298,
+      "learning_rate": 9.795795032481297e-05,
+      "loss": 0.7634603977203369,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9855447680690399,
+      "grad_norm": 1.139446496963501,
+      "learning_rate": 9.79509992176519e-05,
+      "loss": 0.6470903158187866,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9864077669902913,
+      "grad_norm": 1.1564133167266846,
+      "learning_rate": 9.794403654732354e-05,
+      "loss": 0.71474289894104,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9872707659115426,
+      "grad_norm": 1.0605655908584595,
+      "learning_rate": 9.793706231550689e-05,
+      "loss": 0.6413968801498413,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9881337648327939,
+      "grad_norm": 1.1660863161087036,
+      "learning_rate": 9.793007652388378e-05,
+      "loss": 0.6595705151557922,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9889967637540453,
+      "grad_norm": 1.1661914587020874,
+      "learning_rate": 9.792307917413877e-05,
+      "loss": 0.7144066095352173,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9898597626752966,
+      "grad_norm": 1.2250727415084839,
+      "learning_rate": 9.791607026795926e-05,
+      "loss": 0.7956018447875977,
+      "step": 2294
+    },
+    {
+      "epoch": 0.990722761596548,
+      "grad_norm": 0.9593650698661804,
+      "learning_rate": 9.790904980703541e-05,
+      "loss": 0.5786031484603882,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9915857605177993,
+      "grad_norm": 1.0697777271270752,
+      "learning_rate": 9.790201779306018e-05,
+      "loss": 0.6817460656166077,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "grad_norm": 1.2658408880233765,
+      "learning_rate": 9.789497422772931e-05,
+      "loss": 0.6845148801803589,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "eval_loss": 0.7360134720802307,
+      "eval_runtime": 825.8559,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2300
+    },
+    {
+      "epoch": 0.993311758360302,
+      "grad_norm": 1.0869544744491577,
+      "learning_rate": 9.788791911274131e-05,
+      "loss": 0.6689707040786743,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9941747572815534,
+      "grad_norm": 1.0049875974655151,
+      "learning_rate": 9.788085244979748e-05,
+      "loss": 0.6470839381217957,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9950377562028048,
+      "grad_norm": 1.4471046924591064,
+      "learning_rate": 9.787377424060196e-05,
+      "loss": 0.8018016815185547,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9959007551240561,
+      "grad_norm": 1.3750648498535156,
+      "learning_rate": 9.786668448686159e-05,
+      "loss": 0.8078322410583496,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9967637540453075,
+      "grad_norm": 1.098283290863037,
+      "learning_rate": 9.7859583190286e-05,
+      "loss": 0.6740908622741699,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9976267529665588,
+      "grad_norm": 1.0312142372131348,
+      "learning_rate": 9.785247035258773e-05,
+      "loss": 0.6844386458396912,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9984897518878102,
+      "grad_norm": 1.4482319355010986,
+      "learning_rate": 9.784534597548193e-05,
+      "loss": 0.7547479271888733,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9993527508090615,
+      "grad_norm": 1.0897735357284546,
+      "learning_rate": 9.783821006068663e-05,
+      "loss": 0.7713987231254578,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7887805700302124,
+      "learning_rate": 9.783106260992264e-05,
+      "loss": 0.6868399381637573,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0008629989212514,
+      "grad_norm": 1.0021663904190063,
+      "learning_rate": 9.782390362491353e-05,
+      "loss": 0.696355938911438,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0017259978425026,
+      "grad_norm": 1.3650091886520386,
+      "learning_rate": 9.781673310738564e-05,
+      "loss": 0.7315902709960938,
+      "step": 2322
+    },
+    {
+      "epoch": 1.002588996763754,
+      "grad_norm": 1.2247875928878784,
+      "learning_rate": 9.780955105906813e-05,
+      "loss": 0.7667547464370728,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0034519956850054,
+      "grad_norm": 1.1917641162872314,
+      "learning_rate": 9.780235748169291e-05,
+      "loss": 0.7380582094192505,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0043149946062568,
+      "grad_norm": 1.146159052848816,
+      "learning_rate": 9.779515237699466e-05,
+      "loss": 0.6780733466148376,
+      "step": 2328
+    },
+    {
+      "epoch": 1.005177993527508,
+      "grad_norm": 1.4877620935440063,
+      "learning_rate": 9.778793574671089e-05,
+      "loss": 0.8116409182548523,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0060409924487594,
+      "grad_norm": 1.3488554954528809,
+      "learning_rate": 9.778070759258185e-05,
+      "loss": 0.65245521068573,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0069039913700109,
+      "grad_norm": 1.096686840057373,
+      "learning_rate": 9.777346791635056e-05,
+      "loss": 0.584334135055542,
+      "step": 2334
+    },
+    {
+      "epoch": 1.007766990291262,
+      "grad_norm": 1.063981056213379,
+      "learning_rate": 9.776621671976285e-05,
+      "loss": 0.6594714522361755,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0086299892125135,
+      "grad_norm": 1.2811797857284546,
+      "learning_rate": 9.775895400456729e-05,
+      "loss": 0.6963099837303162,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0094929881337649,
+      "grad_norm": 0.9491971731185913,
+      "learning_rate": 9.775167977251529e-05,
+      "loss": 0.6246375441551208,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0103559870550163,
+      "grad_norm": 1.198683500289917,
+      "learning_rate": 9.774439402536098e-05,
+      "loss": 0.6352161169052124,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0112189859762675,
+      "grad_norm": 1.2178813219070435,
+      "learning_rate": 9.773709676486129e-05,
+      "loss": 0.6884894967079163,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0120819848975189,
+      "grad_norm": 1.1905319690704346,
+      "learning_rate": 9.77297879927759e-05,
+      "loss": 0.5945048928260803,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0129449838187703,
+      "grad_norm": 1.210302710533142,
+      "learning_rate": 9.77224677108673e-05,
+      "loss": 0.6533923149108887,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0138079827400215,
+      "grad_norm": 1.15508234500885,
+      "learning_rate": 9.771513592090076e-05,
+      "loss": 0.6619405746459961,
+      "step": 2350
+    },
+    {
+      "epoch": 1.014670981661273,
+      "grad_norm": 1.0833762884140015,
+      "learning_rate": 9.77077926246443e-05,
+      "loss": 0.6581310033798218,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0155339805825243,
+      "grad_norm": 1.0330514907836914,
+      "learning_rate": 9.770043782386869e-05,
+      "loss": 0.6670945882797241,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0163969795037757,
+      "grad_norm": 1.1472195386886597,
+      "learning_rate": 9.769307152034755e-05,
+      "loss": 0.6801814436912537,
+      "step": 2356
+    },
+    {
+      "epoch": 1.017259978425027,
+      "grad_norm": 1.2204132080078125,
+      "learning_rate": 9.76856937158572e-05,
+      "loss": 0.7206465005874634,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0181229773462783,
+      "grad_norm": 1.2335962057113647,
+      "learning_rate": 9.76783044121768e-05,
+      "loss": 0.6884574294090271,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0189859762675297,
+      "grad_norm": 1.0858685970306396,
+      "learning_rate": 9.767090361108822e-05,
+      "loss": 0.655097484588623,
+      "step": 2362
+    },
+    {
+      "epoch": 1.019848975188781,
+      "grad_norm": 1.117907166481018,
+      "learning_rate": 9.766349131437614e-05,
+      "loss": 0.6478561758995056,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0207119741100323,
+      "grad_norm": 1.1995056867599487,
+      "learning_rate": 9.765606752382797e-05,
+      "loss": 0.6310951113700867,
+      "step": 2366
+    },
+    {
+      "epoch": 1.0215749730312838,
+      "grad_norm": 1.354819655418396,
+      "learning_rate": 9.764863224123397e-05,
+      "loss": 0.6373490691184998,
+      "step": 2368
+    },
+    {
+      "epoch": 1.0224379719525352,
+      "grad_norm": 1.2194819450378418,
+      "learning_rate": 9.764118546838712e-05,
+      "loss": 0.6595761179924011,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0233009708737864,
+      "grad_norm": 1.1485158205032349,
+      "learning_rate": 9.763372720708314e-05,
+      "loss": 0.6299480199813843,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0241639697950378,
+      "grad_norm": 1.1153662204742432,
+      "learning_rate": 9.76262574591206e-05,
+      "loss": 0.6788742542266846,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0250269687162892,
+      "grad_norm": 1.0690817832946777,
+      "learning_rate": 9.761877622630075e-05,
+      "loss": 0.559056282043457,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0258899676375404,
+      "grad_norm": 1.1896858215332031,
+      "learning_rate": 9.76112835104277e-05,
+      "loss": 0.7164555191993713,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0267529665587918,
+      "grad_norm": 1.1792134046554565,
+      "learning_rate": 9.760377931330824e-05,
+      "loss": 0.6791456937789917,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0276159654800432,
+      "grad_norm": 1.361763834953308,
+      "learning_rate": 9.759626363675202e-05,
+      "loss": 0.7421526312828064,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0284789644012946,
+      "grad_norm": 1.1960662603378296,
+      "learning_rate": 9.758873648257138e-05,
+      "loss": 0.7032517194747925,
+      "step": 2384
+    },
+    {
+      "epoch": 1.0293419633225458,
+      "grad_norm": 1.323255181312561,
+      "learning_rate": 9.758119785258146e-05,
+      "loss": 0.6953344345092773,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0302049622437972,
+      "grad_norm": 1.2343937158584595,
+      "learning_rate": 9.75736477486002e-05,
+      "loss": 0.7287829518318176,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0310679611650486,
+      "grad_norm": 1.0712062120437622,
+      "learning_rate": 9.756608617244823e-05,
+      "loss": 0.6812928915023804,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0319309600862998,
+      "grad_norm": 1.264726996421814,
+      "learning_rate": 9.755851312594902e-05,
+      "loss": 0.7173963785171509,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0327939590075512,
+      "grad_norm": 1.1563389301300049,
+      "learning_rate": 9.755092861092874e-05,
+      "loss": 0.6509836316108704,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0336569579288026,
+      "grad_norm": 1.326188087463379,
+      "learning_rate": 9.75433326292164e-05,
+      "loss": 0.718437135219574,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0345199568500538,
+      "grad_norm": 1.2263392210006714,
+      "learning_rate": 9.753572518264373e-05,
+      "loss": 0.664516031742096,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "grad_norm": 1.2874478101730347,
+      "learning_rate": 9.752810627304523e-05,
+      "loss": 0.6754546165466309,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "eval_loss": 0.7346783876419067,
+      "eval_runtime": 825.8356,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0362459546925566,
+      "grad_norm": 1.304518461227417,
+      "learning_rate": 9.752047590225814e-05,
+      "loss": 0.708167314529419,
+      "step": 2402
+    },
+    {
+      "epoch": 1.037108953613808,
+      "grad_norm": 1.21835196018219,
+      "learning_rate": 9.751283407212252e-05,
+      "loss": 0.7213018536567688,
+      "step": 2404
+    },
+    {
+      "epoch": 1.0379719525350593,
+      "grad_norm": 1.2947579622268677,
+      "learning_rate": 9.750518078448116e-05,
+      "loss": 0.6739231944084167,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 1.138368010520935,
+      "learning_rate": 9.749751604117959e-05,
+      "loss": 0.7671371102333069,
+      "step": 2408
+    },
+    {
+      "epoch": 1.039697950377562,
+      "grad_norm": 1.1255316734313965,
+      "learning_rate": 9.748983984406616e-05,
+      "loss": 0.6593353748321533,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0405609492988135,
+      "grad_norm": 1.2362794876098633,
+      "learning_rate": 9.748215219499193e-05,
+      "loss": 0.713056743144989,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0414239482200647,
+      "grad_norm": 1.2021304368972778,
+      "learning_rate": 9.747445309581076e-05,
+      "loss": 0.6989394426345825,
+      "step": 2414
+    },
+    {
+      "epoch": 1.042286947141316,
+      "grad_norm": 1.127341389656067,
+      "learning_rate": 9.746674254837925e-05,
+      "loss": 0.623035192489624,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0431499460625675,
+      "grad_norm": 1.1254255771636963,
+      "learning_rate": 9.745902055455672e-05,
+      "loss": 0.6490310430526733,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0440129449838187,
+      "grad_norm": 1.2142512798309326,
+      "learning_rate": 9.745128711620535e-05,
+      "loss": 0.6210505366325378,
+      "step": 2420
+    },
+    {
+      "epoch": 1.04487594390507,
+      "grad_norm": 1.1306709051132202,
+      "learning_rate": 9.744354223519001e-05,
+      "loss": 0.6243326663970947,
+      "step": 2422
+    },
+    {
+      "epoch": 1.0457389428263215,
+      "grad_norm": 1.202621340751648,
+      "learning_rate": 9.743578591337833e-05,
+      "loss": 0.6312575340270996,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0466019417475727,
+      "grad_norm": 1.0556137561798096,
+      "learning_rate": 9.742801815264072e-05,
+      "loss": 0.6190714836120605,
+      "step": 2426
+    },
+    {
+      "epoch": 1.0474649406688241,
+      "grad_norm": 1.5071814060211182,
+      "learning_rate": 9.742023895485033e-05,
+      "loss": 0.7496671080589294,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0483279395900755,
+      "grad_norm": 1.1448185443878174,
+      "learning_rate": 9.74124483218831e-05,
+      "loss": 0.6288847327232361,
+      "step": 2430
+    },
+    {
+      "epoch": 1.049190938511327,
+      "grad_norm": 1.0072964429855347,
+      "learning_rate": 9.74046462556177e-05,
+      "loss": 0.6938424706459045,
+      "step": 2432
+    },
+    {
+      "epoch": 1.0500539374325781,
+      "grad_norm": 1.198836326599121,
+      "learning_rate": 9.739683275793554e-05,
+      "loss": 0.7529639601707458,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0509169363538295,
+      "grad_norm": 1.2170380353927612,
+      "learning_rate": 9.738900783072083e-05,
+      "loss": 0.7193960547447205,
+      "step": 2436
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 1.0384336709976196,
+      "learning_rate": 9.738117147586051e-05,
+      "loss": 0.6488150954246521,
+      "step": 2438
+    },
+    {
+      "epoch": 1.0526429341963321,
+      "grad_norm": 1.1562492847442627,
+      "learning_rate": 9.737332369524429e-05,
+      "loss": 0.6672247052192688,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0535059331175836,
+      "grad_norm": 1.1553802490234375,
+      "learning_rate": 9.736546449076462e-05,
+      "loss": 0.651012122631073,
+      "step": 2442
+    },
+    {
+      "epoch": 1.054368932038835,
+      "grad_norm": 1.136838674545288,
+      "learning_rate": 9.735759386431671e-05,
+      "loss": 0.681246817111969,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0552319309600864,
+      "grad_norm": 1.177869200706482,
+      "learning_rate": 9.734971181779854e-05,
+      "loss": 0.7331019639968872,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0560949298813376,
+      "grad_norm": 1.2983415126800537,
+      "learning_rate": 9.73418183531108e-05,
+      "loss": 0.7316935658454895,
+      "step": 2448
+    },
+    {
+      "epoch": 1.056957928802589,
+      "grad_norm": 1.1189959049224854,
+      "learning_rate": 9.733391347215699e-05,
+      "loss": 0.6513608694076538,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0578209277238404,
+      "grad_norm": 1.3256937265396118,
+      "learning_rate": 9.732599717684334e-05,
+      "loss": 0.6469123363494873,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0586839266450916,
+      "grad_norm": 0.9766954779624939,
+      "learning_rate": 9.731806946907881e-05,
+      "loss": 0.5860153436660767,
+      "step": 2454
+    },
+    {
+      "epoch": 1.059546925566343,
+      "grad_norm": 1.1360441446304321,
+      "learning_rate": 9.731013035077514e-05,
+      "loss": 0.6783872842788696,
+      "step": 2456
+    },
+    {
+      "epoch": 1.0604099244875944,
+      "grad_norm": 1.3483080863952637,
+      "learning_rate": 9.73021798238468e-05,
+      "loss": 0.7894490361213684,
+      "step": 2458
+    },
+    {
+      "epoch": 1.0612729234088458,
+      "grad_norm": 1.1412159204483032,
+      "learning_rate": 9.729421789021105e-05,
+      "loss": 0.7098361849784851,
+      "step": 2460
+    },
+    {
+      "epoch": 1.062135922330097,
+      "grad_norm": 1.2670512199401855,
+      "learning_rate": 9.728624455178785e-05,
+      "loss": 0.7138097286224365,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0629989212513484,
+      "grad_norm": 1.2577205896377563,
+      "learning_rate": 9.727825981049994e-05,
+      "loss": 0.7136340141296387,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0638619201725998,
+      "grad_norm": 1.098997712135315,
+      "learning_rate": 9.727026366827283e-05,
+      "loss": 0.6273087859153748,
+      "step": 2466
+    },
+    {
+      "epoch": 1.064724919093851,
+      "grad_norm": 1.1550003290176392,
+      "learning_rate": 9.72622561270347e-05,
+      "loss": 0.7076449394226074,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0655879180151024,
+      "grad_norm": 1.0861129760742188,
+      "learning_rate": 9.725423718871659e-05,
+      "loss": 0.6856985092163086,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0664509169363539,
+      "grad_norm": 1.2742570638656616,
+      "learning_rate": 9.724620685525219e-05,
+      "loss": 0.7311446070671082,
+      "step": 2472
+    },
+    {
+      "epoch": 1.0673139158576053,
+      "grad_norm": 1.504339337348938,
+      "learning_rate": 9.7238165128578e-05,
+      "loss": 0.7728472948074341,
+      "step": 2474
+    },
+    {
+      "epoch": 1.0681769147788565,
+      "grad_norm": 1.2232030630111694,
+      "learning_rate": 9.723011201063323e-05,
+      "loss": 0.6259000301361084,
+      "step": 2476
+    },
+    {
+      "epoch": 1.0690399137001079,
+      "grad_norm": 1.3238953351974487,
+      "learning_rate": 9.722204750335989e-05,
+      "loss": 0.7132798433303833,
+      "step": 2478
+    },
+    {
+      "epoch": 1.0699029126213593,
+      "grad_norm": 1.2286484241485596,
+      "learning_rate": 9.721397160870265e-05,
+      "loss": 0.6067986488342285,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0707659115426105,
+      "grad_norm": 1.3101580142974854,
+      "learning_rate": 9.7205884328609e-05,
+      "loss": 0.6884648203849792,
+      "step": 2482
+    },
+    {
+      "epoch": 1.0716289104638619,
+      "grad_norm": 1.437833309173584,
+      "learning_rate": 9.719778566502915e-05,
+      "loss": 0.7164831757545471,
+      "step": 2484
+    },
+    {
+      "epoch": 1.0724919093851133,
+      "grad_norm": 1.0980390310287476,
+      "learning_rate": 9.718967561991606e-05,
+      "loss": 0.6326627731323242,
+      "step": 2486
+    },
+    {
+      "epoch": 1.0733549083063647,
+      "grad_norm": 0.9919570088386536,
+      "learning_rate": 9.718155419522543e-05,
+      "loss": 0.5712647438049316,
+      "step": 2488
+    },
+    {
+      "epoch": 1.074217907227616,
+      "grad_norm": 1.0773793458938599,
+      "learning_rate": 9.717342139291569e-05,
+      "loss": 0.6923331022262573,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0750809061488673,
+      "grad_norm": 1.1553318500518799,
+      "learning_rate": 9.716527721494805e-05,
+      "loss": 0.6737485527992249,
+      "step": 2492
+    },
+    {
+      "epoch": 1.0759439050701187,
+      "grad_norm": 1.112865924835205,
+      "learning_rate": 9.715712166328644e-05,
+      "loss": 0.6422063708305359,
+      "step": 2494
+    },
+    {
+      "epoch": 1.07680690399137,
+      "grad_norm": 1.0651512145996094,
+      "learning_rate": 9.71489547398975e-05,
+      "loss": 0.6399799585342407,
+      "step": 2496
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 1.1394895315170288,
+      "learning_rate": 9.71407764467507e-05,
+      "loss": 0.6218312382698059,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "grad_norm": 1.3323278427124023,
+      "learning_rate": 9.713258678581815e-05,
+      "loss": 0.65947425365448,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "eval_loss": 0.7294915914535522,
+      "eval_runtime": 836.4643,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 2500
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.606282976219734e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-2500/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-2500/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-2500/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/README.md b/sft_devstral_24B/checkpoints/checkpoint-3000/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-3000/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-3000/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..29e10b8c2e03836066c6affd2c53a33d379fb97e
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c65bd77c337a84a22331e5bd51cae1ec67e256cc5aa979340e08d04ed1698f42
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-3000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..833c11a1f10b211276f3335903d50909bf8be97a
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdbb0fa9f482cc96353567c4495b179ffe630b391696a98582d4f48909ae95c9
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-3000/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cacf008c00bc69f964ad58256896767f236695a8
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fca7fac4d0a11bc61b7275b19317a4ac989211962bde4cc5c60457d1857a0d66
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-3000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f26315809b767dbe95d10fe37d6cf22de7c36d7d
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43b30e8668249f758637a580360e1d462fc9a0aaee4f333dc3ff507e3a0646c0
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-3000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..bbec06ca240a3e3edcca97323dc0e0153b682d21
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/trainer_state.json
@@ -0,0 +1,10783 @@
+{
+  "best_global_step": 3000,
+  "best_metric": 0.7160727977752686,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-3000",
+  "epoch": 1.2942826321467098,
+  "eval_steps": 100,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4323624595469256,
+      "grad_norm": 1.5475730895996094,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.7443642616271973,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4332254584681769,
+      "grad_norm": 1.4621447324752808,
+      "learning_rate": 9.011680143755616e-05,
+      "loss": 0.8210510015487671,
+      "step": 1004
+    },
+    {
+      "epoch": 0.43408845738942825,
+      "grad_norm": 1.3651922941207886,
+      "learning_rate": 9.029649595687331e-05,
+      "loss": 0.8130766153335571,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4349514563106796,
+      "grad_norm": 1.3587357997894287,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.79145747423172,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 1.464652180671692,
+      "learning_rate": 9.065588499550764e-05,
+      "loss": 0.8286946415901184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4366774541531823,
+      "grad_norm": 1.3912625312805176,
+      "learning_rate": 9.083557951482479e-05,
+      "loss": 0.717319667339325,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4375404530744337,
+      "grad_norm": 1.2045491933822632,
+      "learning_rate": 9.101527403414196e-05,
+      "loss": 0.7294418215751648,
+      "step": 1014
+    },
+    {
+      "epoch": 0.438403451995685,
+      "grad_norm": 1.3449069261550903,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.7366700768470764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43926645091693634,
+      "grad_norm": 2.901384115219116,
+      "learning_rate": 9.137466307277629e-05,
+      "loss": 0.8551293611526489,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 1.6648688316345215,
+      "learning_rate": 9.155435759209344e-05,
+      "loss": 0.8944112658500671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44099244875943905,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 9.17340521114106e-05,
+      "loss": 0.7644214630126953,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4418554476806904,
+      "grad_norm": 1.4643231630325317,
+      "learning_rate": 9.191374663072777e-05,
+      "loss": 0.8644799590110779,
+      "step": 1024
+    },
+    {
+      "epoch": 0.44271844660194176,
+      "grad_norm": 1.4690274000167847,
+      "learning_rate": 9.209344115004492e-05,
+      "loss": 0.8281022906303406,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4435814455231931,
+      "grad_norm": 1.3321986198425293,
+      "learning_rate": 9.227313566936209e-05,
+      "loss": 0.7432141900062561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.8143638372421265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4453074433656958,
+      "grad_norm": 1.747295618057251,
+      "learning_rate": 9.263252470799642e-05,
+      "loss": 0.8102239370346069,
+      "step": 1032
+    },
+    {
+      "epoch": 0.44617044228694713,
+      "grad_norm": 1.3273073434829712,
+      "learning_rate": 9.281221922731357e-05,
+      "loss": 0.7768088579177856,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4470334412081985,
+      "grad_norm": 1.5019903182983398,
+      "learning_rate": 9.299191374663073e-05,
+      "loss": 0.7886938452720642,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44789644012944985,
+      "grad_norm": 1.6711527109146118,
+      "learning_rate": 9.31716082659479e-05,
+      "loss": 0.8152258396148682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 1.8504793643951416,
+      "learning_rate": 9.335130278526505e-05,
+      "loss": 0.8339349627494812,
+      "step": 1040
+    },
+    {
+      "epoch": 0.44962243797195256,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 9.353099730458221e-05,
+      "loss": 0.7798911929130554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.45048543689320386,
+      "grad_norm": 1.3361700773239136,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.7676728963851929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4513484358144552,
+      "grad_norm": 1.338934063911438,
+      "learning_rate": 9.389038634321654e-05,
+      "loss": 0.7729206085205078,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4522114347357066,
+      "grad_norm": 1.4347442388534546,
+      "learning_rate": 9.40700808625337e-05,
+      "loss": 0.7923640608787537,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 1.4451628923416138,
+      "learning_rate": 9.424977538185086e-05,
+      "loss": 0.8068996071815491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4539374325782093,
+      "grad_norm": 1.1904500722885132,
+      "learning_rate": 9.442946990116803e-05,
+      "loss": 0.7532910704612732,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45480043149946064,
+      "grad_norm": 1.4103976488113403,
+      "learning_rate": 9.460916442048518e-05,
+      "loss": 0.8236854672431946,
+      "step": 1054
+    },
+    {
+      "epoch": 0.455663430420712,
+      "grad_norm": 1.3762754201889038,
+      "learning_rate": 9.478885893980234e-05,
+      "loss": 0.7532716393470764,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4565264293419633,
+      "grad_norm": 1.461421012878418,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.8506031632423401,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 1.7207789421081543,
+      "learning_rate": 9.514824797843667e-05,
+      "loss": 0.7824664115905762,
+      "step": 1060
+    },
+    {
+      "epoch": 0.458252427184466,
+      "grad_norm": 1.5946974754333496,
+      "learning_rate": 9.532794249775382e-05,
+      "loss": 0.8674250245094299,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4591154261057174,
+      "grad_norm": 1.3552565574645996,
+      "learning_rate": 9.550763701707099e-05,
+      "loss": 0.7851845622062683,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45997842502696873,
+      "grad_norm": 1.414617657661438,
+      "learning_rate": 9.568733153638815e-05,
+      "loss": 0.8890700936317444,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4608414239482201,
+      "grad_norm": 1.3322749137878418,
+      "learning_rate": 9.58670260557053e-05,
+      "loss": 0.8226593732833862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 1.41420316696167,
+      "learning_rate": 9.604672057502247e-05,
+      "loss": 0.7663303017616272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.46256742179072274,
+      "grad_norm": 1.324967384338379,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.7122516632080078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4634304207119741,
+      "grad_norm": 1.2046154737472534,
+      "learning_rate": 9.640610961365679e-05,
+      "loss": 0.7177618145942688,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46429341963322546,
+      "grad_norm": 1.5221766233444214,
+      "learning_rate": 9.658580413297395e-05,
+      "loss": 0.7618930339813232,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4651564185544768,
+      "grad_norm": 1.3517179489135742,
+      "learning_rate": 9.67654986522911e-05,
+      "loss": 0.7483780384063721,
+      "step": 1078
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.747865080833435,
+      "learning_rate": 9.694519317160827e-05,
+      "loss": 0.8507670164108276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4668824163969795,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 9.712488769092543e-05,
+      "loss": 0.7893540859222412,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46774541531823083,
+      "grad_norm": 1.6290462017059326,
+      "learning_rate": 9.730458221024258e-05,
+      "loss": 0.7891272306442261,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4686084142394822,
+      "grad_norm": 1.6605802774429321,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.841018795967102,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46947141316073354,
+      "grad_norm": 1.388964295387268,
+      "learning_rate": 9.766397124887691e-05,
+      "loss": 0.8099003434181213,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 1.3901978731155396,
+      "learning_rate": 9.784366576819407e-05,
+      "loss": 0.7177658677101135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.47119741100323626,
+      "grad_norm": 1.3341423273086548,
+      "learning_rate": 9.802336028751123e-05,
+      "loss": 0.8285676836967468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4720604099244876,
+      "grad_norm": 1.5451914072036743,
+      "learning_rate": 9.82030548068284e-05,
+      "loss": 0.7582666873931885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47292340884573897,
+      "grad_norm": 1.5395640134811401,
+      "learning_rate": 9.838274932614556e-05,
+      "loss": 0.7293475270271301,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47378640776699027,
+      "grad_norm": 1.5005141496658325,
+      "learning_rate": 9.856244384546271e-05,
+      "loss": 0.8858493566513062,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 1.3562248945236206,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.7176282405853271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "eval_loss": 0.8152187466621399,
+      "eval_runtime": 835.0394,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.475512405609493,
+      "grad_norm": 1.5226716995239258,
+      "learning_rate": 9.892183288409704e-05,
+      "loss": 0.8048524856567383,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47637540453074434,
+      "grad_norm": 1.3544842004776,
+      "learning_rate": 9.91015274034142e-05,
+      "loss": 0.7954329252243042,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4772384034519957,
+      "grad_norm": 1.5169771909713745,
+      "learning_rate": 9.928122192273136e-05,
+      "loss": 0.804015040397644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47810140237324705,
+      "grad_norm": 1.4011081457138062,
+      "learning_rate": 9.946091644204852e-05,
+      "loss": 0.7652344703674316,
+      "step": 1108
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 1.4163414239883423,
+      "learning_rate": 9.964061096136569e-05,
+      "loss": 0.8501335382461548,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4798274002157497,
+      "grad_norm": 1.7406005859375,
+      "learning_rate": 9.982030548068284e-05,
+      "loss": 0.9024044871330261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.48069039913700107,
+      "grad_norm": 1.1490557193756104,
+      "learning_rate": 0.0001,
+      "loss": 0.7308287024497986,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4815533980582524,
+      "grad_norm": 1.2512775659561157,
+      "learning_rate": 9.99999939713627e-05,
+      "loss": 0.7273974418640137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4824163969795038,
+      "grad_norm": 1.3944242000579834,
+      "learning_rate": 9.999997588545227e-05,
+      "loss": 0.8071790337562561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 1.2209112644195557,
+      "learning_rate": 9.999994574227303e-05,
+      "loss": 0.6586678624153137,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4841423948220065,
+      "grad_norm": 1.7328095436096191,
+      "learning_rate": 9.999990354183228e-05,
+      "loss": 0.8529233336448669,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4850053937432578,
+      "grad_norm": 1.3309327363967896,
+      "learning_rate": 9.999984928414021e-05,
+      "loss": 0.7529556155204773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.48586839266450915,
+      "grad_norm": 1.22034752368927,
+      "learning_rate": 9.999978296920987e-05,
+      "loss": 0.6780388355255127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4867313915857605,
+      "grad_norm": 1.975162386894226,
+      "learning_rate": 9.999970459705727e-05,
+      "loss": 0.7941693067550659,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 1.4939875602722168,
+      "learning_rate": 9.999961416770132e-05,
+      "loss": 0.7895156145095825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4884573894282632,
+      "grad_norm": 1.4237034320831299,
+      "learning_rate": 9.999951168116379e-05,
+      "loss": 0.7662149667739868,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4893203883495146,
+      "grad_norm": 1.24979567527771,
+      "learning_rate": 9.999939713746942e-05,
+      "loss": 0.8079231381416321,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49018338727076594,
+      "grad_norm": 1.2252353429794312,
+      "learning_rate": 9.999927053664585e-05,
+      "loss": 0.7616434693336487,
+      "step": 1136
+    },
+    {
+      "epoch": 0.49104638619201724,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.999913187872357e-05,
+      "loss": 0.8576793074607849,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 1.2108005285263062,
+      "learning_rate": 9.999898116373605e-05,
+      "loss": 0.8060292601585388,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49277238403451995,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 9.99988183917196e-05,
+      "loss": 0.7532875537872314,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4936353829557713,
+      "grad_norm": 1.261606216430664,
+      "learning_rate": 9.99986435627135e-05,
+      "loss": 0.7241997718811035,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49449838187702266,
+      "grad_norm": 1.2678134441375732,
+      "learning_rate": 9.99984566767599e-05,
+      "loss": 0.78656005859375,
+      "step": 1146
+    },
+    {
+      "epoch": 0.495361380798274,
+      "grad_norm": 1.4185810089111328,
+      "learning_rate": 9.999825773390387e-05,
+      "loss": 0.7423107624053955,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 1.3218237161636353,
+      "learning_rate": 9.999804673419338e-05,
+      "loss": 0.8178372383117676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4970873786407767,
+      "grad_norm": 1.9698240756988525,
+      "learning_rate": 9.99978236776793e-05,
+      "loss": 0.8016347885131836,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49795037756202803,
+      "grad_norm": 1.2228590250015259,
+      "learning_rate": 9.999758856441544e-05,
+      "loss": 0.782421886920929,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4988133764832794,
+      "grad_norm": 1.2753938436508179,
+      "learning_rate": 9.999734139445849e-05,
+      "loss": 0.7235615849494934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.49967637540453075,
+      "grad_norm": 1.353155493736267,
+      "learning_rate": 9.999708216786805e-05,
+      "loss": 0.7686187028884888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 1.4242770671844482,
+      "learning_rate": 9.999681088470663e-05,
+      "loss": 0.7842718362808228,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5014023732470334,
+      "grad_norm": 1.501827359199524,
+      "learning_rate": 9.999652754503966e-05,
+      "loss": 0.8324156999588013,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5022653721682848,
+      "grad_norm": 1.5403931140899658,
+      "learning_rate": 9.999623214893545e-05,
+      "loss": 0.7797855138778687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5031283710895361,
+      "grad_norm": 1.4075721502304077,
+      "learning_rate": 9.999592469646525e-05,
+      "loss": 0.7667348384857178,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5039913700107875,
+      "grad_norm": 1.4411870241165161,
+      "learning_rate": 9.99956051877032e-05,
+      "loss": 0.8153020739555359,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.4197871685028076,
+      "learning_rate": 9.999527362272632e-05,
+      "loss": 0.8548861742019653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5057173678532901,
+      "grad_norm": 1.5030468702316284,
+      "learning_rate": 9.999493000161458e-05,
+      "loss": 0.7978097796440125,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5065803667745415,
+      "grad_norm": 1.483534574508667,
+      "learning_rate": 9.999457432445088e-05,
+      "loss": 0.7990087866783142,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5074433656957928,
+      "grad_norm": 1.5424920320510864,
+      "learning_rate": 9.999420659132093e-05,
+      "loss": 0.8072807788848877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5083063646170443,
+      "grad_norm": 1.129503846168518,
+      "learning_rate": 9.999382680231344e-05,
+      "loss": 0.7216097712516785,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 1.3208131790161133,
+      "learning_rate": 9.999343495752001e-05,
+      "loss": 0.8637414574623108,
+      "step": 1180
+    },
+    {
+      "epoch": 0.510032362459547,
+      "grad_norm": 1.5239129066467285,
+      "learning_rate": 9.999303105703508e-05,
+      "loss": 0.8561362028121948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5108953613807983,
+      "grad_norm": 1.3193504810333252,
+      "learning_rate": 9.99926151009561e-05,
+      "loss": 0.7514780759811401,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5117583603020496,
+      "grad_norm": 1.312649130821228,
+      "learning_rate": 9.999218708938336e-05,
+      "loss": 0.8634516596794128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.512621359223301,
+      "grad_norm": 1.2677429914474487,
+      "learning_rate": 9.999174702242006e-05,
+      "loss": 0.7493832111358643,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 1.3901816606521606,
+      "learning_rate": 9.999129490017234e-05,
+      "loss": 0.8981389999389648,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5143473570658037,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 9.999083072274921e-05,
+      "loss": 0.6771019697189331,
+      "step": 1192
+    },
+    {
+      "epoch": 0.515210355987055,
+      "grad_norm": 1.1299076080322266,
+      "learning_rate": 9.999035449026261e-05,
+      "loss": 0.789888858795166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5160733549083064,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 9.998986620282738e-05,
+      "loss": 0.6656325459480286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5169363538295577,
+      "grad_norm": 1.3438260555267334,
+      "learning_rate": 9.998936586056127e-05,
+      "loss": 0.795595645904541,
+      "step": 1198
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 1.45097017288208,
+      "learning_rate": 9.998885346358495e-05,
+      "loss": 0.7769836187362671,
+      "step": 1200
+    },
+    {
+      "epoch": 0.517799352750809,
+      "eval_loss": 0.8070361018180847,
+      "eval_runtime": 823.9296,
+      "eval_samples_per_second": 2.5,
+      "eval_steps_per_second": 2.5,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5186623516720604,
+      "grad_norm": 1.3791370391845703,
+      "learning_rate": 9.998832901202195e-05,
+      "loss": 0.7527318000793457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5195253505933117,
+      "grad_norm": 1.2574659585952759,
+      "learning_rate": 9.998779250599876e-05,
+      "loss": 0.8021673560142517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5203883495145631,
+      "grad_norm": 1.3305809497833252,
+      "learning_rate": 9.998724394564476e-05,
+      "loss": 0.776639461517334,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5212513484358144,
+      "grad_norm": 1.2330507040023804,
+      "learning_rate": 9.998668333109223e-05,
+      "loss": 0.7381181120872498,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 1.0482324361801147,
+      "learning_rate": 9.998611066247635e-05,
+      "loss": 0.6595721244812012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5229773462783172,
+      "grad_norm": 1.5456492900848389,
+      "learning_rate": 9.998552593993522e-05,
+      "loss": 0.8388534784317017,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5238403451995685,
+      "grad_norm": 1.3730167150497437,
+      "learning_rate": 9.998492916360986e-05,
+      "loss": 0.7593658566474915,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5247033441208199,
+      "grad_norm": 1.2668508291244507,
+      "learning_rate": 9.998432033364416e-05,
+      "loss": 0.7796505093574524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5255663430420712,
+      "grad_norm": 1.1175068616867065,
+      "learning_rate": 9.998369945018493e-05,
+      "loss": 0.7493594288825989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 1.121848464012146,
+      "learning_rate": 9.99830665133819e-05,
+      "loss": 0.7791322469711304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5272923408845739,
+      "grad_norm": 1.49788236618042,
+      "learning_rate": 9.998242152338772e-05,
+      "loss": 0.8299469947814941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5281553398058253,
+      "grad_norm": 1.227125644683838,
+      "learning_rate": 9.998176448035791e-05,
+      "loss": 0.7911439538002014,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5290183387270766,
+      "grad_norm": 1.176081895828247,
+      "learning_rate": 9.998109538445091e-05,
+      "loss": 0.6487553119659424,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5298813376483279,
+      "grad_norm": 1.169835090637207,
+      "learning_rate": 9.99804142358281e-05,
+      "loss": 0.7331781983375549,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 1.4085001945495605,
+      "learning_rate": 9.997972103465367e-05,
+      "loss": 0.8198121190071106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5316073354908306,
+      "grad_norm": 1.8634135723114014,
+      "learning_rate": 9.997901578109484e-05,
+      "loss": 0.7099791169166565,
+      "step": 1232
+    },
+    {
+      "epoch": 0.532470334412082,
+      "grad_norm": 1.4719511270523071,
+      "learning_rate": 9.997829847532166e-05,
+      "loss": 0.7416951656341553,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.2430769205093384,
+      "learning_rate": 9.99775691175071e-05,
+      "loss": 0.7583408355712891,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5341963322545846,
+      "grad_norm": 1.2502130270004272,
+      "learning_rate": 9.997682770782706e-05,
+      "loss": 0.7639391422271729,
+      "step": 1238
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 1.3577396869659424,
+      "learning_rate": 9.997607424646032e-05,
+      "loss": 0.8304449319839478,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5359223300970873,
+      "grad_norm": 1.0986218452453613,
+      "learning_rate": 9.997530873358854e-05,
+      "loss": 0.7331952452659607,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5367853290183388,
+      "grad_norm": 1.4607514142990112,
+      "learning_rate": 9.997453116939637e-05,
+      "loss": 0.8173373341560364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.53764832793959,
+      "grad_norm": 1.2897552251815796,
+      "learning_rate": 9.997374155407128e-05,
+      "loss": 0.7677372694015503,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5385113268608415,
+      "grad_norm": 1.4581117630004883,
+      "learning_rate": 9.997293988780372e-05,
+      "loss": 0.8524296283721924,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 1.2029366493225098,
+      "learning_rate": 9.997212617078698e-05,
+      "loss": 0.7920866012573242,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5402373247033441,
+      "grad_norm": 1.109483242034912,
+      "learning_rate": 9.997130040321727e-05,
+      "loss": 0.7889736294746399,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5411003236245955,
+      "grad_norm": 1.1930333375930786,
+      "learning_rate": 9.997046258529377e-05,
+      "loss": 0.6593763828277588,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5419633225458468,
+      "grad_norm": 1.3668162822723389,
+      "learning_rate": 9.996961271721845e-05,
+      "loss": 0.7768212556838989,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5428263214670982,
+      "grad_norm": 1.2901886701583862,
+      "learning_rate": 9.996875079919632e-05,
+      "loss": 0.7392467260360718,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.366720199584961,
+      "learning_rate": 9.996787683143518e-05,
+      "loss": 0.6961479187011719,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5445523193096009,
+      "grad_norm": 1.2925949096679688,
+      "learning_rate": 9.99669908141458e-05,
+      "loss": 0.801517128944397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5454153182308522,
+      "grad_norm": 1.1967244148254395,
+      "learning_rate": 9.996609274754182e-05,
+      "loss": 0.7287030816078186,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5462783171521035,
+      "grad_norm": 1.419996738433838,
+      "learning_rate": 9.996518263183985e-05,
+      "loss": 0.7791255116462708,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5471413160733549,
+      "grad_norm": 1.3544845581054688,
+      "learning_rate": 9.996426046725933e-05,
+      "loss": 0.6665297150611877,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 1.6272345781326294,
+      "learning_rate": 9.996332625402262e-05,
+      "loss": 0.7902728915214539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5488673139158576,
+      "grad_norm": 1.4264246225357056,
+      "learning_rate": 9.996237999235502e-05,
+      "loss": 0.8483795523643494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5497303128371089,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 9.996142168248473e-05,
+      "loss": 0.7432113289833069,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5505933117583603,
+      "grad_norm": 0.9958710074424744,
+      "learning_rate": 9.996045132464283e-05,
+      "loss": 0.6721545457839966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5514563106796116,
+      "grad_norm": 1.371719241142273,
+      "learning_rate": 9.995946891906331e-05,
+      "loss": 0.7770468592643738,
+      "step": 1278
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 1.1906969547271729,
+      "learning_rate": 9.995847446598307e-05,
+      "loss": 0.816495418548584,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5531823085221144,
+      "grad_norm": 1.262184977531433,
+      "learning_rate": 9.995746796564195e-05,
+      "loss": 0.7122345566749573,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5540453074433657,
+      "grad_norm": 1.454692006111145,
+      "learning_rate": 9.995644941828262e-05,
+      "loss": 0.8476479649543762,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5549083063646171,
+      "grad_norm": 1.44277024269104,
+      "learning_rate": 9.995541882415072e-05,
+      "loss": 0.7595248818397522,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5557713052858684,
+      "grad_norm": 1.151465654373169,
+      "learning_rate": 9.995437618349478e-05,
+      "loss": 0.7413201928138733,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 1.3160182237625122,
+      "learning_rate": 9.995332149656621e-05,
+      "loss": 0.7242997884750366,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5574973031283711,
+      "grad_norm": 1.2748576402664185,
+      "learning_rate": 9.995225476361936e-05,
+      "loss": 0.7779543995857239,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5583603020496224,
+      "grad_norm": 1.3639990091323853,
+      "learning_rate": 9.995117598491145e-05,
+      "loss": 0.8075780868530273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5592233009708738,
+      "grad_norm": 1.0247999429702759,
+      "learning_rate": 9.995008516070263e-05,
+      "loss": 0.6940524578094482,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5600862998921251,
+      "grad_norm": 1.3383198976516724,
+      "learning_rate": 9.994898229125597e-05,
+      "loss": 0.799100399017334,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 1.37660551071167,
+      "learning_rate": 9.994786737683738e-05,
+      "loss": 0.8038942217826843,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "eval_loss": 0.7962777018547058,
+      "eval_runtime": 827.8798,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5618122977346278,
+      "grad_norm": 1.5424575805664062,
+      "learning_rate": 9.994674041771574e-05,
+      "loss": 0.8553245067596436,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5626752966558792,
+      "grad_norm": 1.1047813892364502,
+      "learning_rate": 9.994560141416282e-05,
+      "loss": 0.6650274991989136,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5635382955771305,
+      "grad_norm": 1.4532393217086792,
+      "learning_rate": 9.994445036645328e-05,
+      "loss": 0.794471025466919,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5644012944983818,
+      "grad_norm": 1.2588927745819092,
+      "learning_rate": 9.994328727486468e-05,
+      "loss": 0.7241487503051758,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 1.4981781244277954,
+      "learning_rate": 9.994211213967748e-05,
+      "loss": 0.7781044244766235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5661272923408845,
+      "grad_norm": 1.426277995109558,
+      "learning_rate": 9.99409249611751e-05,
+      "loss": 0.7772427201271057,
+      "step": 1312
+    },
+    {
+      "epoch": 0.566990291262136,
+      "grad_norm": 1.1987197399139404,
+      "learning_rate": 9.993972573964379e-05,
+      "loss": 0.7443774342536926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5678532901833873,
+      "grad_norm": 1.1490470170974731,
+      "learning_rate": 9.993851447537275e-05,
+      "loss": 0.7203393578529358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5687162891046387,
+      "grad_norm": 1.3586851358413696,
+      "learning_rate": 9.993729116865406e-05,
+      "loss": 0.8038793206214905,
+      "step": 1318
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 1.4160653352737427,
+      "learning_rate": 9.993605581978274e-05,
+      "loss": 0.827308714389801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5704422869471413,
+      "grad_norm": 1.2725251913070679,
+      "learning_rate": 9.993480842905667e-05,
+      "loss": 0.8183196783065796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5713052858683927,
+      "grad_norm": 1.3789756298065186,
+      "learning_rate": 9.993354899677665e-05,
+      "loss": 0.8268020153045654,
+      "step": 1324
+    },
+    {
+      "epoch": 0.572168284789644,
+      "grad_norm": 1.2641910314559937,
+      "learning_rate": 9.993227752324639e-05,
+      "loss": 0.7434715628623962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5730312837108954,
+      "grad_norm": 1.2429786920547485,
+      "learning_rate": 9.99309940087725e-05,
+      "loss": 0.7636955976486206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 1.3281534910202026,
+      "learning_rate": 9.992969845366449e-05,
+      "loss": 0.794617772102356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.574757281553398,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 9.992839085823477e-05,
+      "loss": 0.7658806443214417,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5756202804746494,
+      "grad_norm": 1.4143259525299072,
+      "learning_rate": 9.99270712227987e-05,
+      "loss": 0.8483052849769592,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5764832793959007,
+      "grad_norm": 1.2387818098068237,
+      "learning_rate": 9.992573954767449e-05,
+      "loss": 0.8610434532165527,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5773462783171521,
+      "grad_norm": 1.3307154178619385,
+      "learning_rate": 9.992439583318322e-05,
+      "loss": 0.8021210432052612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 1.293459415435791,
+      "learning_rate": 9.992304007964896e-05,
+      "loss": 0.7536460757255554,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5790722761596548,
+      "grad_norm": 1.1508948802947998,
+      "learning_rate": 9.992167228739865e-05,
+      "loss": 0.7519120573997498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5799352750809061,
+      "grad_norm": 1.5906150341033936,
+      "learning_rate": 9.992029245676211e-05,
+      "loss": 0.784415066242218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5807982740021574,
+      "grad_norm": 1.1421746015548706,
+      "learning_rate": 9.99189005880721e-05,
+      "loss": 0.8217549324035645,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5816612729234089,
+      "grad_norm": 1.0767277479171753,
+      "learning_rate": 9.991749668166422e-05,
+      "loss": 0.7446820735931396,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 1.147594928741455,
+      "learning_rate": 9.991608073787706e-05,
+      "loss": 0.8426011800765991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5833872707659116,
+      "grad_norm": 1.6618242263793945,
+      "learning_rate": 9.991465275705206e-05,
+      "loss": 0.8185623288154602,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5842502696871629,
+      "grad_norm": 1.414926290512085,
+      "learning_rate": 9.991321273953355e-05,
+      "loss": 0.7521756887435913,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5851132686084143,
+      "grad_norm": 1.298245906829834,
+      "learning_rate": 9.991176068566882e-05,
+      "loss": 0.7517471313476562,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5859762675296656,
+      "grad_norm": 1.3243110179901123,
+      "learning_rate": 9.9910296595808e-05,
+      "loss": 0.749305248260498,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 1.1953799724578857,
+      "learning_rate": 9.990882047030413e-05,
+      "loss": 0.6883649230003357,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5877022653721683,
+      "grad_norm": 1.3934388160705566,
+      "learning_rate": 9.990733230951322e-05,
+      "loss": 0.7859632968902588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5885652642934196,
+      "grad_norm": 1.2212340831756592,
+      "learning_rate": 9.990583211379409e-05,
+      "loss": 0.7313099503517151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.589428263214671,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 9.990431988350853e-05,
+      "loss": 0.8007960319519043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5902912621359223,
+      "grad_norm": 1.1826236248016357,
+      "learning_rate": 9.990279561902118e-05,
+      "loss": 0.773026168346405,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 1.1181561946868896,
+      "learning_rate": 9.990125932069964e-05,
+      "loss": 0.7469213604927063,
+      "step": 1370
+    },
+    {
+      "epoch": 0.592017259978425,
+      "grad_norm": 1.3735432624816895,
+      "learning_rate": 9.989971098891439e-05,
+      "loss": 0.7788950204849243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5928802588996763,
+      "grad_norm": 1.1906198263168335,
+      "learning_rate": 9.989815062403877e-05,
+      "loss": 0.8117855191230774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5937432578209277,
+      "grad_norm": 1.326940894126892,
+      "learning_rate": 9.989657822644907e-05,
+      "loss": 0.7536540627479553,
+      "step": 1376
+    },
+    {
+      "epoch": 0.594606256742179,
+      "grad_norm": 1.4508434534072876,
+      "learning_rate": 9.989499379652447e-05,
+      "loss": 0.860320508480072,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 1.114499568939209,
+      "learning_rate": 9.989339733464704e-05,
+      "loss": 0.749508798122406,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5963322545846818,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 9.989178884120176e-05,
+      "loss": 0.7294815182685852,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5971952535059332,
+      "grad_norm": 1.291913628578186,
+      "learning_rate": 9.989016831657652e-05,
+      "loss": 0.7797688245773315,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5980582524271845,
+      "grad_norm": 1.1245452165603638,
+      "learning_rate": 9.988853576116208e-05,
+      "loss": 0.7824407815933228,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5989212513484358,
+      "grad_norm": 1.2384979724884033,
+      "learning_rate": 9.988689117535215e-05,
+      "loss": 0.714794933795929,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 1.1945903301239014,
+      "learning_rate": 9.988523455954332e-05,
+      "loss": 0.7072597742080688,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6006472491909385,
+      "grad_norm": 1.3414266109466553,
+      "learning_rate": 9.988356591413504e-05,
+      "loss": 0.7075814008712769,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6015102481121899,
+      "grad_norm": 1.4575660228729248,
+      "learning_rate": 9.988188523952973e-05,
+      "loss": 0.8237126469612122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6023732470334412,
+      "grad_norm": 1.3836476802825928,
+      "learning_rate": 9.988019253613265e-05,
+      "loss": 0.7991626858711243,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6032362459546926,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.987848780435201e-05,
+      "loss": 0.6936330199241638,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 1.171294093132019,
+      "learning_rate": 9.98767710445989e-05,
+      "loss": 0.7834921479225159,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "eval_loss": 0.7866972088813782,
+      "eval_runtime": 836.0759,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6049622437971952,
+      "grad_norm": 1.5934456586837769,
+      "learning_rate": 9.987504225728727e-05,
+      "loss": 0.765119194984436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6058252427184466,
+      "grad_norm": 1.231372356414795,
+      "learning_rate": 9.987330144283407e-05,
+      "loss": 0.752061665058136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6066882416396979,
+      "grad_norm": 1.315215826034546,
+      "learning_rate": 9.987154860165903e-05,
+      "loss": 0.7475255727767944,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6075512405609493,
+      "grad_norm": 1.2640653848648071,
+      "learning_rate": 9.986978373418488e-05,
+      "loss": 0.777019202709198,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 1.085493803024292,
+      "learning_rate": 9.98680068408372e-05,
+      "loss": 0.6682190299034119,
+      "step": 1410
+    },
+    {
+      "epoch": 0.609277238403452,
+      "grad_norm": 1.4247251749038696,
+      "learning_rate": 9.986621792204447e-05,
+      "loss": 0.7765674591064453,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6101402373247033,
+      "grad_norm": 1.2993831634521484,
+      "learning_rate": 9.986441697823808e-05,
+      "loss": 0.7429938316345215,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6110032362459547,
+      "grad_norm": 2.614980936050415,
+      "learning_rate": 9.986260400985233e-05,
+      "loss": 0.7838386297225952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6118662351672061,
+      "grad_norm": 1.4305615425109863,
+      "learning_rate": 9.98607790173244e-05,
+      "loss": 0.7679475545883179,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 1.1412850618362427,
+      "learning_rate": 9.98589420010944e-05,
+      "loss": 0.6889669895172119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6135922330097088,
+      "grad_norm": 1.372313141822815,
+      "learning_rate": 9.98570929616053e-05,
+      "loss": 0.7659094333648682,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6144552319309601,
+      "grad_norm": 1.2629581689834595,
+      "learning_rate": 9.985523189930297e-05,
+      "loss": 0.7404800653457642,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6153182308522115,
+      "grad_norm": 1.5652254819869995,
+      "learning_rate": 9.985335881463623e-05,
+      "loss": 0.8082308769226074,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6161812297734628,
+      "grad_norm": 1.2268856763839722,
+      "learning_rate": 9.985147370805675e-05,
+      "loss": 0.7773997187614441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 1.2190957069396973,
+      "learning_rate": 9.984957658001912e-05,
+      "loss": 0.7412725687026978,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6179072276159655,
+      "grad_norm": 1.481990098953247,
+      "learning_rate": 9.984766743098081e-05,
+      "loss": 0.7257611751556396,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6187702265372168,
+      "grad_norm": 1.0297027826309204,
+      "learning_rate": 9.984574626140223e-05,
+      "loss": 0.7151464819908142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6196332254584682,
+      "grad_norm": 1.2404571771621704,
+      "learning_rate": 9.984381307174663e-05,
+      "loss": 0.7855883240699768,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6204962243797195,
+      "grad_norm": 1.2952802181243896,
+      "learning_rate": 9.984186786248023e-05,
+      "loss": 0.6360295414924622,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.5405724048614502,
+      "learning_rate": 9.983991063407206e-05,
+      "loss": 0.7280036807060242,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.303699254989624,
+      "learning_rate": 9.983794138699411e-05,
+      "loss": 0.7553068995475769,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6230852211434735,
+      "grad_norm": 1.1452997922897339,
+      "learning_rate": 9.983596012172127e-05,
+      "loss": 0.7329047918319702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.623948220064725,
+      "grad_norm": 1.2648253440856934,
+      "learning_rate": 9.983396683873133e-05,
+      "loss": 0.753168523311615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6248112189859762,
+      "grad_norm": 1.2095013856887817,
+      "learning_rate": 9.983196153850493e-05,
+      "loss": 0.7378705739974976,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 1.2855647802352905,
+      "learning_rate": 9.982994422152564e-05,
+      "loss": 0.7910448908805847,
+      "step": 1450
+    },
+    {
+      "epoch": 0.626537216828479,
+      "grad_norm": 1.3895999193191528,
+      "learning_rate": 9.982791488827994e-05,
+      "loss": 0.785302996635437,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6274002157497303,
+      "grad_norm": 1.2398446798324585,
+      "learning_rate": 9.982587353925719e-05,
+      "loss": 0.7542029023170471,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6282632146709817,
+      "grad_norm": 1.2903637886047363,
+      "learning_rate": 9.982382017494965e-05,
+      "loss": 0.8081737756729126,
+      "step": 1456
+    },
+    {
+      "epoch": 0.629126213592233,
+      "grad_norm": 1.223611831665039,
+      "learning_rate": 9.98217547958525e-05,
+      "loss": 0.7886892557144165,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 1.3036459684371948,
+      "learning_rate": 9.981967740246377e-05,
+      "loss": 0.844115138053894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6308522114347357,
+      "grad_norm": 1.6421090364456177,
+      "learning_rate": 9.98175879952844e-05,
+      "loss": 0.7784068584442139,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6317152103559871,
+      "grad_norm": 1.244330644607544,
+      "learning_rate": 9.981548657481828e-05,
+      "loss": 0.746979832649231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6325782092772384,
+      "grad_norm": 1.0323492288589478,
+      "learning_rate": 9.981337314157215e-05,
+      "loss": 0.6680153608322144,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6334412081984897,
+      "grad_norm": 1.2238314151763916,
+      "learning_rate": 9.981124769605563e-05,
+      "loss": 0.7103193402290344,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 1.2647801637649536,
+      "learning_rate": 9.98091102387813e-05,
+      "loss": 0.7426751255989075,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6351672060409924,
+      "grad_norm": 1.1923259496688843,
+      "learning_rate": 9.980696077026456e-05,
+      "loss": 0.7664520144462585,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6360302049622438,
+      "grad_norm": 1.0503894090652466,
+      "learning_rate": 9.980479929102377e-05,
+      "loss": 0.7348257899284363,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6368932038834951,
+      "grad_norm": 1.2223190069198608,
+      "learning_rate": 9.980262580158014e-05,
+      "loss": 0.687703013420105,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6377562028047465,
+      "grad_norm": 1.1469025611877441,
+      "learning_rate": 9.980044030245782e-05,
+      "loss": 0.7903051972389221,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 1.2581169605255127,
+      "learning_rate": 9.979824279418381e-05,
+      "loss": 0.7920540571212769,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6394822006472491,
+      "grad_norm": 1.2616536617279053,
+      "learning_rate": 9.979603327728806e-05,
+      "loss": 0.7030006647109985,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6403451995685006,
+      "grad_norm": 1.3970606327056885,
+      "learning_rate": 9.979381175230335e-05,
+      "loss": 0.7489339709281921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6412081984897519,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 9.979157821976542e-05,
+      "loss": 0.6733010411262512,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6420711974110033,
+      "grad_norm": 1.4987449645996094,
+      "learning_rate": 9.978933268021286e-05,
+      "loss": 0.8221960663795471,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 1.2843966484069824,
+      "learning_rate": 9.978707513418719e-05,
+      "loss": 0.8829001188278198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.643797195253506,
+      "grad_norm": 1.175879955291748,
+      "learning_rate": 9.978480558223278e-05,
+      "loss": 0.7605764269828796,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6446601941747573,
+      "grad_norm": 1.1991835832595825,
+      "learning_rate": 9.978252402489692e-05,
+      "loss": 0.7618141174316406,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6455231930960086,
+      "grad_norm": 1.167635202407837,
+      "learning_rate": 9.978023046272983e-05,
+      "loss": 0.7071146368980408,
+      "step": 1496
+    },
+    {
+      "epoch": 0.64638619201726,
+      "grad_norm": 1.263371229171753,
+      "learning_rate": 9.977792489628456e-05,
+      "loss": 0.7559816837310791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 1.2108519077301025,
+      "learning_rate": 9.977560732611713e-05,
+      "loss": 0.7253944873809814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "eval_loss": 0.778505802154541,
+      "eval_runtime": 836.2317,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6481121898597627,
+      "grad_norm": 1.4794813394546509,
+      "learning_rate": 9.977327775278636e-05,
+      "loss": 0.8212620615959167,
+      "step": 1502
+    },
+    {
+      "epoch": 0.648975188781014,
+      "grad_norm": 1.2120516300201416,
+      "learning_rate": 9.977093617685404e-05,
+      "loss": 0.7220476865768433,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6498381877022654,
+      "grad_norm": 1.3129044771194458,
+      "learning_rate": 9.976858259888484e-05,
+      "loss": 0.7791266441345215,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6507011866235167,
+      "grad_norm": 1.146490454673767,
+      "learning_rate": 9.976621701944628e-05,
+      "loss": 0.7052001357078552,
+      "step": 1508
+    },
+    {
+      "epoch": 0.651564185544768,
+      "grad_norm": 1.1903722286224365,
+      "learning_rate": 9.976383943910885e-05,
+      "loss": 0.7135450839996338,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6524271844660194,
+      "grad_norm": 1.4106701612472534,
+      "learning_rate": 9.976144985844589e-05,
+      "loss": 0.8114889860153198,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6532901833872707,
+      "grad_norm": 1.245661973953247,
+      "learning_rate": 9.97590482780336e-05,
+      "loss": 0.7501562237739563,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6541531823085222,
+      "grad_norm": 1.2918543815612793,
+      "learning_rate": 9.975663469845114e-05,
+      "loss": 0.7697334289550781,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6550161812297735,
+      "grad_norm": 1.2699558734893799,
+      "learning_rate": 9.975420912028053e-05,
+      "loss": 0.7457168102264404,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6558791801510249,
+      "grad_norm": 1.3968263864517212,
+      "learning_rate": 9.975177154410668e-05,
+      "loss": 0.8368204236030579,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6567421790722762,
+      "grad_norm": 1.2161848545074463,
+      "learning_rate": 9.974932197051741e-05,
+      "loss": 0.7355788350105286,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6576051779935275,
+      "grad_norm": 1.129886507987976,
+      "learning_rate": 9.974686040010341e-05,
+      "loss": 0.8023449182510376,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6584681769147789,
+      "grad_norm": 1.1114557981491089,
+      "learning_rate": 9.974438683345829e-05,
+      "loss": 0.7965474724769592,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6593311758360302,
+      "grad_norm": 1.5741950273513794,
+      "learning_rate": 9.974190127117853e-05,
+      "loss": 0.7176838517189026,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 1.1439597606658936,
+      "learning_rate": 9.973940371386352e-05,
+      "loss": 0.7114462852478027,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6610571736785329,
+      "grad_norm": 1.273659348487854,
+      "learning_rate": 9.973689416211553e-05,
+      "loss": 0.754712700843811,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6619201725997842,
+      "grad_norm": 3.9874372482299805,
+      "learning_rate": 9.973437261653973e-05,
+      "loss": 0.6720641255378723,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6627831715210356,
+      "grad_norm": 1.2506502866744995,
+      "learning_rate": 9.973183907774417e-05,
+      "loss": 0.793043851852417,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6636461704422869,
+      "grad_norm": 1.43257737159729,
+      "learning_rate": 9.972929354633982e-05,
+      "loss": 0.7806380987167358,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6645091693635383,
+      "grad_norm": 1.4794938564300537,
+      "learning_rate": 9.97267360229405e-05,
+      "loss": 0.7176088094711304,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6653721682847896,
+      "grad_norm": 1.2398169040679932,
+      "learning_rate": 9.972416650816297e-05,
+      "loss": 0.7145742774009705,
+      "step": 1542
+    },
+    {
+      "epoch": 0.666235167206041,
+      "grad_norm": 1.1309953927993774,
+      "learning_rate": 9.972158500262683e-05,
+      "loss": 0.8025617003440857,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6670981661272923,
+      "grad_norm": 1.2479451894760132,
+      "learning_rate": 9.971899150695463e-05,
+      "loss": 0.7808589339256287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6679611650485436,
+      "grad_norm": 1.2538350820541382,
+      "learning_rate": 9.971638602177175e-05,
+      "loss": 0.7096936106681824,
+      "step": 1548
+    },
+    {
+      "epoch": 0.668824163969795,
+      "grad_norm": 1.1431525945663452,
+      "learning_rate": 9.97137685477065e-05,
+      "loss": 0.6529367566108704,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6696871628910464,
+      "grad_norm": 1.5966559648513794,
+      "learning_rate": 9.971113908539008e-05,
+      "loss": 0.7780346870422363,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6705501618122978,
+      "grad_norm": 1.186475157737732,
+      "learning_rate": 9.970849763545658e-05,
+      "loss": 0.6889166831970215,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6714131607335491,
+      "grad_norm": 1.1922476291656494,
+      "learning_rate": 9.970584419854295e-05,
+      "loss": 0.7154867649078369,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6722761596548005,
+      "grad_norm": 1.30132257938385,
+      "learning_rate": 9.970317877528906e-05,
+      "loss": 0.7419366240501404,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6731391585760518,
+      "grad_norm": 1.1539888381958008,
+      "learning_rate": 9.970050136633769e-05,
+      "loss": 0.7856602072715759,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6740021574973031,
+      "grad_norm": 1.3056161403656006,
+      "learning_rate": 9.969781197233445e-05,
+      "loss": 0.6834115386009216,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6748651564185545,
+      "grad_norm": 1.2899249792099,
+      "learning_rate": 9.96951105939279e-05,
+      "loss": 0.8490727543830872,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6757281553398058,
+      "grad_norm": 1.2182374000549316,
+      "learning_rate": 9.969239723176946e-05,
+      "loss": 0.7462526559829712,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6765911542610572,
+      "grad_norm": 1.1231273412704468,
+      "learning_rate": 9.968967188651344e-05,
+      "loss": 0.7145209908485413,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6774541531823085,
+      "grad_norm": 1.1776760816574097,
+      "learning_rate": 9.968693455881703e-05,
+      "loss": 0.7911707162857056,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6783171521035599,
+      "grad_norm": 1.3542319536209106,
+      "learning_rate": 9.968418524934035e-05,
+      "loss": 0.6946635246276855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6791801510248112,
+      "grad_norm": 1.3073060512542725,
+      "learning_rate": 9.96814239587464e-05,
+      "loss": 0.7597488164901733,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6800431499460625,
+      "grad_norm": 1.3205567598342896,
+      "learning_rate": 9.967865068770099e-05,
+      "loss": 0.6763246655464172,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6809061488673139,
+      "grad_norm": 1.1734344959259033,
+      "learning_rate": 9.967586543687295e-05,
+      "loss": 0.7560655474662781,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6817691477885652,
+      "grad_norm": 1.445770025253296,
+      "learning_rate": 9.967306820693388e-05,
+      "loss": 0.8050205707550049,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6826321467098166,
+      "grad_norm": 1.1185686588287354,
+      "learning_rate": 9.967025899855836e-05,
+      "loss": 0.6628211140632629,
+      "step": 1582
+    },
+    {
+      "epoch": 0.683495145631068,
+      "grad_norm": 1.3831934928894043,
+      "learning_rate": 9.966743781242379e-05,
+      "loss": 0.7684114575386047,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6843581445523194,
+      "grad_norm": 1.2417019605636597,
+      "learning_rate": 9.966460464921047e-05,
+      "loss": 0.7084140777587891,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6852211434735707,
+      "grad_norm": 1.2138086557388306,
+      "learning_rate": 9.966175950960163e-05,
+      "loss": 0.6817399263381958,
+      "step": 1588
+    },
+    {
+      "epoch": 0.686084142394822,
+      "grad_norm": 1.399778962135315,
+      "learning_rate": 9.965890239428339e-05,
+      "loss": 0.8422354459762573,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6869471413160734,
+      "grad_norm": 1.4348714351654053,
+      "learning_rate": 9.965603330394468e-05,
+      "loss": 0.7432681322097778,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6878101402373247,
+      "grad_norm": 1.192425012588501,
+      "learning_rate": 9.965315223927741e-05,
+      "loss": 0.7571184635162354,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6886731391585761,
+      "grad_norm": 1.2834038734436035,
+      "learning_rate": 9.96502592009763e-05,
+      "loss": 0.802443265914917,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6895361380798274,
+      "grad_norm": 1.1650444269180298,
+      "learning_rate": 9.964735418973901e-05,
+      "loss": 0.8113405704498291,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "grad_norm": 1.1714438199996948,
+      "learning_rate": 9.964443720626606e-05,
+      "loss": 0.7507118582725525,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "eval_loss": 0.7706249952316284,
+      "eval_runtime": 821.4894,
+      "eval_samples_per_second": 2.508,
+      "eval_steps_per_second": 2.508,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6912621359223301,
+      "grad_norm": 1.175459861755371,
+      "learning_rate": 9.964150825126088e-05,
+      "loss": 0.7422796487808228,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6921251348435814,
+      "grad_norm": 1.6418036222457886,
+      "learning_rate": 9.963856732542978e-05,
+      "loss": 0.7125853300094604,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6929881337648328,
+      "grad_norm": 1.3256275653839111,
+      "learning_rate": 9.963561442948193e-05,
+      "loss": 0.757859468460083,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6938511326860841,
+      "grad_norm": 1.19981050491333,
+      "learning_rate": 9.963264956412944e-05,
+      "loss": 0.6849292516708374,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6947141316073355,
+      "grad_norm": 1.0392169952392578,
+      "learning_rate": 9.962967273008723e-05,
+      "loss": 0.7636215090751648,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6955771305285868,
+      "grad_norm": 1.1106646060943604,
+      "learning_rate": 9.962668392807318e-05,
+      "loss": 0.7208617329597473,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6964401294498382,
+      "grad_norm": 1.1569621562957764,
+      "learning_rate": 9.962368315880801e-05,
+      "loss": 0.671805739402771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6973031283710895,
+      "grad_norm": 1.2157769203186035,
+      "learning_rate": 9.962067042301537e-05,
+      "loss": 0.7648762464523315,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6981661272923408,
+      "grad_norm": 1.4804017543792725,
+      "learning_rate": 9.961764572142173e-05,
+      "loss": 0.8490509390830994,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 1.0344431400299072,
+      "learning_rate": 9.961460905475651e-05,
+      "loss": 0.5950891971588135,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6998921251348436,
+      "grad_norm": 1.1109060049057007,
+      "learning_rate": 9.9611560423752e-05,
+      "loss": 0.7288329005241394,
+      "step": 1622
+    },
+    {
+      "epoch": 0.700755124056095,
+      "grad_norm": 1.096996545791626,
+      "learning_rate": 9.960849982914331e-05,
+      "loss": 0.7074542045593262,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7016181229773463,
+      "grad_norm": 1.3109341859817505,
+      "learning_rate": 9.960542727166853e-05,
+      "loss": 0.7348865270614624,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7024811218985976,
+      "grad_norm": 1.3402032852172852,
+      "learning_rate": 9.96023427520686e-05,
+      "loss": 0.7658320665359497,
+      "step": 1628
+    },
+    {
+      "epoch": 0.703344120819849,
+      "grad_norm": 1.3712561130523682,
+      "learning_rate": 9.959924627108732e-05,
+      "loss": 0.809951901435852,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7042071197411003,
+      "grad_norm": 1.2278292179107666,
+      "learning_rate": 9.959613782947139e-05,
+      "loss": 0.6834598779678345,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7050701186623517,
+      "grad_norm": 1.0510832071304321,
+      "learning_rate": 9.95930174279704e-05,
+      "loss": 0.7416030764579773,
+      "step": 1634
+    },
+    {
+      "epoch": 0.705933117583603,
+      "grad_norm": 1.130991816520691,
+      "learning_rate": 9.958988506733684e-05,
+      "loss": 0.7304092645645142,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7067961165048544,
+      "grad_norm": 1.3328672647476196,
+      "learning_rate": 9.958674074832603e-05,
+      "loss": 0.8230003118515015,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7076591154261057,
+      "grad_norm": 1.3842110633850098,
+      "learning_rate": 9.958358447169625e-05,
+      "loss": 0.8413107991218567,
+      "step": 1640
+    },
+    {
+      "epoch": 0.708522114347357,
+      "grad_norm": 1.3143812417984009,
+      "learning_rate": 9.958041623820858e-05,
+      "loss": 0.77432781457901,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7093851132686084,
+      "grad_norm": 1.4727009534835815,
+      "learning_rate": 9.957723604862704e-05,
+      "loss": 0.7563844323158264,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7102481121898597,
+      "grad_norm": 1.1892327070236206,
+      "learning_rate": 9.957404390371854e-05,
+      "loss": 0.7085853219032288,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.124180555343628,
+      "learning_rate": 9.957083980425282e-05,
+      "loss": 0.6837730407714844,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 1.2224774360656738,
+      "learning_rate": 9.956762375100255e-05,
+      "loss": 0.7255740165710449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7128371089536139,
+      "grad_norm": 1.360425591468811,
+      "learning_rate": 9.956439574474326e-05,
+      "loss": 0.7849710583686829,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7137001078748652,
+      "grad_norm": 1.2818151712417603,
+      "learning_rate": 9.956115578625338e-05,
+      "loss": 0.7669665813446045,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7145631067961165,
+      "grad_norm": 1.2415164709091187,
+      "learning_rate": 9.95579038763142e-05,
+      "loss": 0.7052890658378601,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7154261057173679,
+      "grad_norm": 1.3983269929885864,
+      "learning_rate": 9.955464001570992e-05,
+      "loss": 0.8300023674964905,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7162891046386192,
+      "grad_norm": 1.1591373682022095,
+      "learning_rate": 9.955136420522759e-05,
+      "loss": 0.7742894887924194,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7171521035598706,
+      "grad_norm": 1.2503128051757812,
+      "learning_rate": 9.954807644565715e-05,
+      "loss": 0.8458141684532166,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7180151024811219,
+      "grad_norm": 1.2954912185668945,
+      "learning_rate": 9.954477673779146e-05,
+      "loss": 0.7044742107391357,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7188781014023733,
+      "grad_norm": 1.1669788360595703,
+      "learning_rate": 9.954146508242619e-05,
+      "loss": 0.7917183041572571,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7197411003236246,
+      "grad_norm": 1.1466871500015259,
+      "learning_rate": 9.953814148035996e-05,
+      "loss": 0.7174562811851501,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7206040992448759,
+      "grad_norm": 1.367224097251892,
+      "learning_rate": 9.953480593239421e-05,
+      "loss": 0.8925880193710327,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7214670981661273,
+      "grad_norm": 1.049497365951538,
+      "learning_rate": 9.953145843933334e-05,
+      "loss": 0.70949387550354,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7223300970873786,
+      "grad_norm": 0.9995084404945374,
+      "learning_rate": 9.952809900198455e-05,
+      "loss": 0.6731935143470764,
+      "step": 1674
+    },
+    {
+      "epoch": 0.72319309600863,
+      "grad_norm": 1.100899577140808,
+      "learning_rate": 9.952472762115794e-05,
+      "loss": 0.7316275835037231,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7240560949298813,
+      "grad_norm": 1.2594135999679565,
+      "learning_rate": 9.952134429766655e-05,
+      "loss": 0.7379520535469055,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7249190938511327,
+      "grad_norm": 1.3124911785125732,
+      "learning_rate": 9.951794903232621e-05,
+      "loss": 0.7072247266769409,
+      "step": 1680
+    },
+    {
+      "epoch": 0.725782092772384,
+      "grad_norm": 1.212144374847412,
+      "learning_rate": 9.95145418259557e-05,
+      "loss": 0.745680034160614,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7266450916936353,
+      "grad_norm": 1.3913586139678955,
+      "learning_rate": 9.951112267937663e-05,
+      "loss": 0.796852707862854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7275080906148867,
+      "grad_norm": 1.3227289915084839,
+      "learning_rate": 9.950769159341353e-05,
+      "loss": 0.700892984867096,
+      "step": 1686
+    },
+    {
+      "epoch": 0.728371089536138,
+      "grad_norm": 1.4656381607055664,
+      "learning_rate": 9.950424856889378e-05,
+      "loss": 0.7973095178604126,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7292340884573895,
+      "grad_norm": 1.1698973178863525,
+      "learning_rate": 9.950079360664766e-05,
+      "loss": 0.7811378836631775,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7300970873786408,
+      "grad_norm": 1.201695203781128,
+      "learning_rate": 9.949732670750831e-05,
+      "loss": 0.753012478351593,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7309600862998922,
+      "grad_norm": 1.3465325832366943,
+      "learning_rate": 9.949384787231175e-05,
+      "loss": 0.7620642781257629,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7318230852211435,
+      "grad_norm": 1.2191137075424194,
+      "learning_rate": 9.94903571018969e-05,
+      "loss": 0.7479873299598694,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7326860841423948,
+      "grad_norm": 1.1412187814712524,
+      "learning_rate": 9.948685439710554e-05,
+      "loss": 0.7327720522880554,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "grad_norm": 1.310327410697937,
+      "learning_rate": 9.948333975878233e-05,
+      "loss": 0.761475682258606,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "eval_loss": 0.7649155259132385,
+      "eval_runtime": 828.3542,
+      "eval_samples_per_second": 2.487,
+      "eval_steps_per_second": 2.487,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7344120819848975,
+      "grad_norm": 1.3031405210494995,
+      "learning_rate": 9.947981318777482e-05,
+      "loss": 0.7564467191696167,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7352750809061489,
+      "grad_norm": 1.2672377824783325,
+      "learning_rate": 9.94762746849334e-05,
+      "loss": 0.7689576745033264,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7361380798274002,
+      "grad_norm": 1.1159595251083374,
+      "learning_rate": 9.947272425111138e-05,
+      "loss": 0.7923810482025146,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7370010787486516,
+      "grad_norm": 1.1190921068191528,
+      "learning_rate": 9.946916188716495e-05,
+      "loss": 0.6540790796279907,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 1.373832106590271,
+      "learning_rate": 9.946558759395312e-05,
+      "loss": 0.7267473936080933,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7387270765911542,
+      "grad_norm": 1.2346374988555908,
+      "learning_rate": 9.946200137233785e-05,
+      "loss": 0.7693060636520386,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7395900755124056,
+      "grad_norm": 1.1629668474197388,
+      "learning_rate": 9.945840322318391e-05,
+      "loss": 0.7751407027244568,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7404530744336569,
+      "grad_norm": 1.2268708944320679,
+      "learning_rate": 9.945479314735899e-05,
+      "loss": 0.7798545360565186,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7413160733549083,
+      "grad_norm": 0.9968738555908203,
+      "learning_rate": 9.945117114573366e-05,
+      "loss": 0.6128867268562317,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7421790722761596,
+      "grad_norm": 1.2135992050170898,
+      "learning_rate": 9.944753721918133e-05,
+      "loss": 0.7536954283714294,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7430420711974111,
+      "grad_norm": 1.1361218690872192,
+      "learning_rate": 9.94438913685783e-05,
+      "loss": 0.7034479379653931,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7439050701186624,
+      "grad_norm": 1.0728967189788818,
+      "learning_rate": 9.944023359480378e-05,
+      "loss": 0.7285382747650146,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7447680690399137,
+      "grad_norm": 1.3613945245742798,
+      "learning_rate": 9.94365638987398e-05,
+      "loss": 0.7865195274353027,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7456310679611651,
+      "grad_norm": 1.0985898971557617,
+      "learning_rate": 9.94328822812713e-05,
+      "loss": 0.7511224746704102,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7464940668824164,
+      "grad_norm": 1.1506208181381226,
+      "learning_rate": 9.942918874328606e-05,
+      "loss": 0.732453465461731,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7473570658036678,
+      "grad_norm": 1.0995100736618042,
+      "learning_rate": 9.94254832856748e-05,
+      "loss": 0.702409029006958,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7482200647249191,
+      "grad_norm": 1.0075604915618896,
+      "learning_rate": 9.942176590933106e-05,
+      "loss": 0.7515703439712524,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7490830636461704,
+      "grad_norm": 1.3360812664031982,
+      "learning_rate": 9.941803661515126e-05,
+      "loss": 0.6910237073898315,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7499460625674218,
+      "grad_norm": 1.0172319412231445,
+      "learning_rate": 9.941429540403471e-05,
+      "loss": 0.7185606956481934,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7508090614886731,
+      "grad_norm": 1.2721658945083618,
+      "learning_rate": 9.941054227688358e-05,
+      "loss": 0.7136536836624146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7516720604099245,
+      "grad_norm": 1.0775413513183594,
+      "learning_rate": 9.940677723460292e-05,
+      "loss": 0.7595619559288025,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7525350593311758,
+      "grad_norm": 1.3436479568481445,
+      "learning_rate": 9.940300027810067e-05,
+      "loss": 0.7602618336677551,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7533980582524272,
+      "grad_norm": 1.2611958980560303,
+      "learning_rate": 9.93992114082876e-05,
+      "loss": 0.7954604029655457,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7542610571736785,
+      "grad_norm": 1.2794841527938843,
+      "learning_rate": 9.93954106260774e-05,
+      "loss": 0.7661190629005432,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7551240560949298,
+      "grad_norm": 1.3042595386505127,
+      "learning_rate": 9.939159793238661e-05,
+      "loss": 0.6855312585830688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7559870550161812,
+      "grad_norm": 1.2799923419952393,
+      "learning_rate": 9.938777332813463e-05,
+      "loss": 0.7970905900001526,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7568500539374325,
+      "grad_norm": 1.169205904006958,
+      "learning_rate": 9.938393681424375e-05,
+      "loss": 0.7072843313217163,
+      "step": 1754
+    },
+    {
+      "epoch": 0.757713052858684,
+      "grad_norm": 1.3833354711532593,
+      "learning_rate": 9.938008839163913e-05,
+      "loss": 0.7134629487991333,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7585760517799353,
+      "grad_norm": 1.223452091217041,
+      "learning_rate": 9.93762280612488e-05,
+      "loss": 0.7873107194900513,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7594390507011867,
+      "grad_norm": 1.2740365266799927,
+      "learning_rate": 9.937235582400367e-05,
+      "loss": 0.7680654525756836,
+      "step": 1760
+    },
+    {
+      "epoch": 0.760302049622438,
+      "grad_norm": 1.2564697265625,
+      "learning_rate": 9.93684716808375e-05,
+      "loss": 0.734126627445221,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7611650485436893,
+      "grad_norm": 1.3858847618103027,
+      "learning_rate": 9.936457563268694e-05,
+      "loss": 0.7972320914268494,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7620280474649407,
+      "grad_norm": 1.2705003023147583,
+      "learning_rate": 9.93606676804915e-05,
+      "loss": 0.7904276847839355,
+      "step": 1766
+    },
+    {
+      "epoch": 0.762891046386192,
+      "grad_norm": 1.210766077041626,
+      "learning_rate": 9.935674782519358e-05,
+      "loss": 0.7542425990104675,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7637540453074434,
+      "grad_norm": 1.4367541074752808,
+      "learning_rate": 9.93528160677384e-05,
+      "loss": 0.7975004315376282,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7646170442286947,
+      "grad_norm": 1.0994079113006592,
+      "learning_rate": 9.934887240907414e-05,
+      "loss": 0.7537807822227478,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7654800431499461,
+      "grad_norm": 1.111660122871399,
+      "learning_rate": 9.934491685015174e-05,
+      "loss": 0.7333855628967285,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7663430420711974,
+      "grad_norm": 1.2879043817520142,
+      "learning_rate": 9.934094939192508e-05,
+      "loss": 0.7472840547561646,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7672060409924487,
+      "grad_norm": 1.3618478775024414,
+      "learning_rate": 9.933697003535094e-05,
+      "loss": 0.7644904255867004,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7680690399137001,
+      "grad_norm": 1.1235666275024414,
+      "learning_rate": 9.933297878138887e-05,
+      "loss": 0.7872052192687988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7689320388349514,
+      "grad_norm": 1.3741775751113892,
+      "learning_rate": 9.932897563100136e-05,
+      "loss": 0.7343124747276306,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7697950377562028,
+      "grad_norm": 1.2115039825439453,
+      "learning_rate": 9.932496058515375e-05,
+      "loss": 0.7486474514007568,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7706580366774541,
+      "grad_norm": 1.183457851409912,
+      "learning_rate": 9.932093364481426e-05,
+      "loss": 0.7333025932312012,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7715210355987056,
+      "grad_norm": 1.2841869592666626,
+      "learning_rate": 9.931689481095394e-05,
+      "loss": 0.7241485118865967,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7723840345199569,
+      "grad_norm": 1.2312977313995361,
+      "learning_rate": 9.931284408454678e-05,
+      "loss": 0.7510396242141724,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7732470334412082,
+      "grad_norm": 1.144895315170288,
+      "learning_rate": 9.930878146656956e-05,
+      "loss": 0.7108346819877625,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7741100323624596,
+      "grad_norm": 1.18603515625,
+      "learning_rate": 9.9304706958002e-05,
+      "loss": 0.7687607407569885,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7749730312837109,
+      "grad_norm": 1.2827935218811035,
+      "learning_rate": 9.93006205598266e-05,
+      "loss": 0.7826682329177856,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7758360302049623,
+      "grad_norm": 1.164852261543274,
+      "learning_rate": 9.929652227302881e-05,
+      "loss": 0.6970428824424744,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 1.0439956188201904,
+      "learning_rate": 9.92924120985969e-05,
+      "loss": 0.7238826751708984,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "eval_loss": 0.7598537802696228,
+      "eval_runtime": 838.9645,
+      "eval_samples_per_second": 2.455,
+      "eval_steps_per_second": 2.455,
+      "step": 1800
+    },
+    {
+      "epoch": 0.777562028047465,
+      "grad_norm": 1.2464938163757324,
+      "learning_rate": 9.928829003752202e-05,
+      "loss": 0.7723641395568848,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7784250269687163,
+      "grad_norm": 1.2445050477981567,
+      "learning_rate": 9.92841560907982e-05,
+      "loss": 0.7154937982559204,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7792880258899676,
+      "grad_norm": 1.2235661745071411,
+      "learning_rate": 9.928001025942231e-05,
+      "loss": 0.8325800895690918,
+      "step": 1806
+    },
+    {
+      "epoch": 0.780151024811219,
+      "grad_norm": 1.268066644668579,
+      "learning_rate": 9.927585254439409e-05,
+      "loss": 0.7093122601509094,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7810140237324703,
+      "grad_norm": 1.1731390953063965,
+      "learning_rate": 9.927168294671618e-05,
+      "loss": 0.7601775527000427,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7818770226537217,
+      "grad_norm": 1.0685608386993408,
+      "learning_rate": 9.926750146739403e-05,
+      "loss": 0.6889236569404602,
+      "step": 1812
+    },
+    {
+      "epoch": 0.782740021574973,
+      "grad_norm": 1.1098639965057373,
+      "learning_rate": 9.926330810743602e-05,
+      "loss": 0.7397196888923645,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7836030204962244,
+      "grad_norm": 1.3606436252593994,
+      "learning_rate": 9.925910286785332e-05,
+      "loss": 0.7180029153823853,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7844660194174757,
+      "grad_norm": 1.143418550491333,
+      "learning_rate": 9.925488574966004e-05,
+      "loss": 0.7039036750793457,
+      "step": 1818
+    },
+    {
+      "epoch": 0.785329018338727,
+      "grad_norm": 1.180968999862671,
+      "learning_rate": 9.92506567538731e-05,
+      "loss": 0.741872251033783,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7861920172599784,
+      "grad_norm": 1.1953548192977905,
+      "learning_rate": 9.924641588151231e-05,
+      "loss": 0.7580198049545288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7870550161812297,
+      "grad_norm": 1.3049067258834839,
+      "learning_rate": 9.924216313360033e-05,
+      "loss": 0.7479216456413269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7879180151024812,
+      "grad_norm": 1.3962669372558594,
+      "learning_rate": 9.92378985111627e-05,
+      "loss": 0.7382357716560364,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7887810140237325,
+      "grad_norm": 1.1519217491149902,
+      "learning_rate": 9.92336220152278e-05,
+      "loss": 0.6356467604637146,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7896440129449838,
+      "grad_norm": 1.2544089555740356,
+      "learning_rate": 9.922933364682689e-05,
+      "loss": 0.7674193978309631,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7905070118662352,
+      "grad_norm": 1.1358295679092407,
+      "learning_rate": 9.922503340699412e-05,
+      "loss": 0.7083898782730103,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7913700107874865,
+      "grad_norm": 1.1744481325149536,
+      "learning_rate": 9.922072129676644e-05,
+      "loss": 0.6273066401481628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7922330097087379,
+      "grad_norm": 1.20509672164917,
+      "learning_rate": 9.92163973171837e-05,
+      "loss": 0.808181643486023,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7930960086299892,
+      "grad_norm": 1.2035597562789917,
+      "learning_rate": 9.921206146928863e-05,
+      "loss": 0.6631553769111633,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7939590075512406,
+      "grad_norm": 1.058695912361145,
+      "learning_rate": 9.920771375412678e-05,
+      "loss": 0.72016841173172,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7948220064724919,
+      "grad_norm": 1.035791039466858,
+      "learning_rate": 9.92033541727466e-05,
+      "loss": 0.7240060567855835,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7956850053937432,
+      "grad_norm": 1.1845753192901611,
+      "learning_rate": 9.919898272619935e-05,
+      "loss": 0.743224561214447,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7965480043149946,
+      "grad_norm": 1.2942636013031006,
+      "learning_rate": 9.919459941553922e-05,
+      "loss": 0.7064419388771057,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7974110032362459,
+      "grad_norm": 1.2736715078353882,
+      "learning_rate": 9.91902042418232e-05,
+      "loss": 0.805932343006134,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7982740021574973,
+      "grad_norm": 1.2803614139556885,
+      "learning_rate": 9.918579720611119e-05,
+      "loss": 0.8050802946090698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7991370010787486,
+      "grad_norm": 1.2225072383880615,
+      "learning_rate": 9.91813783094659e-05,
+      "loss": 0.7013180255889893,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3375571966171265,
+      "learning_rate": 9.917694755295294e-05,
+      "loss": 0.7415283918380737,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8008629989212513,
+      "grad_norm": 1.2209177017211914,
+      "learning_rate": 9.917250493764079e-05,
+      "loss": 0.7264692783355713,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8017259978425026,
+      "grad_norm": 1.1158864498138428,
+      "learning_rate": 9.916805046460074e-05,
+      "loss": 0.7083719372749329,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8025889967637541,
+      "grad_norm": 1.171000361442566,
+      "learning_rate": 9.916358413490696e-05,
+      "loss": 0.820396900177002,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8034519956850054,
+      "grad_norm": 0.9942941069602966,
+      "learning_rate": 9.91591059496365e-05,
+      "loss": 0.6754014492034912,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8043149946062568,
+      "grad_norm": 1.239795207977295,
+      "learning_rate": 9.915461590986926e-05,
+      "loss": 0.813154399394989,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8051779935275081,
+      "grad_norm": 1.076019525527954,
+      "learning_rate": 9.915011401668796e-05,
+      "loss": 0.745580792427063,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8060409924487595,
+      "grad_norm": 1.2134130001068115,
+      "learning_rate": 9.914560027117826e-05,
+      "loss": 0.7564949989318848,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8069039913700108,
+      "grad_norm": 1.1774994134902954,
+      "learning_rate": 9.914107467442859e-05,
+      "loss": 0.665183424949646,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8077669902912621,
+      "grad_norm": 1.0606465339660645,
+      "learning_rate": 9.91365372275303e-05,
+      "loss": 0.7597744464874268,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8086299892125135,
+      "grad_norm": 1.397161841392517,
+      "learning_rate": 9.913198793157757e-05,
+      "loss": 0.7453591227531433,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8094929881337648,
+      "grad_norm": 1.250003695487976,
+      "learning_rate": 9.912742678766742e-05,
+      "loss": 0.715792179107666,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8103559870550162,
+      "grad_norm": 1.2343473434448242,
+      "learning_rate": 9.912285379689979e-05,
+      "loss": 0.7094268798828125,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8112189859762675,
+      "grad_norm": 1.453764796257019,
+      "learning_rate": 9.911826896037739e-05,
+      "loss": 0.6652816534042358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8120819848975189,
+      "grad_norm": 1.030903935432434,
+      "learning_rate": 9.911367227920587e-05,
+      "loss": 0.6764529347419739,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8129449838187702,
+      "grad_norm": 1.2725969552993774,
+      "learning_rate": 9.91090637544937e-05,
+      "loss": 0.7043389678001404,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8138079827400215,
+      "grad_norm": 1.1610769033432007,
+      "learning_rate": 9.910444338735216e-05,
+      "loss": 0.6960558295249939,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8146709816612729,
+      "grad_norm": 1.2049180269241333,
+      "learning_rate": 9.909981117889548e-05,
+      "loss": 0.8029319643974304,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 1.1671873331069946,
+      "learning_rate": 9.909516713024068e-05,
+      "loss": 0.683114767074585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8163969795037757,
+      "grad_norm": 1.247768759727478,
+      "learning_rate": 9.909051124250764e-05,
+      "loss": 0.7193703055381775,
+      "step": 1892
+    },
+    {
+      "epoch": 0.817259978425027,
+      "grad_norm": 1.1376357078552246,
+      "learning_rate": 9.908584351681911e-05,
+      "loss": 0.7957084774971008,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8181229773462784,
+      "grad_norm": 1.1918323040008545,
+      "learning_rate": 9.90811639543007e-05,
+      "loss": 0.7168112993240356,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8189859762675297,
+      "grad_norm": 1.1249650716781616,
+      "learning_rate": 9.907647255608087e-05,
+      "loss": 0.7519245147705078,
+      "step": 1898
+    },
+    {
+      "epoch": 0.819848975188781,
+      "grad_norm": 1.0688576698303223,
+      "learning_rate": 9.907176932329092e-05,
+      "loss": 0.7496176362037659,
+      "step": 1900
+    },
+    {
+      "epoch": 0.819848975188781,
+      "eval_loss": 0.7533769607543945,
+      "eval_runtime": 830.2894,
+      "eval_samples_per_second": 2.481,
+      "eval_steps_per_second": 2.481,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8207119741100324,
+      "grad_norm": 0.9264270663261414,
+      "learning_rate": 9.906705425706499e-05,
+      "loss": 0.639653205871582,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8215749730312837,
+      "grad_norm": 1.2231628894805908,
+      "learning_rate": 9.906232735854016e-05,
+      "loss": 0.7555551528930664,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8224379719525351,
+      "grad_norm": 1.3369156122207642,
+      "learning_rate": 9.905758862885624e-05,
+      "loss": 0.7113592028617859,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8233009708737864,
+      "grad_norm": 1.217238187789917,
+      "learning_rate": 9.905283806915597e-05,
+      "loss": 0.6823481321334839,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8241639697950378,
+      "grad_norm": 1.2462831735610962,
+      "learning_rate": 9.904807568058495e-05,
+      "loss": 0.7347100377082825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8250269687162891,
+      "grad_norm": 1.2507221698760986,
+      "learning_rate": 9.904330146429157e-05,
+      "loss": 0.7981513738632202,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8258899676375404,
+      "grad_norm": 1.2246814966201782,
+      "learning_rate": 9.903851542142715e-05,
+      "loss": 0.759779155254364,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8267529665587918,
+      "grad_norm": 1.1746087074279785,
+      "learning_rate": 9.903371755314579e-05,
+      "loss": 0.741083562374115,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8276159654800431,
+      "grad_norm": 1.1913539171218872,
+      "learning_rate": 9.902890786060448e-05,
+      "loss": 0.7491101622581482,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8284789644012945,
+      "grad_norm": 1.2342501878738403,
+      "learning_rate": 9.902408634496307e-05,
+      "loss": 0.7655978202819824,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8293419633225458,
+      "grad_norm": 1.075892686843872,
+      "learning_rate": 9.901925300738424e-05,
+      "loss": 0.7199848890304565,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8302049622437971,
+      "grad_norm": 1.1517611742019653,
+      "learning_rate": 9.901440784903353e-05,
+      "loss": 0.8325693011283875,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8310679611650486,
+      "grad_norm": 1.2084628343582153,
+      "learning_rate": 9.900955087107933e-05,
+      "loss": 0.7674157023429871,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8319309600862999,
+      "grad_norm": 1.334679365158081,
+      "learning_rate": 9.900468207469287e-05,
+      "loss": 0.7415181398391724,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8327939590075513,
+      "grad_norm": 1.0505887269973755,
+      "learning_rate": 9.899980146104824e-05,
+      "loss": 0.672336220741272,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8336569579288026,
+      "grad_norm": 1.415797233581543,
+      "learning_rate": 9.899490903132238e-05,
+      "loss": 0.8077969551086426,
+      "step": 1932
+    },
+    {
+      "epoch": 0.834519956850054,
+      "grad_norm": 1.1793439388275146,
+      "learning_rate": 9.899000478669507e-05,
+      "loss": 0.8053045272827148,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8353829557713053,
+      "grad_norm": 1.196115493774414,
+      "learning_rate": 9.898508872834896e-05,
+      "loss": 0.7585914731025696,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8362459546925566,
+      "grad_norm": 1.2181763648986816,
+      "learning_rate": 9.898016085746953e-05,
+      "loss": 0.691072940826416,
+      "step": 1938
+    },
+    {
+      "epoch": 0.837108953613808,
+      "grad_norm": 1.2929240465164185,
+      "learning_rate": 9.897522117524512e-05,
+      "loss": 0.7723121643066406,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8379719525350593,
+      "grad_norm": 1.2859764099121094,
+      "learning_rate": 9.89702696828669e-05,
+      "loss": 0.8236626386642456,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8388349514563107,
+      "grad_norm": 1.2043784856796265,
+      "learning_rate": 9.89653063815289e-05,
+      "loss": 0.7315511703491211,
+      "step": 1944
+    },
+    {
+      "epoch": 0.839697950377562,
+      "grad_norm": 1.2063398361206055,
+      "learning_rate": 9.896033127242802e-05,
+      "loss": 0.6707082986831665,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8405609492988134,
+      "grad_norm": 1.1906715631484985,
+      "learning_rate": 9.895534435676393e-05,
+      "loss": 0.6920633912086487,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 1.122567892074585,
+      "learning_rate": 9.895034563573926e-05,
+      "loss": 0.7134768962860107,
+      "step": 1950
+    },
+    {
+      "epoch": 0.842286947141316,
+      "grad_norm": 1.0681787729263306,
+      "learning_rate": 9.894533511055943e-05,
+      "loss": 0.6333967447280884,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8431499460625674,
+      "grad_norm": 1.5216220617294312,
+      "learning_rate": 9.894031278243266e-05,
+      "loss": 0.7744026780128479,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8440129449838187,
+      "grad_norm": 1.0934535264968872,
+      "learning_rate": 9.89352786525701e-05,
+      "loss": 0.7678086757659912,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8448759439050701,
+      "grad_norm": 1.223285436630249,
+      "learning_rate": 9.89302327221857e-05,
+      "loss": 0.7344630360603333,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8457389428263214,
+      "grad_norm": 1.1123322248458862,
+      "learning_rate": 9.892517499249623e-05,
+      "loss": 0.76033616065979,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8466019417475729,
+      "grad_norm": 1.0472193956375122,
+      "learning_rate": 9.89201054647214e-05,
+      "loss": 0.7176017761230469,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8474649406688242,
+      "grad_norm": 1.8538368940353394,
+      "learning_rate": 9.891502414008365e-05,
+      "loss": 0.7220107316970825,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8483279395900755,
+      "grad_norm": 1.2280668020248413,
+      "learning_rate": 9.890993101980834e-05,
+      "loss": 0.7399556636810303,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8491909385113269,
+      "grad_norm": 1.1482552289962769,
+      "learning_rate": 9.890482610512364e-05,
+      "loss": 0.7700466513633728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8500539374325782,
+      "grad_norm": 1.1877986192703247,
+      "learning_rate": 9.889970939726061e-05,
+      "loss": 0.6666419506072998,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8509169363538296,
+      "grad_norm": 1.407792329788208,
+      "learning_rate": 9.88945808974531e-05,
+      "loss": 0.7883278131484985,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8517799352750809,
+      "grad_norm": 1.1426323652267456,
+      "learning_rate": 9.888944060693781e-05,
+      "loss": 0.7441611289978027,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8526429341963323,
+      "grad_norm": 1.2660622596740723,
+      "learning_rate": 9.888428852695432e-05,
+      "loss": 0.6464171409606934,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8535059331175836,
+      "grad_norm": 1.1639022827148438,
+      "learning_rate": 9.887912465874502e-05,
+      "loss": 0.7242478728294373,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 1.2546422481536865,
+      "learning_rate": 9.887394900355516e-05,
+      "loss": 0.7431875467300415,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8552319309600863,
+      "grad_norm": 1.1057130098342896,
+      "learning_rate": 9.886876156263282e-05,
+      "loss": 0.7662370204925537,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8560949298813376,
+      "grad_norm": 1.1017125844955444,
+      "learning_rate": 9.886356233722894e-05,
+      "loss": 0.7120938897132874,
+      "step": 1984
+    },
+    {
+      "epoch": 0.856957928802589,
+      "grad_norm": 1.3548704385757446,
+      "learning_rate": 9.885835132859727e-05,
+      "loss": 0.789219856262207,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8578209277238403,
+      "grad_norm": 1.1920398473739624,
+      "learning_rate": 9.885312853799443e-05,
+      "loss": 0.7599530816078186,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8586839266450917,
+      "grad_norm": 1.170946478843689,
+      "learning_rate": 9.884789396667988e-05,
+      "loss": 0.8315809369087219,
+      "step": 1990
+    },
+    {
+      "epoch": 0.859546925566343,
+      "grad_norm": 1.2530180215835571,
+      "learning_rate": 9.88426476159159e-05,
+      "loss": 0.713994026184082,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8604099244875943,
+      "grad_norm": 1.1056885719299316,
+      "learning_rate": 9.883738948696765e-05,
+      "loss": 0.6656507849693298,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8612729234088458,
+      "grad_norm": 1.169382929801941,
+      "learning_rate": 9.883211958110308e-05,
+      "loss": 0.7164257764816284,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8621359223300971,
+      "grad_norm": 1.1970325708389282,
+      "learning_rate": 9.882683789959299e-05,
+      "loss": 0.6885558366775513,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "grad_norm": 1.1700761318206787,
+      "learning_rate": 9.882154444371107e-05,
+      "loss": 0.7265812754631042,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "eval_loss": 0.7487027645111084,
+      "eval_runtime": 825.141,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8638619201725998,
+      "grad_norm": 1.3120243549346924,
+      "learning_rate": 9.88162392147338e-05,
+      "loss": 0.7371920347213745,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8647249190938512,
+      "grad_norm": 1.098697304725647,
+      "learning_rate": 9.881092221394049e-05,
+      "loss": 0.7453275918960571,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8655879180151025,
+      "grad_norm": 1.1961677074432373,
+      "learning_rate": 9.880559344261334e-05,
+      "loss": 0.7562432289123535,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8664509169363538,
+      "grad_norm": 1.1655291318893433,
+      "learning_rate": 9.880025290203735e-05,
+      "loss": 0.7654905319213867,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8673139158576052,
+      "grad_norm": 1.0979351997375488,
+      "learning_rate": 9.879490059350036e-05,
+      "loss": 0.6807888746261597,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8681769147788565,
+      "grad_norm": 1.0378475189208984,
+      "learning_rate": 9.878953651829306e-05,
+      "loss": 0.6887364387512207,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8690399137001079,
+      "grad_norm": 2.0228428840637207,
+      "learning_rate": 9.878416067770898e-05,
+      "loss": 0.718578577041626,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8699029126213592,
+      "grad_norm": 1.1610580682754517,
+      "learning_rate": 9.877877307304446e-05,
+      "loss": 0.7424600720405579,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8707659115426106,
+      "grad_norm": 1.1415833234786987,
+      "learning_rate": 9.877337370559873e-05,
+      "loss": 0.7450630068778992,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8716289104638619,
+      "grad_norm": 1.1008672714233398,
+      "learning_rate": 9.876796257667379e-05,
+      "loss": 0.7398725748062134,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8724919093851132,
+      "grad_norm": 1.1662763357162476,
+      "learning_rate": 9.876253968757452e-05,
+      "loss": 0.7122271060943604,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8733549083063646,
+      "grad_norm": 1.1835166215896606,
+      "learning_rate": 9.875710503960862e-05,
+      "loss": 0.7026485800743103,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8742179072276159,
+      "grad_norm": 1.1602214574813843,
+      "learning_rate": 9.875165863408666e-05,
+      "loss": 0.7211011052131653,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8750809061488674,
+      "grad_norm": 1.222589373588562,
+      "learning_rate": 9.874620047232198e-05,
+      "loss": 0.7565322518348694,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8759439050701187,
+      "grad_norm": 1.203661561012268,
+      "learning_rate": 9.87407305556308e-05,
+      "loss": 0.7173075675964355,
+      "step": 2030
+    },
+    {
+      "epoch": 0.87680690399137,
+      "grad_norm": 1.3316307067871094,
+      "learning_rate": 9.87352488853322e-05,
+      "loss": 0.820809006690979,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8776699029126214,
+      "grad_norm": 1.3071057796478271,
+      "learning_rate": 9.872975546274799e-05,
+      "loss": 0.8689352869987488,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8785329018338727,
+      "grad_norm": 1.3557701110839844,
+      "learning_rate": 9.872425028920294e-05,
+      "loss": 0.7090175151824951,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8793959007551241,
+      "grad_norm": 1.2234264612197876,
+      "learning_rate": 9.871873336602459e-05,
+      "loss": 0.6883202195167542,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8802588996763754,
+      "grad_norm": 1.3064714670181274,
+      "learning_rate": 9.87132046945433e-05,
+      "loss": 0.7134175896644592,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8811218985976268,
+      "grad_norm": 1.2502511739730835,
+      "learning_rate": 9.870766427609232e-05,
+      "loss": 0.7127227783203125,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8819848975188781,
+      "grad_norm": 1.1354259252548218,
+      "learning_rate": 9.870211211200766e-05,
+      "loss": 0.719437301158905,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8828478964401294,
+      "grad_norm": 1.1706442832946777,
+      "learning_rate": 9.869654820362821e-05,
+      "loss": 0.6933091282844543,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8837108953613808,
+      "grad_norm": 1.0227864980697632,
+      "learning_rate": 9.86909725522957e-05,
+      "loss": 0.6825079321861267,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8845738942826321,
+      "grad_norm": 1.103418231010437,
+      "learning_rate": 9.868538515935465e-05,
+      "loss": 0.7306133508682251,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8854368932038835,
+      "grad_norm": 1.0938395261764526,
+      "learning_rate": 9.867978602615245e-05,
+      "loss": 0.6956495046615601,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8862998921251348,
+      "grad_norm": 1.115692377090454,
+      "learning_rate": 9.86741751540393e-05,
+      "loss": 0.7740456461906433,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8871628910463862,
+      "grad_norm": 1.0264382362365723,
+      "learning_rate": 9.866855254436824e-05,
+      "loss": 0.6996002197265625,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8880258899676375,
+      "grad_norm": 1.2523163557052612,
+      "learning_rate": 9.866291819849511e-05,
+      "loss": 0.760357677936554,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.0595446825027466,
+      "learning_rate": 9.865727211777864e-05,
+      "loss": 0.6389813423156738,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8897518878101403,
+      "grad_norm": 1.198530912399292,
+      "learning_rate": 9.865161430358036e-05,
+      "loss": 0.7859060168266296,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8906148867313916,
+      "grad_norm": 1.0271193981170654,
+      "learning_rate": 9.864594475726461e-05,
+      "loss": 0.7120063900947571,
+      "step": 2064
+    },
+    {
+      "epoch": 0.891477885652643,
+      "grad_norm": 1.1619046926498413,
+      "learning_rate": 9.864026348019857e-05,
+      "loss": 0.6751878261566162,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8923408845738943,
+      "grad_norm": 1.2812464237213135,
+      "learning_rate": 9.863457047375227e-05,
+      "loss": 0.610360860824585,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 1.112628698348999,
+      "learning_rate": 9.862886573929855e-05,
+      "loss": 0.6540104746818542,
+      "step": 2070
+    },
+    {
+      "epoch": 0.894066882416397,
+      "grad_norm": 1.6257048845291138,
+      "learning_rate": 9.862314927821306e-05,
+      "loss": 0.726277232170105,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8949298813376483,
+      "grad_norm": 1.227334976196289,
+      "learning_rate": 9.861742109187432e-05,
+      "loss": 0.7798765301704407,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8957928802588997,
+      "grad_norm": 1.0728148221969604,
+      "learning_rate": 9.861168118166366e-05,
+      "loss": 0.637127161026001,
+      "step": 2076
+    },
+    {
+      "epoch": 0.896655879180151,
+      "grad_norm": 1.220282793045044,
+      "learning_rate": 9.860592954896523e-05,
+      "loss": 0.7332831025123596,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8975188781014024,
+      "grad_norm": 1.2287596464157104,
+      "learning_rate": 9.8600166195166e-05,
+      "loss": 0.7337175607681274,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8983818770226537,
+      "grad_norm": 1.0701712369918823,
+      "learning_rate": 9.859439112165579e-05,
+      "loss": 0.7422571182250977,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8992448759439051,
+      "grad_norm": 1.1583118438720703,
+      "learning_rate": 9.858860432982721e-05,
+      "loss": 0.7010003328323364,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9001078748651564,
+      "grad_norm": 0.9644797444343567,
+      "learning_rate": 9.858280582107575e-05,
+      "loss": 0.6211312413215637,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9009708737864077,
+      "grad_norm": 1.2448554039001465,
+      "learning_rate": 9.857699559679968e-05,
+      "loss": 0.7299860715866089,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9018338727076591,
+      "grad_norm": 1.2735580205917358,
+      "learning_rate": 9.85711736584001e-05,
+      "loss": 0.7461655735969543,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9026968716289104,
+      "grad_norm": 1.36841881275177,
+      "learning_rate": 9.856534000728093e-05,
+      "loss": 0.7218145132064819,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9035598705501618,
+      "grad_norm": 1.1471478939056396,
+      "learning_rate": 9.855949464484898e-05,
+      "loss": 0.7045709490776062,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9044228694714131,
+      "grad_norm": 1.058254361152649,
+      "learning_rate": 9.85536375725138e-05,
+      "loss": 0.7189744710922241,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9052858683926646,
+      "grad_norm": 1.4070709943771362,
+      "learning_rate": 9.854776879168779e-05,
+      "loss": 0.740775465965271,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 1.2182649374008179,
+      "learning_rate": 9.854188830378622e-05,
+      "loss": 0.8191262483596802,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "eval_loss": 0.7449111342430115,
+      "eval_runtime": 832.5135,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9070118662351672,
+      "grad_norm": 1.139978289604187,
+      "learning_rate": 9.853599611022708e-05,
+      "loss": 0.7295873761177063,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9078748651564186,
+      "grad_norm": 1.1802685260772705,
+      "learning_rate": 9.853009221243127e-05,
+      "loss": 0.7141753435134888,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9087378640776699,
+      "grad_norm": 1.076759934425354,
+      "learning_rate": 9.852417661182253e-05,
+      "loss": 0.7557607293128967,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9096008629989213,
+      "grad_norm": 1.1000943183898926,
+      "learning_rate": 9.851824930982733e-05,
+      "loss": 0.7025325298309326,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9104638619201726,
+      "grad_norm": 1.0235389471054077,
+      "learning_rate": 9.851231030787504e-05,
+      "loss": 0.6967638731002808,
+      "step": 2110
+    },
+    {
+      "epoch": 0.911326860841424,
+      "grad_norm": 1.2311428785324097,
+      "learning_rate": 9.850635960739782e-05,
+      "loss": 0.7890311479568481,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9121898597626753,
+      "grad_norm": 1.173770785331726,
+      "learning_rate": 9.850039720983064e-05,
+      "loss": 0.6824729442596436,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9130528586839266,
+      "grad_norm": 1.2165251970291138,
+      "learning_rate": 9.849442311661131e-05,
+      "loss": 0.7157722115516663,
+      "step": 2116
+    },
+    {
+      "epoch": 0.913915857605178,
+      "grad_norm": 1.251520037651062,
+      "learning_rate": 9.848843732918047e-05,
+      "loss": 0.7164111733436584,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9147788565264293,
+      "grad_norm": 1.1330616474151611,
+      "learning_rate": 9.848243984898153e-05,
+      "loss": 0.6830796003341675,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9156418554476807,
+      "grad_norm": 1.1052451133728027,
+      "learning_rate": 9.847643067746081e-05,
+      "loss": 0.7149171829223633,
+      "step": 2122
+    },
+    {
+      "epoch": 0.916504854368932,
+      "grad_norm": 1.0389703512191772,
+      "learning_rate": 9.847040981606735e-05,
+      "loss": 0.7213287949562073,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9173678532901833,
+      "grad_norm": 1.151895523071289,
+      "learning_rate": 9.846437726625308e-05,
+      "loss": 0.703312873840332,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9182308522114347,
+      "grad_norm": 1.1624029874801636,
+      "learning_rate": 9.845833302947269e-05,
+      "loss": 0.8059951066970825,
+      "step": 2128
+    },
+    {
+      "epoch": 0.919093851132686,
+      "grad_norm": 1.2247511148452759,
+      "learning_rate": 9.845227710718377e-05,
+      "loss": 0.735476553440094,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9199568500539375,
+      "grad_norm": 1.1649967432022095,
+      "learning_rate": 9.844620950084661e-05,
+      "loss": 0.8130571842193604,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9208198489751888,
+      "grad_norm": 1.1915717124938965,
+      "learning_rate": 9.844013021192447e-05,
+      "loss": 0.6908681392669678,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9216828478964402,
+      "grad_norm": 1.211199164390564,
+      "learning_rate": 9.843403924188328e-05,
+      "loss": 0.652173638343811,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9225458468176915,
+      "grad_norm": 1.267242193222046,
+      "learning_rate": 9.842793659219186e-05,
+      "loss": 0.6672233939170837,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9234088457389428,
+      "grad_norm": 1.7038761377334595,
+      "learning_rate": 9.842182226432186e-05,
+      "loss": 0.7274143695831299,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9242718446601942,
+      "grad_norm": 1.151426076889038,
+      "learning_rate": 9.841569625974769e-05,
+      "loss": 0.707998514175415,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9251348435814455,
+      "grad_norm": 1.0414749383926392,
+      "learning_rate": 9.840955857994663e-05,
+      "loss": 0.6639942526817322,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9259978425026969,
+      "grad_norm": 1.1045551300048828,
+      "learning_rate": 9.840340922639875e-05,
+      "loss": 0.7096070647239685,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9268608414239482,
+      "grad_norm": 1.1193996667861938,
+      "learning_rate": 9.839724820058695e-05,
+      "loss": 0.71246737241745,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9277238403451996,
+      "grad_norm": 1.2609117031097412,
+      "learning_rate": 9.839107550399692e-05,
+      "loss": 0.7351052761077881,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9285868392664509,
+      "grad_norm": 1.1127749681472778,
+      "learning_rate": 9.838489113811717e-05,
+      "loss": 0.7125008702278137,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9294498381877022,
+      "grad_norm": 1.4771665334701538,
+      "learning_rate": 9.837869510443905e-05,
+      "loss": 0.7891085743904114,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9303128371089536,
+      "grad_norm": 1.0701470375061035,
+      "learning_rate": 9.83724874044567e-05,
+      "loss": 0.7370918393135071,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9311758360302049,
+      "grad_norm": 1.2045820951461792,
+      "learning_rate": 9.836626803966708e-05,
+      "loss": 0.7543193697929382,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.1944687366485596,
+      "learning_rate": 9.836003701156996e-05,
+      "loss": 0.6903626322746277,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9329018338727076,
+      "grad_norm": 1.1035394668579102,
+      "learning_rate": 9.83537943216679e-05,
+      "loss": 0.6923050880432129,
+      "step": 2162
+    },
+    {
+      "epoch": 0.933764832793959,
+      "grad_norm": 1.3766103982925415,
+      "learning_rate": 9.834753997146633e-05,
+      "loss": 0.8099993467330933,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9346278317152104,
+      "grad_norm": 1.0874576568603516,
+      "learning_rate": 9.834127396247346e-05,
+      "loss": 0.7896219491958618,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9354908306364617,
+      "grad_norm": 1.1633459329605103,
+      "learning_rate": 9.833499629620028e-05,
+      "loss": 0.7659367918968201,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9363538295577131,
+      "grad_norm": 1.405423641204834,
+      "learning_rate": 9.832870697416066e-05,
+      "loss": 0.7474715709686279,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9372168284789644,
+      "grad_norm": 1.1794270277023315,
+      "learning_rate": 9.83224059978712e-05,
+      "loss": 0.7018750905990601,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9380798274002158,
+      "grad_norm": 1.082466959953308,
+      "learning_rate": 9.831609336885139e-05,
+      "loss": 0.6477128267288208,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9389428263214671,
+      "grad_norm": 0.9830297827720642,
+      "learning_rate": 9.830976908862346e-05,
+      "loss": 0.6516029238700867,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9398058252427185,
+      "grad_norm": 1.112289309501648,
+      "learning_rate": 9.83034331587125e-05,
+      "loss": 0.6793080568313599,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9406688241639698,
+      "grad_norm": 1.1212154626846313,
+      "learning_rate": 9.829708558064638e-05,
+      "loss": 0.7220560312271118,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9415318230852211,
+      "grad_norm": 1.246471643447876,
+      "learning_rate": 9.829072635595582e-05,
+      "loss": 0.7621577978134155,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9423948220064725,
+      "grad_norm": 1.1732288599014282,
+      "learning_rate": 9.828435548617428e-05,
+      "loss": 0.6988050937652588,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9432578209277238,
+      "grad_norm": 1.1767359972000122,
+      "learning_rate": 9.827797297283808e-05,
+      "loss": 0.7909345626831055,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9441208198489752,
+      "grad_norm": 1.007215976715088,
+      "learning_rate": 9.827157881748633e-05,
+      "loss": 0.6651493310928345,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9449838187702265,
+      "grad_norm": 1.184118628501892,
+      "learning_rate": 9.826517302166096e-05,
+      "loss": 0.7742198705673218,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9458468176914779,
+      "grad_norm": 1.1227030754089355,
+      "learning_rate": 9.82587555869067e-05,
+      "loss": 0.7614272236824036,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9467098166127292,
+      "grad_norm": 1.1030738353729248,
+      "learning_rate": 9.825232651477109e-05,
+      "loss": 0.7302108407020569,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9475728155339805,
+      "grad_norm": 1.5209563970565796,
+      "learning_rate": 9.824588580680445e-05,
+      "loss": 0.7408936023712158,
+      "step": 2196
+    },
+    {
+      "epoch": 0.948435814455232,
+      "grad_norm": 1.1853431463241577,
+      "learning_rate": 9.823943346455995e-05,
+      "loss": 0.8050196766853333,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "grad_norm": 1.285551905632019,
+      "learning_rate": 9.823296948959354e-05,
+      "loss": 0.726731538772583,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "eval_loss": 0.7400083541870117,
+      "eval_runtime": 838.7633,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9501618122977347,
+      "grad_norm": 1.2159804105758667,
+      "learning_rate": 9.822649388346396e-05,
+      "loss": 0.7061861753463745,
+      "step": 2202
+    },
+    {
+      "epoch": 0.951024811218986,
+      "grad_norm": 1.1189815998077393,
+      "learning_rate": 9.822000664773278e-05,
+      "loss": 0.824293851852417,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9518878101402374,
+      "grad_norm": 0.9968506097793579,
+      "learning_rate": 9.821350778396439e-05,
+      "loss": 0.7230689525604248,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9527508090614887,
+      "grad_norm": 1.101393222808838,
+      "learning_rate": 9.820699729372593e-05,
+      "loss": 0.6633790135383606,
+      "step": 2208
+    },
+    {
+      "epoch": 0.95361380798274,
+      "grad_norm": 1.084307312965393,
+      "learning_rate": 9.820047517858741e-05,
+      "loss": 0.6548227071762085,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9544768069039914,
+      "grad_norm": 1.1291496753692627,
+      "learning_rate": 9.819394144012158e-05,
+      "loss": 0.6434295177459717,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9553398058252427,
+      "grad_norm": 2.303483724594116,
+      "learning_rate": 9.818739607990404e-05,
+      "loss": 0.8122013211250305,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9562028047464941,
+      "grad_norm": 1.244541883468628,
+      "learning_rate": 9.818083909951315e-05,
+      "loss": 0.6120955348014832,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9570658036677454,
+      "grad_norm": 1.1051502227783203,
+      "learning_rate": 9.817427050053012e-05,
+      "loss": 0.7569628953933716,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9579288025889967,
+      "grad_norm": 1.3161911964416504,
+      "learning_rate": 9.816769028453893e-05,
+      "loss": 0.7588980197906494,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9587918015102481,
+      "grad_norm": 1.2526928186416626,
+      "learning_rate": 9.816109845312636e-05,
+      "loss": 0.743370771408081,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9596548004314994,
+      "grad_norm": 1.196215033531189,
+      "learning_rate": 9.815449500788202e-05,
+      "loss": 0.80516517162323,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9605177993527508,
+      "grad_norm": 1.3204832077026367,
+      "learning_rate": 9.81478799503983e-05,
+      "loss": 0.7331632971763611,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9613807982740021,
+      "grad_norm": 1.2442790269851685,
+      "learning_rate": 9.814125328227035e-05,
+      "loss": 0.6810339093208313,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9622437971952535,
+      "grad_norm": 1.1739158630371094,
+      "learning_rate": 9.813461500509623e-05,
+      "loss": 0.7269608378410339,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9631067961165048,
+      "grad_norm": 1.1016241312026978,
+      "learning_rate": 9.812796512047666e-05,
+      "loss": 0.7159227132797241,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9639697950377561,
+      "grad_norm": 1.1654549837112427,
+      "learning_rate": 9.812130363001529e-05,
+      "loss": 0.7145758867263794,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9648327939590076,
+      "grad_norm": 1.0521938800811768,
+      "learning_rate": 9.811463053531846e-05,
+      "loss": 0.6873835325241089,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9656957928802589,
+      "grad_norm": 1.2285776138305664,
+      "learning_rate": 9.810794583799538e-05,
+      "loss": 0.6683027148246765,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9665587918015103,
+      "grad_norm": 0.9826368689537048,
+      "learning_rate": 9.810124953965805e-05,
+      "loss": 0.663141131401062,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9674217907227616,
+      "grad_norm": 1.0102336406707764,
+      "learning_rate": 9.809454164192123e-05,
+      "loss": 0.6688778400421143,
+      "step": 2242
+    },
+    {
+      "epoch": 0.968284789644013,
+      "grad_norm": 0.9781575202941895,
+      "learning_rate": 9.808782214640251e-05,
+      "loss": 0.671240508556366,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9691477885652643,
+      "grad_norm": 1.137385368347168,
+      "learning_rate": 9.808109105472225e-05,
+      "loss": 0.7673252820968628,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9700107874865156,
+      "grad_norm": 1.3359450101852417,
+      "learning_rate": 9.807434836850362e-05,
+      "loss": 0.8085554838180542,
+      "step": 2248
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 1.150694727897644,
+      "learning_rate": 9.806759408937262e-05,
+      "loss": 0.6750202178955078,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9717367853290183,
+      "grad_norm": 1.080316424369812,
+      "learning_rate": 9.8060828218958e-05,
+      "loss": 0.5928418040275574,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9725997842502697,
+      "grad_norm": 1.0489610433578491,
+      "learning_rate": 9.805405075889129e-05,
+      "loss": 0.6947471499443054,
+      "step": 2254
+    },
+    {
+      "epoch": 0.973462783171521,
+      "grad_norm": 1.199119210243225,
+      "learning_rate": 9.80472617108069e-05,
+      "loss": 0.7273924350738525,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9743257820927724,
+      "grad_norm": 0.9830896258354187,
+      "learning_rate": 9.804046107634193e-05,
+      "loss": 0.668613612651825,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9751887810140237,
+      "grad_norm": 1.1655877828598022,
+      "learning_rate": 9.803364885713633e-05,
+      "loss": 0.7255772352218628,
+      "step": 2260
+    },
+    {
+      "epoch": 0.976051779935275,
+      "grad_norm": 1.0289462804794312,
+      "learning_rate": 9.802682505483284e-05,
+      "loss": 0.6865858435630798,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9769147788565264,
+      "grad_norm": 1.1020126342773438,
+      "learning_rate": 9.801998967107702e-05,
+      "loss": 0.6680663228034973,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 1.0846027135849,
+      "learning_rate": 9.801314270751716e-05,
+      "loss": 0.7476468086242676,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9786407766990292,
+      "grad_norm": 1.1834278106689453,
+      "learning_rate": 9.800628416580437e-05,
+      "loss": 0.7173228859901428,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9795037756202805,
+      "grad_norm": 1.0082321166992188,
+      "learning_rate": 9.799941404759257e-05,
+      "loss": 0.6768618226051331,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9803667745415319,
+      "grad_norm": 1.1573102474212646,
+      "learning_rate": 9.799253235453846e-05,
+      "loss": 0.7877311110496521,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9812297734627832,
+      "grad_norm": 1.2030092477798462,
+      "learning_rate": 9.798563908830152e-05,
+      "loss": 0.6843966841697693,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9820927723840345,
+      "grad_norm": 1.080560564994812,
+      "learning_rate": 9.797873425054404e-05,
+      "loss": 0.6084349155426025,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9829557713052859,
+      "grad_norm": 1.1488033533096313,
+      "learning_rate": 9.797181784293107e-05,
+      "loss": 0.6784868836402893,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9838187702265372,
+      "grad_norm": 1.144773006439209,
+      "learning_rate": 9.796488986713049e-05,
+      "loss": 0.6111500263214111,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9846817691477886,
+      "grad_norm": 1.1356254816055298,
+      "learning_rate": 9.795795032481297e-05,
+      "loss": 0.7634603977203369,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9855447680690399,
+      "grad_norm": 1.139446496963501,
+      "learning_rate": 9.79509992176519e-05,
+      "loss": 0.6470903158187866,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9864077669902913,
+      "grad_norm": 1.1564133167266846,
+      "learning_rate": 9.794403654732354e-05,
+      "loss": 0.71474289894104,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9872707659115426,
+      "grad_norm": 1.0605655908584595,
+      "learning_rate": 9.793706231550689e-05,
+      "loss": 0.6413968801498413,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9881337648327939,
+      "grad_norm": 1.1660863161087036,
+      "learning_rate": 9.793007652388378e-05,
+      "loss": 0.6595705151557922,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9889967637540453,
+      "grad_norm": 1.1661914587020874,
+      "learning_rate": 9.792307917413877e-05,
+      "loss": 0.7144066095352173,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9898597626752966,
+      "grad_norm": 1.2250727415084839,
+      "learning_rate": 9.791607026795926e-05,
+      "loss": 0.7956018447875977,
+      "step": 2294
+    },
+    {
+      "epoch": 0.990722761596548,
+      "grad_norm": 0.9593650698661804,
+      "learning_rate": 9.790904980703541e-05,
+      "loss": 0.5786031484603882,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9915857605177993,
+      "grad_norm": 1.0697777271270752,
+      "learning_rate": 9.790201779306018e-05,
+      "loss": 0.6817460656166077,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "grad_norm": 1.2658408880233765,
+      "learning_rate": 9.789497422772931e-05,
+      "loss": 0.6845148801803589,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "eval_loss": 0.7360134720802307,
+      "eval_runtime": 825.8559,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2300
+    },
+    {
+      "epoch": 0.993311758360302,
+      "grad_norm": 1.0869544744491577,
+      "learning_rate": 9.788791911274131e-05,
+      "loss": 0.6689707040786743,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9941747572815534,
+      "grad_norm": 1.0049875974655151,
+      "learning_rate": 9.788085244979748e-05,
+      "loss": 0.6470839381217957,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9950377562028048,
+      "grad_norm": 1.4471046924591064,
+      "learning_rate": 9.787377424060196e-05,
+      "loss": 0.8018016815185547,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9959007551240561,
+      "grad_norm": 1.3750648498535156,
+      "learning_rate": 9.786668448686159e-05,
+      "loss": 0.8078322410583496,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9967637540453075,
+      "grad_norm": 1.098283290863037,
+      "learning_rate": 9.7859583190286e-05,
+      "loss": 0.6740908622741699,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9976267529665588,
+      "grad_norm": 1.0312142372131348,
+      "learning_rate": 9.785247035258773e-05,
+      "loss": 0.6844386458396912,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9984897518878102,
+      "grad_norm": 1.4482319355010986,
+      "learning_rate": 9.784534597548193e-05,
+      "loss": 0.7547479271888733,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9993527508090615,
+      "grad_norm": 1.0897735357284546,
+      "learning_rate": 9.783821006068663e-05,
+      "loss": 0.7713987231254578,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7887805700302124,
+      "learning_rate": 9.783106260992264e-05,
+      "loss": 0.6868399381637573,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0008629989212514,
+      "grad_norm": 1.0021663904190063,
+      "learning_rate": 9.782390362491353e-05,
+      "loss": 0.696355938911438,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0017259978425026,
+      "grad_norm": 1.3650091886520386,
+      "learning_rate": 9.781673310738564e-05,
+      "loss": 0.7315902709960938,
+      "step": 2322
+    },
+    {
+      "epoch": 1.002588996763754,
+      "grad_norm": 1.2247875928878784,
+      "learning_rate": 9.780955105906813e-05,
+      "loss": 0.7667547464370728,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0034519956850054,
+      "grad_norm": 1.1917641162872314,
+      "learning_rate": 9.780235748169291e-05,
+      "loss": 0.7380582094192505,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0043149946062568,
+      "grad_norm": 1.146159052848816,
+      "learning_rate": 9.779515237699466e-05,
+      "loss": 0.6780733466148376,
+      "step": 2328
+    },
+    {
+      "epoch": 1.005177993527508,
+      "grad_norm": 1.4877620935440063,
+      "learning_rate": 9.778793574671089e-05,
+      "loss": 0.8116409182548523,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0060409924487594,
+      "grad_norm": 1.3488554954528809,
+      "learning_rate": 9.778070759258185e-05,
+      "loss": 0.65245521068573,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0069039913700109,
+      "grad_norm": 1.096686840057373,
+      "learning_rate": 9.777346791635056e-05,
+      "loss": 0.584334135055542,
+      "step": 2334
+    },
+    {
+      "epoch": 1.007766990291262,
+      "grad_norm": 1.063981056213379,
+      "learning_rate": 9.776621671976285e-05,
+      "loss": 0.6594714522361755,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0086299892125135,
+      "grad_norm": 1.2811797857284546,
+      "learning_rate": 9.775895400456729e-05,
+      "loss": 0.6963099837303162,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0094929881337649,
+      "grad_norm": 0.9491971731185913,
+      "learning_rate": 9.775167977251529e-05,
+      "loss": 0.6246375441551208,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0103559870550163,
+      "grad_norm": 1.198683500289917,
+      "learning_rate": 9.774439402536098e-05,
+      "loss": 0.6352161169052124,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0112189859762675,
+      "grad_norm": 1.2178813219070435,
+      "learning_rate": 9.773709676486129e-05,
+      "loss": 0.6884894967079163,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0120819848975189,
+      "grad_norm": 1.1905319690704346,
+      "learning_rate": 9.77297879927759e-05,
+      "loss": 0.5945048928260803,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0129449838187703,
+      "grad_norm": 1.210302710533142,
+      "learning_rate": 9.77224677108673e-05,
+      "loss": 0.6533923149108887,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0138079827400215,
+      "grad_norm": 1.15508234500885,
+      "learning_rate": 9.771513592090076e-05,
+      "loss": 0.6619405746459961,
+      "step": 2350
+    },
+    {
+      "epoch": 1.014670981661273,
+      "grad_norm": 1.0833762884140015,
+      "learning_rate": 9.77077926246443e-05,
+      "loss": 0.6581310033798218,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0155339805825243,
+      "grad_norm": 1.0330514907836914,
+      "learning_rate": 9.770043782386869e-05,
+      "loss": 0.6670945882797241,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0163969795037757,
+      "grad_norm": 1.1472195386886597,
+      "learning_rate": 9.769307152034755e-05,
+      "loss": 0.6801814436912537,
+      "step": 2356
+    },
+    {
+      "epoch": 1.017259978425027,
+      "grad_norm": 1.2204132080078125,
+      "learning_rate": 9.76856937158572e-05,
+      "loss": 0.7206465005874634,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0181229773462783,
+      "grad_norm": 1.2335962057113647,
+      "learning_rate": 9.76783044121768e-05,
+      "loss": 0.6884574294090271,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0189859762675297,
+      "grad_norm": 1.0858685970306396,
+      "learning_rate": 9.767090361108822e-05,
+      "loss": 0.655097484588623,
+      "step": 2362
+    },
+    {
+      "epoch": 1.019848975188781,
+      "grad_norm": 1.117907166481018,
+      "learning_rate": 9.766349131437614e-05,
+      "loss": 0.6478561758995056,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0207119741100323,
+      "grad_norm": 1.1995056867599487,
+      "learning_rate": 9.765606752382797e-05,
+      "loss": 0.6310951113700867,
+      "step": 2366
+    },
+    {
+      "epoch": 1.0215749730312838,
+      "grad_norm": 1.354819655418396,
+      "learning_rate": 9.764863224123397e-05,
+      "loss": 0.6373490691184998,
+      "step": 2368
+    },
+    {
+      "epoch": 1.0224379719525352,
+      "grad_norm": 1.2194819450378418,
+      "learning_rate": 9.764118546838712e-05,
+      "loss": 0.6595761179924011,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0233009708737864,
+      "grad_norm": 1.1485158205032349,
+      "learning_rate": 9.763372720708314e-05,
+      "loss": 0.6299480199813843,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0241639697950378,
+      "grad_norm": 1.1153662204742432,
+      "learning_rate": 9.76262574591206e-05,
+      "loss": 0.6788742542266846,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0250269687162892,
+      "grad_norm": 1.0690817832946777,
+      "learning_rate": 9.761877622630075e-05,
+      "loss": 0.559056282043457,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0258899676375404,
+      "grad_norm": 1.1896858215332031,
+      "learning_rate": 9.76112835104277e-05,
+      "loss": 0.7164555191993713,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0267529665587918,
+      "grad_norm": 1.1792134046554565,
+      "learning_rate": 9.760377931330824e-05,
+      "loss": 0.6791456937789917,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0276159654800432,
+      "grad_norm": 1.361763834953308,
+      "learning_rate": 9.759626363675202e-05,
+      "loss": 0.7421526312828064,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0284789644012946,
+      "grad_norm": 1.1960662603378296,
+      "learning_rate": 9.758873648257138e-05,
+      "loss": 0.7032517194747925,
+      "step": 2384
+    },
+    {
+      "epoch": 1.0293419633225458,
+      "grad_norm": 1.323255181312561,
+      "learning_rate": 9.758119785258146e-05,
+      "loss": 0.6953344345092773,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0302049622437972,
+      "grad_norm": 1.2343937158584595,
+      "learning_rate": 9.75736477486002e-05,
+      "loss": 0.7287829518318176,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0310679611650486,
+      "grad_norm": 1.0712062120437622,
+      "learning_rate": 9.756608617244823e-05,
+      "loss": 0.6812928915023804,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0319309600862998,
+      "grad_norm": 1.264726996421814,
+      "learning_rate": 9.755851312594902e-05,
+      "loss": 0.7173963785171509,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0327939590075512,
+      "grad_norm": 1.1563389301300049,
+      "learning_rate": 9.755092861092874e-05,
+      "loss": 0.6509836316108704,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0336569579288026,
+      "grad_norm": 1.326188087463379,
+      "learning_rate": 9.75433326292164e-05,
+      "loss": 0.718437135219574,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0345199568500538,
+      "grad_norm": 1.2263392210006714,
+      "learning_rate": 9.753572518264373e-05,
+      "loss": 0.664516031742096,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "grad_norm": 1.2874478101730347,
+      "learning_rate": 9.752810627304523e-05,
+      "loss": 0.6754546165466309,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "eval_loss": 0.7346783876419067,
+      "eval_runtime": 825.8356,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0362459546925566,
+      "grad_norm": 1.304518461227417,
+      "learning_rate": 9.752047590225814e-05,
+      "loss": 0.708167314529419,
+      "step": 2402
+    },
+    {
+      "epoch": 1.037108953613808,
+      "grad_norm": 1.21835196018219,
+      "learning_rate": 9.751283407212252e-05,
+      "loss": 0.7213018536567688,
+      "step": 2404
+    },
+    {
+      "epoch": 1.0379719525350593,
+      "grad_norm": 1.2947579622268677,
+      "learning_rate": 9.750518078448116e-05,
+      "loss": 0.6739231944084167,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 1.138368010520935,
+      "learning_rate": 9.749751604117959e-05,
+      "loss": 0.7671371102333069,
+      "step": 2408
+    },
+    {
+      "epoch": 1.039697950377562,
+      "grad_norm": 1.1255316734313965,
+      "learning_rate": 9.748983984406616e-05,
+      "loss": 0.6593353748321533,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0405609492988135,
+      "grad_norm": 1.2362794876098633,
+      "learning_rate": 9.748215219499193e-05,
+      "loss": 0.713056743144989,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0414239482200647,
+      "grad_norm": 1.2021304368972778,
+      "learning_rate": 9.747445309581076e-05,
+      "loss": 0.6989394426345825,
+      "step": 2414
+    },
+    {
+      "epoch": 1.042286947141316,
+      "grad_norm": 1.127341389656067,
+      "learning_rate": 9.746674254837925e-05,
+      "loss": 0.623035192489624,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0431499460625675,
+      "grad_norm": 1.1254255771636963,
+      "learning_rate": 9.745902055455672e-05,
+      "loss": 0.6490310430526733,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0440129449838187,
+      "grad_norm": 1.2142512798309326,
+      "learning_rate": 9.745128711620535e-05,
+      "loss": 0.6210505366325378,
+      "step": 2420
+    },
+    {
+      "epoch": 1.04487594390507,
+      "grad_norm": 1.1306709051132202,
+      "learning_rate": 9.744354223519001e-05,
+      "loss": 0.6243326663970947,
+      "step": 2422
+    },
+    {
+      "epoch": 1.0457389428263215,
+      "grad_norm": 1.202621340751648,
+      "learning_rate": 9.743578591337833e-05,
+      "loss": 0.6312575340270996,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0466019417475727,
+      "grad_norm": 1.0556137561798096,
+      "learning_rate": 9.742801815264072e-05,
+      "loss": 0.6190714836120605,
+      "step": 2426
+    },
+    {
+      "epoch": 1.0474649406688241,
+      "grad_norm": 1.5071814060211182,
+      "learning_rate": 9.742023895485033e-05,
+      "loss": 0.7496671080589294,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0483279395900755,
+      "grad_norm": 1.1448185443878174,
+      "learning_rate": 9.74124483218831e-05,
+      "loss": 0.6288847327232361,
+      "step": 2430
+    },
+    {
+      "epoch": 1.049190938511327,
+      "grad_norm": 1.0072964429855347,
+      "learning_rate": 9.74046462556177e-05,
+      "loss": 0.6938424706459045,
+      "step": 2432
+    },
+    {
+      "epoch": 1.0500539374325781,
+      "grad_norm": 1.198836326599121,
+      "learning_rate": 9.739683275793554e-05,
+      "loss": 0.7529639601707458,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0509169363538295,
+      "grad_norm": 1.2170380353927612,
+      "learning_rate": 9.738900783072083e-05,
+      "loss": 0.7193960547447205,
+      "step": 2436
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 1.0384336709976196,
+      "learning_rate": 9.738117147586051e-05,
+      "loss": 0.6488150954246521,
+      "step": 2438
+    },
+    {
+      "epoch": 1.0526429341963321,
+      "grad_norm": 1.1562492847442627,
+      "learning_rate": 9.737332369524429e-05,
+      "loss": 0.6672247052192688,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0535059331175836,
+      "grad_norm": 1.1553802490234375,
+      "learning_rate": 9.736546449076462e-05,
+      "loss": 0.651012122631073,
+      "step": 2442
+    },
+    {
+      "epoch": 1.054368932038835,
+      "grad_norm": 1.136838674545288,
+      "learning_rate": 9.735759386431671e-05,
+      "loss": 0.681246817111969,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0552319309600864,
+      "grad_norm": 1.177869200706482,
+      "learning_rate": 9.734971181779854e-05,
+      "loss": 0.7331019639968872,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0560949298813376,
+      "grad_norm": 1.2983415126800537,
+      "learning_rate": 9.73418183531108e-05,
+      "loss": 0.7316935658454895,
+      "step": 2448
+    },
+    {
+      "epoch": 1.056957928802589,
+      "grad_norm": 1.1189959049224854,
+      "learning_rate": 9.733391347215699e-05,
+      "loss": 0.6513608694076538,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0578209277238404,
+      "grad_norm": 1.3256937265396118,
+      "learning_rate": 9.732599717684334e-05,
+      "loss": 0.6469123363494873,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0586839266450916,
+      "grad_norm": 0.9766954779624939,
+      "learning_rate": 9.731806946907881e-05,
+      "loss": 0.5860153436660767,
+      "step": 2454
+    },
+    {
+      "epoch": 1.059546925566343,
+      "grad_norm": 1.1360441446304321,
+      "learning_rate": 9.731013035077514e-05,
+      "loss": 0.6783872842788696,
+      "step": 2456
+    },
+    {
+      "epoch": 1.0604099244875944,
+      "grad_norm": 1.3483080863952637,
+      "learning_rate": 9.73021798238468e-05,
+      "loss": 0.7894490361213684,
+      "step": 2458
+    },
+    {
+      "epoch": 1.0612729234088458,
+      "grad_norm": 1.1412159204483032,
+      "learning_rate": 9.729421789021105e-05,
+      "loss": 0.7098361849784851,
+      "step": 2460
+    },
+    {
+      "epoch": 1.062135922330097,
+      "grad_norm": 1.2670512199401855,
+      "learning_rate": 9.728624455178785e-05,
+      "loss": 0.7138097286224365,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0629989212513484,
+      "grad_norm": 1.2577205896377563,
+      "learning_rate": 9.727825981049994e-05,
+      "loss": 0.7136340141296387,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0638619201725998,
+      "grad_norm": 1.098997712135315,
+      "learning_rate": 9.727026366827283e-05,
+      "loss": 0.6273087859153748,
+      "step": 2466
+    },
+    {
+      "epoch": 1.064724919093851,
+      "grad_norm": 1.1550003290176392,
+      "learning_rate": 9.72622561270347e-05,
+      "loss": 0.7076449394226074,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0655879180151024,
+      "grad_norm": 1.0861129760742188,
+      "learning_rate": 9.725423718871659e-05,
+      "loss": 0.6856985092163086,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0664509169363539,
+      "grad_norm": 1.2742570638656616,
+      "learning_rate": 9.724620685525219e-05,
+      "loss": 0.7311446070671082,
+      "step": 2472
+    },
+    {
+      "epoch": 1.0673139158576053,
+      "grad_norm": 1.504339337348938,
+      "learning_rate": 9.7238165128578e-05,
+      "loss": 0.7728472948074341,
+      "step": 2474
+    },
+    {
+      "epoch": 1.0681769147788565,
+      "grad_norm": 1.2232030630111694,
+      "learning_rate": 9.723011201063323e-05,
+      "loss": 0.6259000301361084,
+      "step": 2476
+    },
+    {
+      "epoch": 1.0690399137001079,
+      "grad_norm": 1.3238953351974487,
+      "learning_rate": 9.722204750335989e-05,
+      "loss": 0.7132798433303833,
+      "step": 2478
+    },
+    {
+      "epoch": 1.0699029126213593,
+      "grad_norm": 1.2286484241485596,
+      "learning_rate": 9.721397160870265e-05,
+      "loss": 0.6067986488342285,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0707659115426105,
+      "grad_norm": 1.3101580142974854,
+      "learning_rate": 9.7205884328609e-05,
+      "loss": 0.6884648203849792,
+      "step": 2482
+    },
+    {
+      "epoch": 1.0716289104638619,
+      "grad_norm": 1.437833309173584,
+      "learning_rate": 9.719778566502915e-05,
+      "loss": 0.7164831757545471,
+      "step": 2484
+    },
+    {
+      "epoch": 1.0724919093851133,
+      "grad_norm": 1.0980390310287476,
+      "learning_rate": 9.718967561991606e-05,
+      "loss": 0.6326627731323242,
+      "step": 2486
+    },
+    {
+      "epoch": 1.0733549083063647,
+      "grad_norm": 0.9919570088386536,
+      "learning_rate": 9.718155419522543e-05,
+      "loss": 0.5712647438049316,
+      "step": 2488
+    },
+    {
+      "epoch": 1.074217907227616,
+      "grad_norm": 1.0773793458938599,
+      "learning_rate": 9.717342139291569e-05,
+      "loss": 0.6923331022262573,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0750809061488673,
+      "grad_norm": 1.1553318500518799,
+      "learning_rate": 9.716527721494805e-05,
+      "loss": 0.6737485527992249,
+      "step": 2492
+    },
+    {
+      "epoch": 1.0759439050701187,
+      "grad_norm": 1.112865924835205,
+      "learning_rate": 9.715712166328644e-05,
+      "loss": 0.6422063708305359,
+      "step": 2494
+    },
+    {
+      "epoch": 1.07680690399137,
+      "grad_norm": 1.0651512145996094,
+      "learning_rate": 9.71489547398975e-05,
+      "loss": 0.6399799585342407,
+      "step": 2496
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 1.1394895315170288,
+      "learning_rate": 9.71407764467507e-05,
+      "loss": 0.6218312382698059,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "grad_norm": 1.3323278427124023,
+      "learning_rate": 9.713258678581815e-05,
+      "loss": 0.65947425365448,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "eval_loss": 0.7294915914535522,
+      "eval_runtime": 836.4643,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0793959007551241,
+      "grad_norm": 1.3122533559799194,
+      "learning_rate": 9.712438575907479e-05,
+      "loss": 0.66681307554245,
+      "step": 2502
+    },
+    {
+      "epoch": 1.0802588996763753,
+      "grad_norm": 1.3631974458694458,
+      "learning_rate": 9.711617336849824e-05,
+      "loss": 0.6926553249359131,
+      "step": 2504
+    },
+    {
+      "epoch": 1.0811218985976268,
+      "grad_norm": 1.2383692264556885,
+      "learning_rate": 9.710794961606886e-05,
+      "loss": 0.6706596612930298,
+      "step": 2506
+    },
+    {
+      "epoch": 1.0819848975188782,
+      "grad_norm": 1.1660504341125488,
+      "learning_rate": 9.709971450376982e-05,
+      "loss": 0.6011359095573425,
+      "step": 2508
+    },
+    {
+      "epoch": 1.0828478964401294,
+      "grad_norm": 1.3021246194839478,
+      "learning_rate": 9.709146803358693e-05,
+      "loss": 0.689009428024292,
+      "step": 2510
+    },
+    {
+      "epoch": 1.0837108953613808,
+      "grad_norm": 1.0553252696990967,
+      "learning_rate": 9.708321020750883e-05,
+      "loss": 0.6216740012168884,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0845738942826322,
+      "grad_norm": 1.1862177848815918,
+      "learning_rate": 9.707494102752683e-05,
+      "loss": 0.6439616680145264,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0854368932038836,
+      "grad_norm": 1.1166558265686035,
+      "learning_rate": 9.706666049563504e-05,
+      "loss": 0.6825361251831055,
+      "step": 2516
+    },
+    {
+      "epoch": 1.0862998921251348,
+      "grad_norm": 1.128812551498413,
+      "learning_rate": 9.705836861383022e-05,
+      "loss": 0.6471308469772339,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0871628910463862,
+      "grad_norm": 1.324802279472351,
+      "learning_rate": 9.705006538411197e-05,
+      "loss": 0.6535511016845703,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0880258899676376,
+      "grad_norm": 1.3586591482162476,
+      "learning_rate": 9.704175080848255e-05,
+      "loss": 0.678022563457489,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 1.1334232091903687,
+      "learning_rate": 9.703342488894697e-05,
+      "loss": 0.5904073119163513,
+      "step": 2524
+    },
+    {
+      "epoch": 1.0897518878101402,
+      "grad_norm": 1.182215929031372,
+      "learning_rate": 9.702508762751303e-05,
+      "loss": 0.6072601079940796,
+      "step": 2526
+    },
+    {
+      "epoch": 1.0906148867313916,
+      "grad_norm": 1.2525769472122192,
+      "learning_rate": 9.701673902619119e-05,
+      "loss": 0.6752429604530334,
+      "step": 2528
+    },
+    {
+      "epoch": 1.091477885652643,
+      "grad_norm": 1.2811026573181152,
+      "learning_rate": 9.70083790869947e-05,
+      "loss": 0.7192611694335938,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0923408845738942,
+      "grad_norm": 1.1434061527252197,
+      "learning_rate": 9.700000781193949e-05,
+      "loss": 0.6991751194000244,
+      "step": 2532
+    },
+    {
+      "epoch": 1.0932038834951456,
+      "grad_norm": 1.26658296585083,
+      "learning_rate": 9.699162520304428e-05,
+      "loss": 0.7010498046875,
+      "step": 2534
+    },
+    {
+      "epoch": 1.094066882416397,
+      "grad_norm": 1.2083529233932495,
+      "learning_rate": 9.698323126233048e-05,
+      "loss": 0.6945615410804749,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0949298813376482,
+      "grad_norm": 1.2064114809036255,
+      "learning_rate": 9.697482599182227e-05,
+      "loss": 0.689496636390686,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0957928802588996,
+      "grad_norm": 1.17113196849823,
+      "learning_rate": 9.696640939354655e-05,
+      "loss": 0.6966884136199951,
+      "step": 2540
+    },
+    {
+      "epoch": 1.096655879180151,
+      "grad_norm": 1.1551356315612793,
+      "learning_rate": 9.695798146953291e-05,
+      "loss": 0.6524666547775269,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0975188781014025,
+      "grad_norm": 1.0966521501541138,
+      "learning_rate": 9.694954222181374e-05,
+      "loss": 0.5769726037979126,
+      "step": 2544
+    },
+    {
+      "epoch": 1.0983818770226537,
+      "grad_norm": 1.320669174194336,
+      "learning_rate": 9.694109165242411e-05,
+      "loss": 0.6986333727836609,
+      "step": 2546
+    },
+    {
+      "epoch": 1.099244875943905,
+      "grad_norm": 1.1678471565246582,
+      "learning_rate": 9.693262976340183e-05,
+      "loss": 0.7014455199241638,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1001078748651565,
+      "grad_norm": 1.1527155637741089,
+      "learning_rate": 9.692415655678745e-05,
+      "loss": 0.5786730647087097,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1009708737864077,
+      "grad_norm": 1.2511217594146729,
+      "learning_rate": 9.691567203462428e-05,
+      "loss": 0.755945086479187,
+      "step": 2552
+    },
+    {
+      "epoch": 1.101833872707659,
+      "grad_norm": 1.0910221338272095,
+      "learning_rate": 9.690717619895828e-05,
+      "loss": 0.6487610936164856,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1026968716289105,
+      "grad_norm": 1.3682197332382202,
+      "learning_rate": 9.68986690518382e-05,
+      "loss": 0.6346665024757385,
+      "step": 2556
+    },
+    {
+      "epoch": 1.103559870550162,
+      "grad_norm": 1.2161564826965332,
+      "learning_rate": 9.68901505953155e-05,
+      "loss": 0.647144615650177,
+      "step": 2558
+    },
+    {
+      "epoch": 1.104422869471413,
+      "grad_norm": 1.2559088468551636,
+      "learning_rate": 9.688162083144436e-05,
+      "loss": 0.6461755633354187,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1052858683926645,
+      "grad_norm": 1.1260225772857666,
+      "learning_rate": 9.68730797622817e-05,
+      "loss": 0.6888763904571533,
+      "step": 2562
+    },
+    {
+      "epoch": 1.106148867313916,
+      "grad_norm": 1.1499868631362915,
+      "learning_rate": 9.686452738988716e-05,
+      "loss": 0.6318327784538269,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1070118662351671,
+      "grad_norm": 1.1775025129318237,
+      "learning_rate": 9.68559637163231e-05,
+      "loss": 0.6538340449333191,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1078748651564185,
+      "grad_norm": 1.264656901359558,
+      "learning_rate": 9.684738874365465e-05,
+      "loss": 0.6706726551055908,
+      "step": 2568
+    },
+    {
+      "epoch": 1.10873786407767,
+      "grad_norm": 1.1130270957946777,
+      "learning_rate": 9.683880247394956e-05,
+      "loss": 0.7573289275169373,
+      "step": 2570
+    },
+    {
+      "epoch": 1.1096008629989211,
+      "grad_norm": 1.2781217098236084,
+      "learning_rate": 9.683020490927841e-05,
+      "loss": 0.6603306531906128,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1104638619201725,
+      "grad_norm": 1.251062035560608,
+      "learning_rate": 9.682159605171446e-05,
+      "loss": 0.6523779630661011,
+      "step": 2574
+    },
+    {
+      "epoch": 1.111326860841424,
+      "grad_norm": 1.2631903886795044,
+      "learning_rate": 9.68129759033337e-05,
+      "loss": 0.677550733089447,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1121898597626754,
+      "grad_norm": 1.0212602615356445,
+      "learning_rate": 9.680434446621483e-05,
+      "loss": 0.666540801525116,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1130528586839266,
+      "grad_norm": 1.221718430519104,
+      "learning_rate": 9.679570174243929e-05,
+      "loss": 0.7467086315155029,
+      "step": 2580
+    },
+    {
+      "epoch": 1.113915857605178,
+      "grad_norm": 1.1175146102905273,
+      "learning_rate": 9.678704773409124e-05,
+      "loss": 0.673492968082428,
+      "step": 2582
+    },
+    {
+      "epoch": 1.1147788565264294,
+      "grad_norm": 1.1274988651275635,
+      "learning_rate": 9.677838244325754e-05,
+      "loss": 0.5724847912788391,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1156418554476808,
+      "grad_norm": 1.297654390335083,
+      "learning_rate": 9.676970587202779e-05,
+      "loss": 0.634985625743866,
+      "step": 2586
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 1.1908061504364014,
+      "learning_rate": 9.67610180224943e-05,
+      "loss": 0.6813890933990479,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1173678532901834,
+      "grad_norm": 1.1787118911743164,
+      "learning_rate": 9.675231889675212e-05,
+      "loss": 0.6738101243972778,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1182308522114348,
+      "grad_norm": 1.1327152252197266,
+      "learning_rate": 9.674360849689899e-05,
+      "loss": 0.6662938594818115,
+      "step": 2592
+    },
+    {
+      "epoch": 1.119093851132686,
+      "grad_norm": 1.1246299743652344,
+      "learning_rate": 9.67348868250354e-05,
+      "loss": 0.677423357963562,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1199568500539374,
+      "grad_norm": 1.1411702632904053,
+      "learning_rate": 9.672615388326454e-05,
+      "loss": 0.6257020235061646,
+      "step": 2596
+    },
+    {
+      "epoch": 1.1208198489751888,
+      "grad_norm": 1.1607346534729004,
+      "learning_rate": 9.671740967369231e-05,
+      "loss": 0.7029470205307007,
+      "step": 2598
+    },
+    {
+      "epoch": 1.12168284789644,
+      "grad_norm": 1.047934651374817,
+      "learning_rate": 9.670865419842735e-05,
+      "loss": 0.6040990352630615,
+      "step": 2600
+    },
+    {
+      "epoch": 1.12168284789644,
+      "eval_loss": 0.7287447452545166,
+      "eval_runtime": 831.8138,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2600
+    },
+    {
+      "epoch": 1.1225458468176914,
+      "grad_norm": 1.0571783781051636,
+      "learning_rate": 9.669988745958098e-05,
+      "loss": 0.6872879266738892,
+      "step": 2602
+    },
+    {
+      "epoch": 1.1234088457389428,
+      "grad_norm": 1.0940403938293457,
+      "learning_rate": 9.669110945926727e-05,
+      "loss": 0.6173275113105774,
+      "step": 2604
+    },
+    {
+      "epoch": 1.1242718446601943,
+      "grad_norm": 1.1180222034454346,
+      "learning_rate": 9.6682320199603e-05,
+      "loss": 0.6142664551734924,
+      "step": 2606
+    },
+    {
+      "epoch": 1.1251348435814454,
+      "grad_norm": 1.251098871231079,
+      "learning_rate": 9.667351968270768e-05,
+      "loss": 0.7225713133811951,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1259978425026969,
+      "grad_norm": 1.2333059310913086,
+      "learning_rate": 9.666470791070347e-05,
+      "loss": 0.6830481290817261,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1268608414239483,
+      "grad_norm": 1.2904123067855835,
+      "learning_rate": 9.665588488571531e-05,
+      "loss": 0.6835238337516785,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1277238403451997,
+      "grad_norm": 1.4779443740844727,
+      "learning_rate": 9.664705060987085e-05,
+      "loss": 0.6993544697761536,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1285868392664509,
+      "grad_norm": 1.267021656036377,
+      "learning_rate": 9.663820508530042e-05,
+      "loss": 0.6819394826889038,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1294498381877023,
+      "grad_norm": 1.0755046606063843,
+      "learning_rate": 9.662934831413708e-05,
+      "loss": 0.6200308799743652,
+      "step": 2618
+    },
+    {
+      "epoch": 1.1303128371089537,
+      "grad_norm": 1.2043033838272095,
+      "learning_rate": 9.66204802985166e-05,
+      "loss": 0.7138658165931702,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1311758360302049,
+      "grad_norm": 1.2641128301620483,
+      "learning_rate": 9.661160104057744e-05,
+      "loss": 0.7238332033157349,
+      "step": 2622
+    },
+    {
+      "epoch": 1.1320388349514563,
+      "grad_norm": 1.0422990322113037,
+      "learning_rate": 9.660271054246085e-05,
+      "loss": 0.6401007175445557,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1329018338727077,
+      "grad_norm": 1.2298130989074707,
+      "learning_rate": 9.659380880631068e-05,
+      "loss": 0.6305879950523376,
+      "step": 2626
+    },
+    {
+      "epoch": 1.133764832793959,
+      "grad_norm": 1.254573106765747,
+      "learning_rate": 9.658489583427357e-05,
+      "loss": 0.5751872062683105,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1346278317152103,
+      "grad_norm": 1.15679132938385,
+      "learning_rate": 9.657597162849883e-05,
+      "loss": 0.6629595756530762,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1354908306364617,
+      "grad_norm": 1.193629264831543,
+      "learning_rate": 9.656703619113849e-05,
+      "loss": 0.6832258105278015,
+      "step": 2632
+    },
+    {
+      "epoch": 1.1363538295577131,
+      "grad_norm": 1.2330398559570312,
+      "learning_rate": 9.655808952434732e-05,
+      "loss": 0.6823931932449341,
+      "step": 2634
+    },
+    {
+      "epoch": 1.1372168284789643,
+      "grad_norm": 1.3926069736480713,
+      "learning_rate": 9.654913163028272e-05,
+      "loss": 0.6671395897865295,
+      "step": 2636
+    },
+    {
+      "epoch": 1.1380798274002157,
+      "grad_norm": 1.2995383739471436,
+      "learning_rate": 9.65401625111049e-05,
+      "loss": 0.662219226360321,
+      "step": 2638
+    },
+    {
+      "epoch": 1.1389428263214671,
+      "grad_norm": 1.1584415435791016,
+      "learning_rate": 9.653118216897667e-05,
+      "loss": 0.6419461965560913,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1398058252427186,
+      "grad_norm": 1.3584444522857666,
+      "learning_rate": 9.652219060606364e-05,
+      "loss": 0.6844428181648254,
+      "step": 2642
+    },
+    {
+      "epoch": 1.1406688241639698,
+      "grad_norm": 1.2619858980178833,
+      "learning_rate": 9.651318782453407e-05,
+      "loss": 0.6761999130249023,
+      "step": 2644
+    },
+    {
+      "epoch": 1.1415318230852212,
+      "grad_norm": 1.2202177047729492,
+      "learning_rate": 9.650417382655892e-05,
+      "loss": 0.7258123159408569,
+      "step": 2646
+    },
+    {
+      "epoch": 1.1423948220064726,
+      "grad_norm": 1.0225578546524048,
+      "learning_rate": 9.64951486143119e-05,
+      "loss": 0.5709917545318604,
+      "step": 2648
+    },
+    {
+      "epoch": 1.1432578209277238,
+      "grad_norm": 1.0961005687713623,
+      "learning_rate": 9.64861121899694e-05,
+      "loss": 0.7353709936141968,
+      "step": 2650
+    },
+    {
+      "epoch": 1.1441208198489752,
+      "grad_norm": 0.9393200874328613,
+      "learning_rate": 9.64770645557105e-05,
+      "loss": 0.7083833813667297,
+      "step": 2652
+    },
+    {
+      "epoch": 1.1449838187702266,
+      "grad_norm": 1.089733362197876,
+      "learning_rate": 9.646800571371701e-05,
+      "loss": 0.6405303478240967,
+      "step": 2654
+    },
+    {
+      "epoch": 1.1458468176914778,
+      "grad_norm": 1.1581398248672485,
+      "learning_rate": 9.645893566617341e-05,
+      "loss": 0.617733359336853,
+      "step": 2656
+    },
+    {
+      "epoch": 1.1467098166127292,
+      "grad_norm": 1.1336220502853394,
+      "learning_rate": 9.644985441526693e-05,
+      "loss": 0.6948232650756836,
+      "step": 2658
+    },
+    {
+      "epoch": 1.1475728155339806,
+      "grad_norm": 1.028591513633728,
+      "learning_rate": 9.644076196318743e-05,
+      "loss": 0.6392919421195984,
+      "step": 2660
+    },
+    {
+      "epoch": 1.148435814455232,
+      "grad_norm": 1.2088173627853394,
+      "learning_rate": 9.643165831212756e-05,
+      "loss": 0.6384931206703186,
+      "step": 2662
+    },
+    {
+      "epoch": 1.1492988133764832,
+      "grad_norm": 0.9905838966369629,
+      "learning_rate": 9.642254346428259e-05,
+      "loss": 0.5909387469291687,
+      "step": 2664
+    },
+    {
+      "epoch": 1.1501618122977346,
+      "grad_norm": 1.1670424938201904,
+      "learning_rate": 9.641341742185054e-05,
+      "loss": 0.700904130935669,
+      "step": 2666
+    },
+    {
+      "epoch": 1.151024811218986,
+      "grad_norm": 1.164746880531311,
+      "learning_rate": 9.640428018703211e-05,
+      "loss": 0.6036953330039978,
+      "step": 2668
+    },
+    {
+      "epoch": 1.1518878101402374,
+      "grad_norm": 1.00895357131958,
+      "learning_rate": 9.639513176203072e-05,
+      "loss": 0.6158549785614014,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1527508090614886,
+      "grad_norm": 1.2236207723617554,
+      "learning_rate": 9.638597214905244e-05,
+      "loss": 0.7280431389808655,
+      "step": 2672
+    },
+    {
+      "epoch": 1.15361380798274,
+      "grad_norm": 1.0930732488632202,
+      "learning_rate": 9.637680135030609e-05,
+      "loss": 0.6434410810470581,
+      "step": 2674
+    },
+    {
+      "epoch": 1.1544768069039915,
+      "grad_norm": 1.366325855255127,
+      "learning_rate": 9.636761936800315e-05,
+      "loss": 0.6771191954612732,
+      "step": 2676
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 2.249835968017578,
+      "learning_rate": 9.635842620435785e-05,
+      "loss": 0.6382432579994202,
+      "step": 2678
+    },
+    {
+      "epoch": 1.156202804746494,
+      "grad_norm": 1.274483561515808,
+      "learning_rate": 9.634922186158704e-05,
+      "loss": 0.6733657121658325,
+      "step": 2680
+    },
+    {
+      "epoch": 1.1570658036677455,
+      "grad_norm": 1.3715782165527344,
+      "learning_rate": 9.634000634191033e-05,
+      "loss": 0.7601200938224792,
+      "step": 2682
+    },
+    {
+      "epoch": 1.1579288025889967,
+      "grad_norm": 1.1312460899353027,
+      "learning_rate": 9.633077964754998e-05,
+      "loss": 0.5959098935127258,
+      "step": 2684
+    },
+    {
+      "epoch": 1.158791801510248,
+      "grad_norm": 1.063233494758606,
+      "learning_rate": 9.632154178073098e-05,
+      "loss": 0.615096926689148,
+      "step": 2686
+    },
+    {
+      "epoch": 1.1596548004314995,
+      "grad_norm": 1.074887752532959,
+      "learning_rate": 9.6312292743681e-05,
+      "loss": 0.6553784608840942,
+      "step": 2688
+    },
+    {
+      "epoch": 1.160517799352751,
+      "grad_norm": 1.17954683303833,
+      "learning_rate": 9.630303253863042e-05,
+      "loss": 0.6963269114494324,
+      "step": 2690
+    },
+    {
+      "epoch": 1.161380798274002,
+      "grad_norm": 1.2549158334732056,
+      "learning_rate": 9.629376116781226e-05,
+      "loss": 0.6633999347686768,
+      "step": 2692
+    },
+    {
+      "epoch": 1.1622437971952535,
+      "grad_norm": 1.237196683883667,
+      "learning_rate": 9.628447863346229e-05,
+      "loss": 0.6727597117424011,
+      "step": 2694
+    },
+    {
+      "epoch": 1.163106796116505,
+      "grad_norm": 1.4047741889953613,
+      "learning_rate": 9.627518493781895e-05,
+      "loss": 0.7018169164657593,
+      "step": 2696
+    },
+    {
+      "epoch": 1.1639697950377563,
+      "grad_norm": 1.15412175655365,
+      "learning_rate": 9.626588008312339e-05,
+      "loss": 0.6566510796546936,
+      "step": 2698
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "grad_norm": 1.194110631942749,
+      "learning_rate": 9.625656407161938e-05,
+      "loss": 0.7017931342124939,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "eval_loss": 0.72404944896698,
+      "eval_runtime": 824.2877,
+      "eval_samples_per_second": 2.499,
+      "eval_steps_per_second": 2.499,
+      "step": 2700
+    },
+    {
+      "epoch": 1.165695792880259,
+      "grad_norm": 1.1512954235076904,
+      "learning_rate": 9.62472369055535e-05,
+      "loss": 0.7058532238006592,
+      "step": 2702
+    },
+    {
+      "epoch": 1.1665587918015103,
+      "grad_norm": 1.4743460416793823,
+      "learning_rate": 9.623789858717492e-05,
+      "loss": 0.7458376288414001,
+      "step": 2704
+    },
+    {
+      "epoch": 1.1674217907227615,
+      "grad_norm": 1.0526450872421265,
+      "learning_rate": 9.622854911873551e-05,
+      "loss": 0.6526424288749695,
+      "step": 2706
+    },
+    {
+      "epoch": 1.168284789644013,
+      "grad_norm": 1.0408949851989746,
+      "learning_rate": 9.62191885024899e-05,
+      "loss": 0.6055451035499573,
+      "step": 2708
+    },
+    {
+      "epoch": 1.1691477885652644,
+      "grad_norm": 1.214052677154541,
+      "learning_rate": 9.620981674069534e-05,
+      "loss": 0.666462779045105,
+      "step": 2710
+    },
+    {
+      "epoch": 1.1700107874865155,
+      "grad_norm": 1.2526963949203491,
+      "learning_rate": 9.62004338356118e-05,
+      "loss": 0.7208677530288696,
+      "step": 2712
+    },
+    {
+      "epoch": 1.170873786407767,
+      "grad_norm": 1.2097395658493042,
+      "learning_rate": 9.61910397895019e-05,
+      "loss": 0.6269269585609436,
+      "step": 2714
+    },
+    {
+      "epoch": 1.1717367853290184,
+      "grad_norm": 1.035733699798584,
+      "learning_rate": 9.618163460463096e-05,
+      "loss": 0.6063299179077148,
+      "step": 2716
+    },
+    {
+      "epoch": 1.1725997842502698,
+      "grad_norm": 1.2881336212158203,
+      "learning_rate": 9.617221828326706e-05,
+      "loss": 0.7084641456604004,
+      "step": 2718
+    },
+    {
+      "epoch": 1.173462783171521,
+      "grad_norm": 1.2808257341384888,
+      "learning_rate": 9.616279082768085e-05,
+      "loss": 0.6515718102455139,
+      "step": 2720
+    },
+    {
+      "epoch": 1.1743257820927724,
+      "grad_norm": 1.1174249649047852,
+      "learning_rate": 9.615335224014574e-05,
+      "loss": 0.6745902299880981,
+      "step": 2722
+    },
+    {
+      "epoch": 1.1751887810140238,
+      "grad_norm": 1.1222290992736816,
+      "learning_rate": 9.61439025229378e-05,
+      "loss": 0.6824861168861389,
+      "step": 2724
+    },
+    {
+      "epoch": 1.176051779935275,
+      "grad_norm": 1.156187891960144,
+      "learning_rate": 9.613444167833579e-05,
+      "loss": 0.6020817160606384,
+      "step": 2726
+    },
+    {
+      "epoch": 1.1769147788565264,
+      "grad_norm": 1.2502888441085815,
+      "learning_rate": 9.612496970862113e-05,
+      "loss": 0.6430572271347046,
+      "step": 2728
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 1.6157052516937256,
+      "learning_rate": 9.611548661607797e-05,
+      "loss": 0.7195103168487549,
+      "step": 2730
+    },
+    {
+      "epoch": 1.1786407766990292,
+      "grad_norm": 1.3196234703063965,
+      "learning_rate": 9.610599240299308e-05,
+      "loss": 0.710418164730072,
+      "step": 2732
+    },
+    {
+      "epoch": 1.1795037756202804,
+      "grad_norm": 1.2117719650268555,
+      "learning_rate": 9.6096487071656e-05,
+      "loss": 0.650095522403717,
+      "step": 2734
+    },
+    {
+      "epoch": 1.1803667745415318,
+      "grad_norm": 2.2723631858825684,
+      "learning_rate": 9.608697062435884e-05,
+      "loss": 0.7150285243988037,
+      "step": 2736
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 1.0523152351379395,
+      "learning_rate": 9.607744306339647e-05,
+      "loss": 0.6271997690200806,
+      "step": 2738
+    },
+    {
+      "epoch": 1.1820927723840344,
+      "grad_norm": 1.2567129135131836,
+      "learning_rate": 9.606790439106646e-05,
+      "loss": 0.659799337387085,
+      "step": 2740
+    },
+    {
+      "epoch": 1.1829557713052858,
+      "grad_norm": 1.1448849439620972,
+      "learning_rate": 9.605835460966896e-05,
+      "loss": 0.7097326517105103,
+      "step": 2742
+    },
+    {
+      "epoch": 1.1838187702265373,
+      "grad_norm": 1.3698351383209229,
+      "learning_rate": 9.604879372150688e-05,
+      "loss": 0.6692788004875183,
+      "step": 2744
+    },
+    {
+      "epoch": 1.1846817691477884,
+      "grad_norm": 1.1424204111099243,
+      "learning_rate": 9.603922172888577e-05,
+      "loss": 0.6453474760055542,
+      "step": 2746
+    },
+    {
+      "epoch": 1.1855447680690399,
+      "grad_norm": 1.2701301574707031,
+      "learning_rate": 9.60296386341139e-05,
+      "loss": 0.6197408437728882,
+      "step": 2748
+    },
+    {
+      "epoch": 1.1864077669902913,
+      "grad_norm": 1.1171884536743164,
+      "learning_rate": 9.602004443950217e-05,
+      "loss": 0.6525035500526428,
+      "step": 2750
+    },
+    {
+      "epoch": 1.1872707659115427,
+      "grad_norm": 1.132633090019226,
+      "learning_rate": 9.60104391473642e-05,
+      "loss": 0.6459797620773315,
+      "step": 2752
+    },
+    {
+      "epoch": 1.1881337648327939,
+      "grad_norm": 1.1362011432647705,
+      "learning_rate": 9.600082276001623e-05,
+      "loss": 0.6175895929336548,
+      "step": 2754
+    },
+    {
+      "epoch": 1.1889967637540453,
+      "grad_norm": 1.1216920614242554,
+      "learning_rate": 9.599119527977722e-05,
+      "loss": 0.6782001852989197,
+      "step": 2756
+    },
+    {
+      "epoch": 1.1898597626752967,
+      "grad_norm": 1.024833083152771,
+      "learning_rate": 9.59815567089688e-05,
+      "loss": 0.6387497782707214,
+      "step": 2758
+    },
+    {
+      "epoch": 1.190722761596548,
+      "grad_norm": 1.3539615869522095,
+      "learning_rate": 9.597190704991527e-05,
+      "loss": 0.6974280476570129,
+      "step": 2760
+    },
+    {
+      "epoch": 1.1915857605177993,
+      "grad_norm": 1.002167820930481,
+      "learning_rate": 9.59622463049436e-05,
+      "loss": 0.6039770245552063,
+      "step": 2762
+    },
+    {
+      "epoch": 1.1924487594390507,
+      "grad_norm": 1.150636076927185,
+      "learning_rate": 9.595257447638343e-05,
+      "loss": 0.6543651819229126,
+      "step": 2764
+    },
+    {
+      "epoch": 1.1933117583603021,
+      "grad_norm": 1.413001537322998,
+      "learning_rate": 9.594289156656709e-05,
+      "loss": 0.7388567328453064,
+      "step": 2766
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 1.2041958570480347,
+      "learning_rate": 9.593319757782955e-05,
+      "loss": 0.7128161191940308,
+      "step": 2768
+    },
+    {
+      "epoch": 1.1950377562028047,
+      "grad_norm": 1.2151374816894531,
+      "learning_rate": 9.592349251250847e-05,
+      "loss": 0.6593868136405945,
+      "step": 2770
+    },
+    {
+      "epoch": 1.1959007551240561,
+      "grad_norm": 1.6225353479385376,
+      "learning_rate": 9.59137763729442e-05,
+      "loss": 0.6128022074699402,
+      "step": 2772
+    },
+    {
+      "epoch": 1.1967637540453073,
+      "grad_norm": 1.1904171705245972,
+      "learning_rate": 9.590404916147976e-05,
+      "loss": 0.6537567377090454,
+      "step": 2774
+    },
+    {
+      "epoch": 1.1976267529665587,
+      "grad_norm": 1.159506916999817,
+      "learning_rate": 9.589431088046079e-05,
+      "loss": 0.6325924396514893,
+      "step": 2776
+    },
+    {
+      "epoch": 1.1984897518878102,
+      "grad_norm": 1.1314560174942017,
+      "learning_rate": 9.588456153223565e-05,
+      "loss": 0.6438963413238525,
+      "step": 2778
+    },
+    {
+      "epoch": 1.1993527508090616,
+      "grad_norm": 1.0182478427886963,
+      "learning_rate": 9.587480111915532e-05,
+      "loss": 0.6703475117683411,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2002157497303128,
+      "grad_norm": 1.5506517887115479,
+      "learning_rate": 9.586502964357352e-05,
+      "loss": 0.7694190144538879,
+      "step": 2782
+    },
+    {
+      "epoch": 1.2010787486515642,
+      "grad_norm": 1.0769826173782349,
+      "learning_rate": 9.585524710784659e-05,
+      "loss": 0.6117724180221558,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2019417475728156,
+      "grad_norm": 1.207175850868225,
+      "learning_rate": 9.584545351433352e-05,
+      "loss": 0.6375927329063416,
+      "step": 2786
+    },
+    {
+      "epoch": 1.202804746494067,
+      "grad_norm": 1.1291100978851318,
+      "learning_rate": 9.583564886539602e-05,
+      "loss": 0.6526537537574768,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2036677454153182,
+      "grad_norm": 1.3076926469802856,
+      "learning_rate": 9.582583316339843e-05,
+      "loss": 0.7080113887786865,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2045307443365696,
+      "grad_norm": 1.166823148727417,
+      "learning_rate": 9.581600641070773e-05,
+      "loss": 0.6415606141090393,
+      "step": 2792
+    },
+    {
+      "epoch": 1.205393743257821,
+      "grad_norm": 1.1248292922973633,
+      "learning_rate": 9.580616860969364e-05,
+      "loss": 0.6538968086242676,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2062567421790722,
+      "grad_norm": 1.406731128692627,
+      "learning_rate": 9.579631976272849e-05,
+      "loss": 0.694507896900177,
+      "step": 2796
+    },
+    {
+      "epoch": 1.2071197411003236,
+      "grad_norm": 1.185895562171936,
+      "learning_rate": 9.578645987218728e-05,
+      "loss": 0.7064871788024902,
+      "step": 2798
+    },
+    {
+      "epoch": 1.207982740021575,
+      "grad_norm": 1.0860708951950073,
+      "learning_rate": 9.577658894044768e-05,
+      "loss": 0.5790154933929443,
+      "step": 2800
+    },
+    {
+      "epoch": 1.207982740021575,
+      "eval_loss": 0.7220398187637329,
+      "eval_runtime": 831.5812,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2088457389428262,
+      "grad_norm": 1.2761744260787964,
+      "learning_rate": 9.576670696989001e-05,
+      "loss": 0.6645335555076599,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2097087378640776,
+      "grad_norm": 1.3502917289733887,
+      "learning_rate": 9.575681396289728e-05,
+      "loss": 0.6320346593856812,
+      "step": 2804
+    },
+    {
+      "epoch": 1.210571736785329,
+      "grad_norm": 1.2638593912124634,
+      "learning_rate": 9.574690992185514e-05,
+      "loss": 0.6297362446784973,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2114347357065804,
+      "grad_norm": 1.0787774324417114,
+      "learning_rate": 9.573699484915189e-05,
+      "loss": 0.7569923996925354,
+      "step": 2808
+    },
+    {
+      "epoch": 1.2122977346278316,
+      "grad_norm": 1.097254991531372,
+      "learning_rate": 9.572706874717851e-05,
+      "loss": 0.6712964773178101,
+      "step": 2810
+    },
+    {
+      "epoch": 1.213160733549083,
+      "grad_norm": 1.0995726585388184,
+      "learning_rate": 9.571713161832865e-05,
+      "loss": 0.6775450706481934,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2140237324703345,
+      "grad_norm": 1.7386797666549683,
+      "learning_rate": 9.570718346499859e-05,
+      "loss": 0.6691418290138245,
+      "step": 2814
+    },
+    {
+      "epoch": 1.2148867313915859,
+      "grad_norm": 1.2281049489974976,
+      "learning_rate": 9.569722428958728e-05,
+      "loss": 0.6406403183937073,
+      "step": 2816
+    },
+    {
+      "epoch": 1.215749730312837,
+      "grad_norm": 1.1326305866241455,
+      "learning_rate": 9.568725409449635e-05,
+      "loss": 0.7008584141731262,
+      "step": 2818
+    },
+    {
+      "epoch": 1.2166127292340885,
+      "grad_norm": 1.3039381504058838,
+      "learning_rate": 9.567727288213005e-05,
+      "loss": 0.7438415288925171,
+      "step": 2820
+    },
+    {
+      "epoch": 1.2174757281553399,
+      "grad_norm": 1.219136357307434,
+      "learning_rate": 9.566728065489531e-05,
+      "loss": 0.6937777996063232,
+      "step": 2822
+    },
+    {
+      "epoch": 1.218338727076591,
+      "grad_norm": 1.321285605430603,
+      "learning_rate": 9.56572774152017e-05,
+      "loss": 0.7586468458175659,
+      "step": 2824
+    },
+    {
+      "epoch": 1.2192017259978425,
+      "grad_norm": 1.3538484573364258,
+      "learning_rate": 9.564726316546147e-05,
+      "loss": 0.6356340050697327,
+      "step": 2826
+    },
+    {
+      "epoch": 1.220064724919094,
+      "grad_norm": 1.1015816926956177,
+      "learning_rate": 9.563723790808952e-05,
+      "loss": 0.6304616928100586,
+      "step": 2828
+    },
+    {
+      "epoch": 1.220927723840345,
+      "grad_norm": 0.9658390283584595,
+      "learning_rate": 9.562720164550337e-05,
+      "loss": 0.6459048986434937,
+      "step": 2830
+    },
+    {
+      "epoch": 1.2217907227615965,
+      "grad_norm": 1.3981443643569946,
+      "learning_rate": 9.561715438012323e-05,
+      "loss": 0.7034182548522949,
+      "step": 2832
+    },
+    {
+      "epoch": 1.222653721682848,
+      "grad_norm": 1.1737093925476074,
+      "learning_rate": 9.560709611437195e-05,
+      "loss": 0.6629860997200012,
+      "step": 2834
+    },
+    {
+      "epoch": 1.2235167206040993,
+      "grad_norm": 1.2071303129196167,
+      "learning_rate": 9.559702685067506e-05,
+      "loss": 0.7166537642478943,
+      "step": 2836
+    },
+    {
+      "epoch": 1.2243797195253505,
+      "grad_norm": 1.2520930767059326,
+      "learning_rate": 9.558694659146068e-05,
+      "loss": 0.6711867451667786,
+      "step": 2838
+    },
+    {
+      "epoch": 1.225242718446602,
+      "grad_norm": 1.2304543256759644,
+      "learning_rate": 9.557685533915964e-05,
+      "loss": 0.7135099768638611,
+      "step": 2840
+    },
+    {
+      "epoch": 1.2261057173678533,
+      "grad_norm": 1.1509613990783691,
+      "learning_rate": 9.556675309620539e-05,
+      "loss": 0.6279205083847046,
+      "step": 2842
+    },
+    {
+      "epoch": 1.2269687162891048,
+      "grad_norm": 1.2024176120758057,
+      "learning_rate": 9.555663986503406e-05,
+      "loss": 0.6848570704460144,
+      "step": 2844
+    },
+    {
+      "epoch": 1.227831715210356,
+      "grad_norm": 1.360257863998413,
+      "learning_rate": 9.55465156480844e-05,
+      "loss": 0.7109206914901733,
+      "step": 2846
+    },
+    {
+      "epoch": 1.2286947141316074,
+      "grad_norm": 1.1556848287582397,
+      "learning_rate": 9.55363804477978e-05,
+      "loss": 0.5983237028121948,
+      "step": 2848
+    },
+    {
+      "epoch": 1.2295577130528588,
+      "grad_norm": 1.131439447402954,
+      "learning_rate": 9.552623426661836e-05,
+      "loss": 0.6387401223182678,
+      "step": 2850
+    },
+    {
+      "epoch": 1.23042071197411,
+      "grad_norm": 1.1854766607284546,
+      "learning_rate": 9.551607710699273e-05,
+      "loss": 0.6584593057632446,
+      "step": 2852
+    },
+    {
+      "epoch": 1.2312837108953614,
+      "grad_norm": 1.1811782121658325,
+      "learning_rate": 9.55059089713703e-05,
+      "loss": 0.6284405589103699,
+      "step": 2854
+    },
+    {
+      "epoch": 1.2321467098166128,
+      "grad_norm": 1.1038708686828613,
+      "learning_rate": 9.549572986220307e-05,
+      "loss": 0.6183417439460754,
+      "step": 2856
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 1.3645228147506714,
+      "learning_rate": 9.548553978194567e-05,
+      "loss": 0.7266764640808105,
+      "step": 2858
+    },
+    {
+      "epoch": 1.2338727076591154,
+      "grad_norm": 1.1896798610687256,
+      "learning_rate": 9.547533873305542e-05,
+      "loss": 0.6580738425254822,
+      "step": 2860
+    },
+    {
+      "epoch": 1.2347357065803668,
+      "grad_norm": 1.1351168155670166,
+      "learning_rate": 9.546512671799224e-05,
+      "loss": 0.669810950756073,
+      "step": 2862
+    },
+    {
+      "epoch": 1.2355987055016182,
+      "grad_norm": 1.1937936544418335,
+      "learning_rate": 9.545490373921869e-05,
+      "loss": 0.730329155921936,
+      "step": 2864
+    },
+    {
+      "epoch": 1.2364617044228694,
+      "grad_norm": 1.094347357749939,
+      "learning_rate": 9.544466979920002e-05,
+      "loss": 0.6867446303367615,
+      "step": 2866
+    },
+    {
+      "epoch": 1.2373247033441208,
+      "grad_norm": 1.4509600400924683,
+      "learning_rate": 9.54344249004041e-05,
+      "loss": 0.735468327999115,
+      "step": 2868
+    },
+    {
+      "epoch": 1.2381877022653722,
+      "grad_norm": 1.1195857524871826,
+      "learning_rate": 9.542416904530142e-05,
+      "loss": 0.6316167116165161,
+      "step": 2870
+    },
+    {
+      "epoch": 1.2390507011866236,
+      "grad_norm": 1.0227746963500977,
+      "learning_rate": 9.541390223636517e-05,
+      "loss": 0.6637873649597168,
+      "step": 2872
+    },
+    {
+      "epoch": 1.2399137001078748,
+      "grad_norm": 1.1871414184570312,
+      "learning_rate": 9.540362447607111e-05,
+      "loss": 0.6579877734184265,
+      "step": 2874
+    },
+    {
+      "epoch": 1.2407766990291262,
+      "grad_norm": 1.3265694379806519,
+      "learning_rate": 9.539333576689768e-05,
+      "loss": 0.7563158273696899,
+      "step": 2876
+    },
+    {
+      "epoch": 1.2416396979503777,
+      "grad_norm": 1.0074095726013184,
+      "learning_rate": 9.538303611132598e-05,
+      "loss": 0.6373291015625,
+      "step": 2878
+    },
+    {
+      "epoch": 1.2425026968716288,
+      "grad_norm": 1.421204924583435,
+      "learning_rate": 9.537272551183971e-05,
+      "loss": 0.6989303827285767,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2433656957928803,
+      "grad_norm": 1.256486415863037,
+      "learning_rate": 9.536240397092522e-05,
+      "loss": 0.7590261697769165,
+      "step": 2882
+    },
+    {
+      "epoch": 1.2442286947141317,
+      "grad_norm": 1.1513274908065796,
+      "learning_rate": 9.53520714910715e-05,
+      "loss": 0.6652005314826965,
+      "step": 2884
+    },
+    {
+      "epoch": 1.2450916936353829,
+      "grad_norm": 1.2129496335983276,
+      "learning_rate": 9.53417280747702e-05,
+      "loss": 0.6510938405990601,
+      "step": 2886
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 1.1393296718597412,
+      "learning_rate": 9.533137372451557e-05,
+      "loss": 0.6405928730964661,
+      "step": 2888
+    },
+    {
+      "epoch": 1.2468176914778857,
+      "grad_norm": 1.1467162370681763,
+      "learning_rate": 9.532100844280451e-05,
+      "loss": 0.6819307208061218,
+      "step": 2890
+    },
+    {
+      "epoch": 1.247680690399137,
+      "grad_norm": 1.1516613960266113,
+      "learning_rate": 9.531063223213659e-05,
+      "loss": 0.6333560347557068,
+      "step": 2892
+    },
+    {
+      "epoch": 1.2485436893203883,
+      "grad_norm": 1.463189959526062,
+      "learning_rate": 9.530024509501396e-05,
+      "loss": 0.7259839177131653,
+      "step": 2894
+    },
+    {
+      "epoch": 1.2494066882416397,
+      "grad_norm": 1.0450547933578491,
+      "learning_rate": 9.528984703394144e-05,
+      "loss": 0.661693811416626,
+      "step": 2896
+    },
+    {
+      "epoch": 1.250269687162891,
+      "grad_norm": 1.0525774955749512,
+      "learning_rate": 9.527943805142648e-05,
+      "loss": 0.5889303684234619,
+      "step": 2898
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "grad_norm": 1.130048394203186,
+      "learning_rate": 9.526901814997914e-05,
+      "loss": 0.6380610466003418,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "eval_loss": 0.7182289958000183,
+      "eval_runtime": 840.8156,
+      "eval_samples_per_second": 2.45,
+      "eval_steps_per_second": 2.45,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2519956850053937,
+      "grad_norm": 1.0194777250289917,
+      "learning_rate": 9.525858733211217e-05,
+      "loss": 0.6397356986999512,
+      "step": 2902
+    },
+    {
+      "epoch": 1.2528586839266451,
+      "grad_norm": 1.3902885913848877,
+      "learning_rate": 9.524814560034087e-05,
+      "loss": 0.7931577563285828,
+      "step": 2904
+    },
+    {
+      "epoch": 1.2537216828478965,
+      "grad_norm": 1.258169412612915,
+      "learning_rate": 9.523769295718325e-05,
+      "loss": 0.6997087597846985,
+      "step": 2906
+    },
+    {
+      "epoch": 1.2545846817691477,
+      "grad_norm": 1.3403711318969727,
+      "learning_rate": 9.522722940515991e-05,
+      "loss": 0.6946825385093689,
+      "step": 2908
+    },
+    {
+      "epoch": 1.2554476806903991,
+      "grad_norm": 1.1724950075149536,
+      "learning_rate": 9.521675494679409e-05,
+      "loss": 0.6309347748756409,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2563106796116505,
+      "grad_norm": 1.210274338722229,
+      "learning_rate": 9.520626958461163e-05,
+      "loss": 0.6453542113304138,
+      "step": 2912
+    },
+    {
+      "epoch": 1.2571736785329017,
+      "grad_norm": 1.233112096786499,
+      "learning_rate": 9.519577332114106e-05,
+      "loss": 0.6747867465019226,
+      "step": 2914
+    },
+    {
+      "epoch": 1.2580366774541532,
+      "grad_norm": 1.1785951852798462,
+      "learning_rate": 9.51852661589135e-05,
+      "loss": 0.6729235053062439,
+      "step": 2916
+    },
+    {
+      "epoch": 1.2588996763754046,
+      "grad_norm": 1.0781456232070923,
+      "learning_rate": 9.51747481004627e-05,
+      "loss": 0.6581047177314758,
+      "step": 2918
+    },
+    {
+      "epoch": 1.2597626752966558,
+      "grad_norm": 1.1122184991836548,
+      "learning_rate": 9.516421914832506e-05,
+      "loss": 0.6576398015022278,
+      "step": 2920
+    },
+    {
+      "epoch": 1.2606256742179072,
+      "grad_norm": 1.3378703594207764,
+      "learning_rate": 9.515367930503956e-05,
+      "loss": 0.7026275396347046,
+      "step": 2922
+    },
+    {
+      "epoch": 1.2614886731391586,
+      "grad_norm": 1.127657175064087,
+      "learning_rate": 9.514312857314785e-05,
+      "loss": 0.6111703515052795,
+      "step": 2924
+    },
+    {
+      "epoch": 1.26235167206041,
+      "grad_norm": 1.2058607339859009,
+      "learning_rate": 9.513256695519419e-05,
+      "loss": 0.6698760986328125,
+      "step": 2926
+    },
+    {
+      "epoch": 1.2632146709816614,
+      "grad_norm": 1.2206004858016968,
+      "learning_rate": 9.512199445372546e-05,
+      "loss": 0.6793603897094727,
+      "step": 2928
+    },
+    {
+      "epoch": 1.2640776699029126,
+      "grad_norm": 1.167311668395996,
+      "learning_rate": 9.511141107129119e-05,
+      "loss": 0.709406316280365,
+      "step": 2930
+    },
+    {
+      "epoch": 1.264940668824164,
+      "grad_norm": 1.3347517251968384,
+      "learning_rate": 9.51008168104435e-05,
+      "loss": 0.685570240020752,
+      "step": 2932
+    },
+    {
+      "epoch": 1.2658036677454154,
+      "grad_norm": 1.186468482017517,
+      "learning_rate": 9.509021167373714e-05,
+      "loss": 0.6775280833244324,
+      "step": 2934
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 1.3934271335601807,
+      "learning_rate": 9.507959566372953e-05,
+      "loss": 0.6587507128715515,
+      "step": 2936
+    },
+    {
+      "epoch": 1.267529665587918,
+      "grad_norm": 1.3766281604766846,
+      "learning_rate": 9.506896878298061e-05,
+      "loss": 0.692707896232605,
+      "step": 2938
+    },
+    {
+      "epoch": 1.2683926645091694,
+      "grad_norm": 1.034883975982666,
+      "learning_rate": 9.505833103405307e-05,
+      "loss": 0.6481655836105347,
+      "step": 2940
+    },
+    {
+      "epoch": 1.2692556634304206,
+      "grad_norm": 1.113044261932373,
+      "learning_rate": 9.504768241951212e-05,
+      "loss": 0.5867606997489929,
+      "step": 2942
+    },
+    {
+      "epoch": 1.270118662351672,
+      "grad_norm": 1.0769455432891846,
+      "learning_rate": 9.503702294192563e-05,
+      "loss": 0.6147398948669434,
+      "step": 2944
+    },
+    {
+      "epoch": 1.2709816612729234,
+      "grad_norm": 1.100616216659546,
+      "learning_rate": 9.502635260386408e-05,
+      "loss": 0.650299608707428,
+      "step": 2946
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 1.1982520818710327,
+      "learning_rate": 9.501567140790057e-05,
+      "loss": 0.6896426677703857,
+      "step": 2948
+    },
+    {
+      "epoch": 1.272707659115426,
+      "grad_norm": 1.276911735534668,
+      "learning_rate": 9.500497935661085e-05,
+      "loss": 0.6780133247375488,
+      "step": 2950
+    },
+    {
+      "epoch": 1.2735706580366775,
+      "grad_norm": 1.131345272064209,
+      "learning_rate": 9.499427645257323e-05,
+      "loss": 0.6906163692474365,
+      "step": 2952
+    },
+    {
+      "epoch": 1.2744336569579289,
+      "grad_norm": 1.125251293182373,
+      "learning_rate": 9.498356269836868e-05,
+      "loss": 0.6418634057044983,
+      "step": 2954
+    },
+    {
+      "epoch": 1.2752966558791803,
+      "grad_norm": 1.2508443593978882,
+      "learning_rate": 9.497283809658078e-05,
+      "loss": 0.690779983997345,
+      "step": 2956
+    },
+    {
+      "epoch": 1.2761596548004315,
+      "grad_norm": 0.9558717608451843,
+      "learning_rate": 9.49621026497957e-05,
+      "loss": 0.645685613155365,
+      "step": 2958
+    },
+    {
+      "epoch": 1.2770226537216829,
+      "grad_norm": 1.1781591176986694,
+      "learning_rate": 9.495135636060226e-05,
+      "loss": 0.6313972473144531,
+      "step": 2960
+    },
+    {
+      "epoch": 1.2778856526429343,
+      "grad_norm": 0.9713736772537231,
+      "learning_rate": 9.494059923159189e-05,
+      "loss": 0.5853794813156128,
+      "step": 2962
+    },
+    {
+      "epoch": 1.2787486515641855,
+      "grad_norm": 1.3793153762817383,
+      "learning_rate": 9.492983126535859e-05,
+      "loss": 0.6966362595558167,
+      "step": 2964
+    },
+    {
+      "epoch": 1.279611650485437,
+      "grad_norm": 1.250563383102417,
+      "learning_rate": 9.491905246449903e-05,
+      "loss": 0.7281587719917297,
+      "step": 2966
+    },
+    {
+      "epoch": 1.2804746494066883,
+      "grad_norm": 1.3268930912017822,
+      "learning_rate": 9.490826283161247e-05,
+      "loss": 0.7322974801063538,
+      "step": 2968
+    },
+    {
+      "epoch": 1.2813376483279395,
+      "grad_norm": 1.3876935243606567,
+      "learning_rate": 9.489746236930077e-05,
+      "loss": 0.7875081896781921,
+      "step": 2970
+    },
+    {
+      "epoch": 1.282200647249191,
+      "grad_norm": 1.02049720287323,
+      "learning_rate": 9.488665108016842e-05,
+      "loss": 0.601652979850769,
+      "step": 2972
+    },
+    {
+      "epoch": 1.2830636461704423,
+      "grad_norm": 1.0868756771087646,
+      "learning_rate": 9.487582896682252e-05,
+      "loss": 0.6255780458450317,
+      "step": 2974
+    },
+    {
+      "epoch": 1.2839266450916935,
+      "grad_norm": 1.00597083568573,
+      "learning_rate": 9.486499603187275e-05,
+      "loss": 0.5967418551445007,
+      "step": 2976
+    },
+    {
+      "epoch": 1.284789644012945,
+      "grad_norm": 1.3068969249725342,
+      "learning_rate": 9.485415227793145e-05,
+      "loss": 0.719573438167572,
+      "step": 2978
+    },
+    {
+      "epoch": 1.2856526429341963,
+      "grad_norm": 1.2478386163711548,
+      "learning_rate": 9.484329770761353e-05,
+      "loss": 0.665522038936615,
+      "step": 2980
+    },
+    {
+      "epoch": 1.2865156418554478,
+      "grad_norm": 1.1220918893814087,
+      "learning_rate": 9.483243232353652e-05,
+      "loss": 0.6798994541168213,
+      "step": 2982
+    },
+    {
+      "epoch": 1.287378640776699,
+      "grad_norm": 1.1369609832763672,
+      "learning_rate": 9.482155612832057e-05,
+      "loss": 0.672002911567688,
+      "step": 2984
+    },
+    {
+      "epoch": 1.2882416396979504,
+      "grad_norm": 1.2077715396881104,
+      "learning_rate": 9.48106691245884e-05,
+      "loss": 0.6755662560462952,
+      "step": 2986
+    },
+    {
+      "epoch": 1.2891046386192018,
+      "grad_norm": 1.3299074172973633,
+      "learning_rate": 9.479977131496539e-05,
+      "loss": 0.7504231333732605,
+      "step": 2988
+    },
+    {
+      "epoch": 1.2899676375404532,
+      "grad_norm": 1.08982253074646,
+      "learning_rate": 9.478886270207948e-05,
+      "loss": 0.6837405562400818,
+      "step": 2990
+    },
+    {
+      "epoch": 1.2908306364617044,
+      "grad_norm": 1.1015464067459106,
+      "learning_rate": 9.477794328856124e-05,
+      "loss": 0.6552648544311523,
+      "step": 2992
+    },
+    {
+      "epoch": 1.2916936353829558,
+      "grad_norm": 1.1793229579925537,
+      "learning_rate": 9.476701307704384e-05,
+      "loss": 0.6160483360290527,
+      "step": 2994
+    },
+    {
+      "epoch": 1.2925566343042072,
+      "grad_norm": 1.4781638383865356,
+      "learning_rate": 9.475607207016306e-05,
+      "loss": 0.7406170964241028,
+      "step": 2996
+    },
+    {
+      "epoch": 1.2934196332254584,
+      "grad_norm": 1.2391449213027954,
+      "learning_rate": 9.474512027055724e-05,
+      "loss": 0.5950883626937866,
+      "step": 2998
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "grad_norm": 1.166535496711731,
+      "learning_rate": 9.473415768086738e-05,
+      "loss": 0.6587089896202087,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "eval_loss": 0.7160727977752686,
+      "eval_runtime": 830.8186,
+      "eval_samples_per_second": 2.479,
+      "eval_steps_per_second": 2.479,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.1265100707984364e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3000/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-3000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/README.md b/sft_devstral_24B/checkpoints/checkpoint-3500/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-3500/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-3500/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..9a597db228a0e2d2f3a069688754e932d49a304a
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed4f0946d7554d2eb8dc25ff656bda1bc499858c3b08b2d09e9b3c13e41ad4b8
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-3500/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f8bc37898798737cbb593cef20b8a7692312a916
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:339367ba5676b49a084955aff2688d0f4f258b3db053f8a02ce2b85942acedcf
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-3500/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..08e3fa893e5464de9dc0bc8d38176f77f75b3dd7
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecad9507556708b06aaa4af823c254c9618f43d9383dcf7f8d3563bdd2adf264
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-3500/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..cfa39b38eac1e159bec88d08297dc4e2b1d20205
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5d033ea53af659a1c711c957f940dab790290f335af89412b27ed8e8c83f0bf
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-3500/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b6bb4dd3707b3aebb8ac8908b18c7f987b4f0ce
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/trainer_state.json
@@ -0,0 +1,12573 @@
+{
+  "best_global_step": 3500,
+  "best_metric": 0.7047909498214722,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-3500",
+  "epoch": 1.510032362459547,
+  "eval_steps": 100,
+  "global_step": 3500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4323624595469256,
+      "grad_norm": 1.5475730895996094,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.7443642616271973,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4332254584681769,
+      "grad_norm": 1.4621447324752808,
+      "learning_rate": 9.011680143755616e-05,
+      "loss": 0.8210510015487671,
+      "step": 1004
+    },
+    {
+      "epoch": 0.43408845738942825,
+      "grad_norm": 1.3651922941207886,
+      "learning_rate": 9.029649595687331e-05,
+      "loss": 0.8130766153335571,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4349514563106796,
+      "grad_norm": 1.3587357997894287,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.79145747423172,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 1.464652180671692,
+      "learning_rate": 9.065588499550764e-05,
+      "loss": 0.8286946415901184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4366774541531823,
+      "grad_norm": 1.3912625312805176,
+      "learning_rate": 9.083557951482479e-05,
+      "loss": 0.717319667339325,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4375404530744337,
+      "grad_norm": 1.2045491933822632,
+      "learning_rate": 9.101527403414196e-05,
+      "loss": 0.7294418215751648,
+      "step": 1014
+    },
+    {
+      "epoch": 0.438403451995685,
+      "grad_norm": 1.3449069261550903,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.7366700768470764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43926645091693634,
+      "grad_norm": 2.901384115219116,
+      "learning_rate": 9.137466307277629e-05,
+      "loss": 0.8551293611526489,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 1.6648688316345215,
+      "learning_rate": 9.155435759209344e-05,
+      "loss": 0.8944112658500671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44099244875943905,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 9.17340521114106e-05,
+      "loss": 0.7644214630126953,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4418554476806904,
+      "grad_norm": 1.4643231630325317,
+      "learning_rate": 9.191374663072777e-05,
+      "loss": 0.8644799590110779,
+      "step": 1024
+    },
+    {
+      "epoch": 0.44271844660194176,
+      "grad_norm": 1.4690274000167847,
+      "learning_rate": 9.209344115004492e-05,
+      "loss": 0.8281022906303406,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4435814455231931,
+      "grad_norm": 1.3321986198425293,
+      "learning_rate": 9.227313566936209e-05,
+      "loss": 0.7432141900062561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.8143638372421265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4453074433656958,
+      "grad_norm": 1.747295618057251,
+      "learning_rate": 9.263252470799642e-05,
+      "loss": 0.8102239370346069,
+      "step": 1032
+    },
+    {
+      "epoch": 0.44617044228694713,
+      "grad_norm": 1.3273073434829712,
+      "learning_rate": 9.281221922731357e-05,
+      "loss": 0.7768088579177856,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4470334412081985,
+      "grad_norm": 1.5019903182983398,
+      "learning_rate": 9.299191374663073e-05,
+      "loss": 0.7886938452720642,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44789644012944985,
+      "grad_norm": 1.6711527109146118,
+      "learning_rate": 9.31716082659479e-05,
+      "loss": 0.8152258396148682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 1.8504793643951416,
+      "learning_rate": 9.335130278526505e-05,
+      "loss": 0.8339349627494812,
+      "step": 1040
+    },
+    {
+      "epoch": 0.44962243797195256,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 9.353099730458221e-05,
+      "loss": 0.7798911929130554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.45048543689320386,
+      "grad_norm": 1.3361700773239136,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.7676728963851929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4513484358144552,
+      "grad_norm": 1.338934063911438,
+      "learning_rate": 9.389038634321654e-05,
+      "loss": 0.7729206085205078,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4522114347357066,
+      "grad_norm": 1.4347442388534546,
+      "learning_rate": 9.40700808625337e-05,
+      "loss": 0.7923640608787537,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 1.4451628923416138,
+      "learning_rate": 9.424977538185086e-05,
+      "loss": 0.8068996071815491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4539374325782093,
+      "grad_norm": 1.1904500722885132,
+      "learning_rate": 9.442946990116803e-05,
+      "loss": 0.7532910704612732,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45480043149946064,
+      "grad_norm": 1.4103976488113403,
+      "learning_rate": 9.460916442048518e-05,
+      "loss": 0.8236854672431946,
+      "step": 1054
+    },
+    {
+      "epoch": 0.455663430420712,
+      "grad_norm": 1.3762754201889038,
+      "learning_rate": 9.478885893980234e-05,
+      "loss": 0.7532716393470764,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4565264293419633,
+      "grad_norm": 1.461421012878418,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.8506031632423401,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 1.7207789421081543,
+      "learning_rate": 9.514824797843667e-05,
+      "loss": 0.7824664115905762,
+      "step": 1060
+    },
+    {
+      "epoch": 0.458252427184466,
+      "grad_norm": 1.5946974754333496,
+      "learning_rate": 9.532794249775382e-05,
+      "loss": 0.8674250245094299,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4591154261057174,
+      "grad_norm": 1.3552565574645996,
+      "learning_rate": 9.550763701707099e-05,
+      "loss": 0.7851845622062683,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45997842502696873,
+      "grad_norm": 1.414617657661438,
+      "learning_rate": 9.568733153638815e-05,
+      "loss": 0.8890700936317444,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4608414239482201,
+      "grad_norm": 1.3322749137878418,
+      "learning_rate": 9.58670260557053e-05,
+      "loss": 0.8226593732833862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 1.41420316696167,
+      "learning_rate": 9.604672057502247e-05,
+      "loss": 0.7663303017616272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.46256742179072274,
+      "grad_norm": 1.324967384338379,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.7122516632080078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4634304207119741,
+      "grad_norm": 1.2046154737472534,
+      "learning_rate": 9.640610961365679e-05,
+      "loss": 0.7177618145942688,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46429341963322546,
+      "grad_norm": 1.5221766233444214,
+      "learning_rate": 9.658580413297395e-05,
+      "loss": 0.7618930339813232,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4651564185544768,
+      "grad_norm": 1.3517179489135742,
+      "learning_rate": 9.67654986522911e-05,
+      "loss": 0.7483780384063721,
+      "step": 1078
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.747865080833435,
+      "learning_rate": 9.694519317160827e-05,
+      "loss": 0.8507670164108276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4668824163969795,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 9.712488769092543e-05,
+      "loss": 0.7893540859222412,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46774541531823083,
+      "grad_norm": 1.6290462017059326,
+      "learning_rate": 9.730458221024258e-05,
+      "loss": 0.7891272306442261,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4686084142394822,
+      "grad_norm": 1.6605802774429321,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.841018795967102,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46947141316073354,
+      "grad_norm": 1.388964295387268,
+      "learning_rate": 9.766397124887691e-05,
+      "loss": 0.8099003434181213,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 1.3901978731155396,
+      "learning_rate": 9.784366576819407e-05,
+      "loss": 0.7177658677101135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.47119741100323626,
+      "grad_norm": 1.3341423273086548,
+      "learning_rate": 9.802336028751123e-05,
+      "loss": 0.8285676836967468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4720604099244876,
+      "grad_norm": 1.5451914072036743,
+      "learning_rate": 9.82030548068284e-05,
+      "loss": 0.7582666873931885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47292340884573897,
+      "grad_norm": 1.5395640134811401,
+      "learning_rate": 9.838274932614556e-05,
+      "loss": 0.7293475270271301,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47378640776699027,
+      "grad_norm": 1.5005141496658325,
+      "learning_rate": 9.856244384546271e-05,
+      "loss": 0.8858493566513062,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 1.3562248945236206,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.7176282405853271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "eval_loss": 0.8152187466621399,
+      "eval_runtime": 835.0394,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.475512405609493,
+      "grad_norm": 1.5226716995239258,
+      "learning_rate": 9.892183288409704e-05,
+      "loss": 0.8048524856567383,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47637540453074434,
+      "grad_norm": 1.3544842004776,
+      "learning_rate": 9.91015274034142e-05,
+      "loss": 0.7954329252243042,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4772384034519957,
+      "grad_norm": 1.5169771909713745,
+      "learning_rate": 9.928122192273136e-05,
+      "loss": 0.804015040397644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47810140237324705,
+      "grad_norm": 1.4011081457138062,
+      "learning_rate": 9.946091644204852e-05,
+      "loss": 0.7652344703674316,
+      "step": 1108
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 1.4163414239883423,
+      "learning_rate": 9.964061096136569e-05,
+      "loss": 0.8501335382461548,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4798274002157497,
+      "grad_norm": 1.7406005859375,
+      "learning_rate": 9.982030548068284e-05,
+      "loss": 0.9024044871330261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.48069039913700107,
+      "grad_norm": 1.1490557193756104,
+      "learning_rate": 0.0001,
+      "loss": 0.7308287024497986,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4815533980582524,
+      "grad_norm": 1.2512775659561157,
+      "learning_rate": 9.99999939713627e-05,
+      "loss": 0.7273974418640137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4824163969795038,
+      "grad_norm": 1.3944242000579834,
+      "learning_rate": 9.999997588545227e-05,
+      "loss": 0.8071790337562561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 1.2209112644195557,
+      "learning_rate": 9.999994574227303e-05,
+      "loss": 0.6586678624153137,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4841423948220065,
+      "grad_norm": 1.7328095436096191,
+      "learning_rate": 9.999990354183228e-05,
+      "loss": 0.8529233336448669,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4850053937432578,
+      "grad_norm": 1.3309327363967896,
+      "learning_rate": 9.999984928414021e-05,
+      "loss": 0.7529556155204773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.48586839266450915,
+      "grad_norm": 1.22034752368927,
+      "learning_rate": 9.999978296920987e-05,
+      "loss": 0.6780388355255127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4867313915857605,
+      "grad_norm": 1.975162386894226,
+      "learning_rate": 9.999970459705727e-05,
+      "loss": 0.7941693067550659,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 1.4939875602722168,
+      "learning_rate": 9.999961416770132e-05,
+      "loss": 0.7895156145095825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4884573894282632,
+      "grad_norm": 1.4237034320831299,
+      "learning_rate": 9.999951168116379e-05,
+      "loss": 0.7662149667739868,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4893203883495146,
+      "grad_norm": 1.24979567527771,
+      "learning_rate": 9.999939713746942e-05,
+      "loss": 0.8079231381416321,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49018338727076594,
+      "grad_norm": 1.2252353429794312,
+      "learning_rate": 9.999927053664585e-05,
+      "loss": 0.7616434693336487,
+      "step": 1136
+    },
+    {
+      "epoch": 0.49104638619201724,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.999913187872357e-05,
+      "loss": 0.8576793074607849,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 1.2108005285263062,
+      "learning_rate": 9.999898116373605e-05,
+      "loss": 0.8060292601585388,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49277238403451995,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 9.99988183917196e-05,
+      "loss": 0.7532875537872314,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4936353829557713,
+      "grad_norm": 1.261606216430664,
+      "learning_rate": 9.99986435627135e-05,
+      "loss": 0.7241997718811035,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49449838187702266,
+      "grad_norm": 1.2678134441375732,
+      "learning_rate": 9.99984566767599e-05,
+      "loss": 0.78656005859375,
+      "step": 1146
+    },
+    {
+      "epoch": 0.495361380798274,
+      "grad_norm": 1.4185810089111328,
+      "learning_rate": 9.999825773390387e-05,
+      "loss": 0.7423107624053955,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 1.3218237161636353,
+      "learning_rate": 9.999804673419338e-05,
+      "loss": 0.8178372383117676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4970873786407767,
+      "grad_norm": 1.9698240756988525,
+      "learning_rate": 9.99978236776793e-05,
+      "loss": 0.8016347885131836,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49795037756202803,
+      "grad_norm": 1.2228590250015259,
+      "learning_rate": 9.999758856441544e-05,
+      "loss": 0.782421886920929,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4988133764832794,
+      "grad_norm": 1.2753938436508179,
+      "learning_rate": 9.999734139445849e-05,
+      "loss": 0.7235615849494934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.49967637540453075,
+      "grad_norm": 1.353155493736267,
+      "learning_rate": 9.999708216786805e-05,
+      "loss": 0.7686187028884888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 1.4242770671844482,
+      "learning_rate": 9.999681088470663e-05,
+      "loss": 0.7842718362808228,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5014023732470334,
+      "grad_norm": 1.501827359199524,
+      "learning_rate": 9.999652754503966e-05,
+      "loss": 0.8324156999588013,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5022653721682848,
+      "grad_norm": 1.5403931140899658,
+      "learning_rate": 9.999623214893545e-05,
+      "loss": 0.7797855138778687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5031283710895361,
+      "grad_norm": 1.4075721502304077,
+      "learning_rate": 9.999592469646525e-05,
+      "loss": 0.7667348384857178,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5039913700107875,
+      "grad_norm": 1.4411870241165161,
+      "learning_rate": 9.99956051877032e-05,
+      "loss": 0.8153020739555359,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.4197871685028076,
+      "learning_rate": 9.999527362272632e-05,
+      "loss": 0.8548861742019653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5057173678532901,
+      "grad_norm": 1.5030468702316284,
+      "learning_rate": 9.999493000161458e-05,
+      "loss": 0.7978097796440125,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5065803667745415,
+      "grad_norm": 1.483534574508667,
+      "learning_rate": 9.999457432445088e-05,
+      "loss": 0.7990087866783142,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5074433656957928,
+      "grad_norm": 1.5424920320510864,
+      "learning_rate": 9.999420659132093e-05,
+      "loss": 0.8072807788848877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5083063646170443,
+      "grad_norm": 1.129503846168518,
+      "learning_rate": 9.999382680231344e-05,
+      "loss": 0.7216097712516785,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 1.3208131790161133,
+      "learning_rate": 9.999343495752001e-05,
+      "loss": 0.8637414574623108,
+      "step": 1180
+    },
+    {
+      "epoch": 0.510032362459547,
+      "grad_norm": 1.5239129066467285,
+      "learning_rate": 9.999303105703508e-05,
+      "loss": 0.8561362028121948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5108953613807983,
+      "grad_norm": 1.3193504810333252,
+      "learning_rate": 9.99926151009561e-05,
+      "loss": 0.7514780759811401,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5117583603020496,
+      "grad_norm": 1.312649130821228,
+      "learning_rate": 9.999218708938336e-05,
+      "loss": 0.8634516596794128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.512621359223301,
+      "grad_norm": 1.2677429914474487,
+      "learning_rate": 9.999174702242006e-05,
+      "loss": 0.7493832111358643,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 1.3901816606521606,
+      "learning_rate": 9.999129490017234e-05,
+      "loss": 0.8981389999389648,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5143473570658037,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 9.999083072274921e-05,
+      "loss": 0.6771019697189331,
+      "step": 1192
+    },
+    {
+      "epoch": 0.515210355987055,
+      "grad_norm": 1.1299076080322266,
+      "learning_rate": 9.999035449026261e-05,
+      "loss": 0.789888858795166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5160733549083064,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 9.998986620282738e-05,
+      "loss": 0.6656325459480286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5169363538295577,
+      "grad_norm": 1.3438260555267334,
+      "learning_rate": 9.998936586056127e-05,
+      "loss": 0.795595645904541,
+      "step": 1198
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 1.45097017288208,
+      "learning_rate": 9.998885346358495e-05,
+      "loss": 0.7769836187362671,
+      "step": 1200
+    },
+    {
+      "epoch": 0.517799352750809,
+      "eval_loss": 0.8070361018180847,
+      "eval_runtime": 823.9296,
+      "eval_samples_per_second": 2.5,
+      "eval_steps_per_second": 2.5,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5186623516720604,
+      "grad_norm": 1.3791370391845703,
+      "learning_rate": 9.998832901202195e-05,
+      "loss": 0.7527318000793457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5195253505933117,
+      "grad_norm": 1.2574659585952759,
+      "learning_rate": 9.998779250599876e-05,
+      "loss": 0.8021673560142517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5203883495145631,
+      "grad_norm": 1.3305809497833252,
+      "learning_rate": 9.998724394564476e-05,
+      "loss": 0.776639461517334,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5212513484358144,
+      "grad_norm": 1.2330507040023804,
+      "learning_rate": 9.998668333109223e-05,
+      "loss": 0.7381181120872498,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 1.0482324361801147,
+      "learning_rate": 9.998611066247635e-05,
+      "loss": 0.6595721244812012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5229773462783172,
+      "grad_norm": 1.5456492900848389,
+      "learning_rate": 9.998552593993522e-05,
+      "loss": 0.8388534784317017,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5238403451995685,
+      "grad_norm": 1.3730167150497437,
+      "learning_rate": 9.998492916360986e-05,
+      "loss": 0.7593658566474915,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5247033441208199,
+      "grad_norm": 1.2668508291244507,
+      "learning_rate": 9.998432033364416e-05,
+      "loss": 0.7796505093574524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5255663430420712,
+      "grad_norm": 1.1175068616867065,
+      "learning_rate": 9.998369945018493e-05,
+      "loss": 0.7493594288825989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 1.121848464012146,
+      "learning_rate": 9.99830665133819e-05,
+      "loss": 0.7791322469711304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5272923408845739,
+      "grad_norm": 1.49788236618042,
+      "learning_rate": 9.998242152338772e-05,
+      "loss": 0.8299469947814941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5281553398058253,
+      "grad_norm": 1.227125644683838,
+      "learning_rate": 9.998176448035791e-05,
+      "loss": 0.7911439538002014,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5290183387270766,
+      "grad_norm": 1.176081895828247,
+      "learning_rate": 9.998109538445091e-05,
+      "loss": 0.6487553119659424,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5298813376483279,
+      "grad_norm": 1.169835090637207,
+      "learning_rate": 9.99804142358281e-05,
+      "loss": 0.7331781983375549,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 1.4085001945495605,
+      "learning_rate": 9.997972103465367e-05,
+      "loss": 0.8198121190071106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5316073354908306,
+      "grad_norm": 1.8634135723114014,
+      "learning_rate": 9.997901578109484e-05,
+      "loss": 0.7099791169166565,
+      "step": 1232
+    },
+    {
+      "epoch": 0.532470334412082,
+      "grad_norm": 1.4719511270523071,
+      "learning_rate": 9.997829847532166e-05,
+      "loss": 0.7416951656341553,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.2430769205093384,
+      "learning_rate": 9.99775691175071e-05,
+      "loss": 0.7583408355712891,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5341963322545846,
+      "grad_norm": 1.2502130270004272,
+      "learning_rate": 9.997682770782706e-05,
+      "loss": 0.7639391422271729,
+      "step": 1238
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 1.3577396869659424,
+      "learning_rate": 9.997607424646032e-05,
+      "loss": 0.8304449319839478,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5359223300970873,
+      "grad_norm": 1.0986218452453613,
+      "learning_rate": 9.997530873358854e-05,
+      "loss": 0.7331952452659607,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5367853290183388,
+      "grad_norm": 1.4607514142990112,
+      "learning_rate": 9.997453116939637e-05,
+      "loss": 0.8173373341560364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.53764832793959,
+      "grad_norm": 1.2897552251815796,
+      "learning_rate": 9.997374155407128e-05,
+      "loss": 0.7677372694015503,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5385113268608415,
+      "grad_norm": 1.4581117630004883,
+      "learning_rate": 9.997293988780372e-05,
+      "loss": 0.8524296283721924,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 1.2029366493225098,
+      "learning_rate": 9.997212617078698e-05,
+      "loss": 0.7920866012573242,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5402373247033441,
+      "grad_norm": 1.109483242034912,
+      "learning_rate": 9.997130040321727e-05,
+      "loss": 0.7889736294746399,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5411003236245955,
+      "grad_norm": 1.1930333375930786,
+      "learning_rate": 9.997046258529377e-05,
+      "loss": 0.6593763828277588,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5419633225458468,
+      "grad_norm": 1.3668162822723389,
+      "learning_rate": 9.996961271721845e-05,
+      "loss": 0.7768212556838989,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5428263214670982,
+      "grad_norm": 1.2901886701583862,
+      "learning_rate": 9.996875079919632e-05,
+      "loss": 0.7392467260360718,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.366720199584961,
+      "learning_rate": 9.996787683143518e-05,
+      "loss": 0.6961479187011719,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5445523193096009,
+      "grad_norm": 1.2925949096679688,
+      "learning_rate": 9.99669908141458e-05,
+      "loss": 0.801517128944397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5454153182308522,
+      "grad_norm": 1.1967244148254395,
+      "learning_rate": 9.996609274754182e-05,
+      "loss": 0.7287030816078186,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5462783171521035,
+      "grad_norm": 1.419996738433838,
+      "learning_rate": 9.996518263183985e-05,
+      "loss": 0.7791255116462708,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5471413160733549,
+      "grad_norm": 1.3544845581054688,
+      "learning_rate": 9.996426046725933e-05,
+      "loss": 0.6665297150611877,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 1.6272345781326294,
+      "learning_rate": 9.996332625402262e-05,
+      "loss": 0.7902728915214539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5488673139158576,
+      "grad_norm": 1.4264246225357056,
+      "learning_rate": 9.996237999235502e-05,
+      "loss": 0.8483795523643494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5497303128371089,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 9.996142168248473e-05,
+      "loss": 0.7432113289833069,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5505933117583603,
+      "grad_norm": 0.9958710074424744,
+      "learning_rate": 9.996045132464283e-05,
+      "loss": 0.6721545457839966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5514563106796116,
+      "grad_norm": 1.371719241142273,
+      "learning_rate": 9.995946891906331e-05,
+      "loss": 0.7770468592643738,
+      "step": 1278
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 1.1906969547271729,
+      "learning_rate": 9.995847446598307e-05,
+      "loss": 0.816495418548584,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5531823085221144,
+      "grad_norm": 1.262184977531433,
+      "learning_rate": 9.995746796564195e-05,
+      "loss": 0.7122345566749573,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5540453074433657,
+      "grad_norm": 1.454692006111145,
+      "learning_rate": 9.995644941828262e-05,
+      "loss": 0.8476479649543762,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5549083063646171,
+      "grad_norm": 1.44277024269104,
+      "learning_rate": 9.995541882415072e-05,
+      "loss": 0.7595248818397522,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5557713052858684,
+      "grad_norm": 1.151465654373169,
+      "learning_rate": 9.995437618349478e-05,
+      "loss": 0.7413201928138733,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 1.3160182237625122,
+      "learning_rate": 9.995332149656621e-05,
+      "loss": 0.7242997884750366,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5574973031283711,
+      "grad_norm": 1.2748576402664185,
+      "learning_rate": 9.995225476361936e-05,
+      "loss": 0.7779543995857239,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5583603020496224,
+      "grad_norm": 1.3639990091323853,
+      "learning_rate": 9.995117598491145e-05,
+      "loss": 0.8075780868530273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5592233009708738,
+      "grad_norm": 1.0247999429702759,
+      "learning_rate": 9.995008516070263e-05,
+      "loss": 0.6940524578094482,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5600862998921251,
+      "grad_norm": 1.3383198976516724,
+      "learning_rate": 9.994898229125597e-05,
+      "loss": 0.799100399017334,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 1.37660551071167,
+      "learning_rate": 9.994786737683738e-05,
+      "loss": 0.8038942217826843,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "eval_loss": 0.7962777018547058,
+      "eval_runtime": 827.8798,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5618122977346278,
+      "grad_norm": 1.5424575805664062,
+      "learning_rate": 9.994674041771574e-05,
+      "loss": 0.8553245067596436,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5626752966558792,
+      "grad_norm": 1.1047813892364502,
+      "learning_rate": 9.994560141416282e-05,
+      "loss": 0.6650274991989136,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5635382955771305,
+      "grad_norm": 1.4532393217086792,
+      "learning_rate": 9.994445036645328e-05,
+      "loss": 0.794471025466919,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5644012944983818,
+      "grad_norm": 1.2588927745819092,
+      "learning_rate": 9.994328727486468e-05,
+      "loss": 0.7241487503051758,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 1.4981781244277954,
+      "learning_rate": 9.994211213967748e-05,
+      "loss": 0.7781044244766235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5661272923408845,
+      "grad_norm": 1.426277995109558,
+      "learning_rate": 9.99409249611751e-05,
+      "loss": 0.7772427201271057,
+      "step": 1312
+    },
+    {
+      "epoch": 0.566990291262136,
+      "grad_norm": 1.1987197399139404,
+      "learning_rate": 9.993972573964379e-05,
+      "loss": 0.7443774342536926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5678532901833873,
+      "grad_norm": 1.1490470170974731,
+      "learning_rate": 9.993851447537275e-05,
+      "loss": 0.7203393578529358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5687162891046387,
+      "grad_norm": 1.3586851358413696,
+      "learning_rate": 9.993729116865406e-05,
+      "loss": 0.8038793206214905,
+      "step": 1318
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 1.4160653352737427,
+      "learning_rate": 9.993605581978274e-05,
+      "loss": 0.827308714389801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5704422869471413,
+      "grad_norm": 1.2725251913070679,
+      "learning_rate": 9.993480842905667e-05,
+      "loss": 0.8183196783065796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5713052858683927,
+      "grad_norm": 1.3789756298065186,
+      "learning_rate": 9.993354899677665e-05,
+      "loss": 0.8268020153045654,
+      "step": 1324
+    },
+    {
+      "epoch": 0.572168284789644,
+      "grad_norm": 1.2641910314559937,
+      "learning_rate": 9.993227752324639e-05,
+      "loss": 0.7434715628623962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5730312837108954,
+      "grad_norm": 1.2429786920547485,
+      "learning_rate": 9.99309940087725e-05,
+      "loss": 0.7636955976486206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 1.3281534910202026,
+      "learning_rate": 9.992969845366449e-05,
+      "loss": 0.794617772102356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.574757281553398,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 9.992839085823477e-05,
+      "loss": 0.7658806443214417,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5756202804746494,
+      "grad_norm": 1.4143259525299072,
+      "learning_rate": 9.99270712227987e-05,
+      "loss": 0.8483052849769592,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5764832793959007,
+      "grad_norm": 1.2387818098068237,
+      "learning_rate": 9.992573954767449e-05,
+      "loss": 0.8610434532165527,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5773462783171521,
+      "grad_norm": 1.3307154178619385,
+      "learning_rate": 9.992439583318322e-05,
+      "loss": 0.8021210432052612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 1.293459415435791,
+      "learning_rate": 9.992304007964896e-05,
+      "loss": 0.7536460757255554,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5790722761596548,
+      "grad_norm": 1.1508948802947998,
+      "learning_rate": 9.992167228739865e-05,
+      "loss": 0.7519120573997498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5799352750809061,
+      "grad_norm": 1.5906150341033936,
+      "learning_rate": 9.992029245676211e-05,
+      "loss": 0.784415066242218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5807982740021574,
+      "grad_norm": 1.1421746015548706,
+      "learning_rate": 9.99189005880721e-05,
+      "loss": 0.8217549324035645,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5816612729234089,
+      "grad_norm": 1.0767277479171753,
+      "learning_rate": 9.991749668166422e-05,
+      "loss": 0.7446820735931396,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 1.147594928741455,
+      "learning_rate": 9.991608073787706e-05,
+      "loss": 0.8426011800765991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5833872707659116,
+      "grad_norm": 1.6618242263793945,
+      "learning_rate": 9.991465275705206e-05,
+      "loss": 0.8185623288154602,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5842502696871629,
+      "grad_norm": 1.414926290512085,
+      "learning_rate": 9.991321273953355e-05,
+      "loss": 0.7521756887435913,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5851132686084143,
+      "grad_norm": 1.298245906829834,
+      "learning_rate": 9.991176068566882e-05,
+      "loss": 0.7517471313476562,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5859762675296656,
+      "grad_norm": 1.3243110179901123,
+      "learning_rate": 9.9910296595808e-05,
+      "loss": 0.749305248260498,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 1.1953799724578857,
+      "learning_rate": 9.990882047030413e-05,
+      "loss": 0.6883649230003357,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5877022653721683,
+      "grad_norm": 1.3934388160705566,
+      "learning_rate": 9.990733230951322e-05,
+      "loss": 0.7859632968902588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5885652642934196,
+      "grad_norm": 1.2212340831756592,
+      "learning_rate": 9.990583211379409e-05,
+      "loss": 0.7313099503517151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.589428263214671,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 9.990431988350853e-05,
+      "loss": 0.8007960319519043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5902912621359223,
+      "grad_norm": 1.1826236248016357,
+      "learning_rate": 9.990279561902118e-05,
+      "loss": 0.773026168346405,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 1.1181561946868896,
+      "learning_rate": 9.990125932069964e-05,
+      "loss": 0.7469213604927063,
+      "step": 1370
+    },
+    {
+      "epoch": 0.592017259978425,
+      "grad_norm": 1.3735432624816895,
+      "learning_rate": 9.989971098891439e-05,
+      "loss": 0.7788950204849243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5928802588996763,
+      "grad_norm": 1.1906198263168335,
+      "learning_rate": 9.989815062403877e-05,
+      "loss": 0.8117855191230774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5937432578209277,
+      "grad_norm": 1.326940894126892,
+      "learning_rate": 9.989657822644907e-05,
+      "loss": 0.7536540627479553,
+      "step": 1376
+    },
+    {
+      "epoch": 0.594606256742179,
+      "grad_norm": 1.4508434534072876,
+      "learning_rate": 9.989499379652447e-05,
+      "loss": 0.860320508480072,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 1.114499568939209,
+      "learning_rate": 9.989339733464704e-05,
+      "loss": 0.749508798122406,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5963322545846818,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 9.989178884120176e-05,
+      "loss": 0.7294815182685852,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5971952535059332,
+      "grad_norm": 1.291913628578186,
+      "learning_rate": 9.989016831657652e-05,
+      "loss": 0.7797688245773315,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5980582524271845,
+      "grad_norm": 1.1245452165603638,
+      "learning_rate": 9.988853576116208e-05,
+      "loss": 0.7824407815933228,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5989212513484358,
+      "grad_norm": 1.2384979724884033,
+      "learning_rate": 9.988689117535215e-05,
+      "loss": 0.714794933795929,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 1.1945903301239014,
+      "learning_rate": 9.988523455954332e-05,
+      "loss": 0.7072597742080688,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6006472491909385,
+      "grad_norm": 1.3414266109466553,
+      "learning_rate": 9.988356591413504e-05,
+      "loss": 0.7075814008712769,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6015102481121899,
+      "grad_norm": 1.4575660228729248,
+      "learning_rate": 9.988188523952973e-05,
+      "loss": 0.8237126469612122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6023732470334412,
+      "grad_norm": 1.3836476802825928,
+      "learning_rate": 9.988019253613265e-05,
+      "loss": 0.7991626858711243,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6032362459546926,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.987848780435201e-05,
+      "loss": 0.6936330199241638,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 1.171294093132019,
+      "learning_rate": 9.98767710445989e-05,
+      "loss": 0.7834921479225159,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "eval_loss": 0.7866972088813782,
+      "eval_runtime": 836.0759,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6049622437971952,
+      "grad_norm": 1.5934456586837769,
+      "learning_rate": 9.987504225728727e-05,
+      "loss": 0.765119194984436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6058252427184466,
+      "grad_norm": 1.231372356414795,
+      "learning_rate": 9.987330144283407e-05,
+      "loss": 0.752061665058136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6066882416396979,
+      "grad_norm": 1.315215826034546,
+      "learning_rate": 9.987154860165903e-05,
+      "loss": 0.7475255727767944,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6075512405609493,
+      "grad_norm": 1.2640653848648071,
+      "learning_rate": 9.986978373418488e-05,
+      "loss": 0.777019202709198,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 1.085493803024292,
+      "learning_rate": 9.98680068408372e-05,
+      "loss": 0.6682190299034119,
+      "step": 1410
+    },
+    {
+      "epoch": 0.609277238403452,
+      "grad_norm": 1.4247251749038696,
+      "learning_rate": 9.986621792204447e-05,
+      "loss": 0.7765674591064453,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6101402373247033,
+      "grad_norm": 1.2993831634521484,
+      "learning_rate": 9.986441697823808e-05,
+      "loss": 0.7429938316345215,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6110032362459547,
+      "grad_norm": 2.614980936050415,
+      "learning_rate": 9.986260400985233e-05,
+      "loss": 0.7838386297225952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6118662351672061,
+      "grad_norm": 1.4305615425109863,
+      "learning_rate": 9.98607790173244e-05,
+      "loss": 0.7679475545883179,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 1.1412850618362427,
+      "learning_rate": 9.98589420010944e-05,
+      "loss": 0.6889669895172119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6135922330097088,
+      "grad_norm": 1.372313141822815,
+      "learning_rate": 9.98570929616053e-05,
+      "loss": 0.7659094333648682,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6144552319309601,
+      "grad_norm": 1.2629581689834595,
+      "learning_rate": 9.985523189930297e-05,
+      "loss": 0.7404800653457642,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6153182308522115,
+      "grad_norm": 1.5652254819869995,
+      "learning_rate": 9.985335881463623e-05,
+      "loss": 0.8082308769226074,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6161812297734628,
+      "grad_norm": 1.2268856763839722,
+      "learning_rate": 9.985147370805675e-05,
+      "loss": 0.7773997187614441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 1.2190957069396973,
+      "learning_rate": 9.984957658001912e-05,
+      "loss": 0.7412725687026978,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6179072276159655,
+      "grad_norm": 1.481990098953247,
+      "learning_rate": 9.984766743098081e-05,
+      "loss": 0.7257611751556396,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6187702265372168,
+      "grad_norm": 1.0297027826309204,
+      "learning_rate": 9.984574626140223e-05,
+      "loss": 0.7151464819908142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6196332254584682,
+      "grad_norm": 1.2404571771621704,
+      "learning_rate": 9.984381307174663e-05,
+      "loss": 0.7855883240699768,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6204962243797195,
+      "grad_norm": 1.2952802181243896,
+      "learning_rate": 9.984186786248023e-05,
+      "loss": 0.6360295414924622,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.5405724048614502,
+      "learning_rate": 9.983991063407206e-05,
+      "loss": 0.7280036807060242,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.303699254989624,
+      "learning_rate": 9.983794138699411e-05,
+      "loss": 0.7553068995475769,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6230852211434735,
+      "grad_norm": 1.1452997922897339,
+      "learning_rate": 9.983596012172127e-05,
+      "loss": 0.7329047918319702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.623948220064725,
+      "grad_norm": 1.2648253440856934,
+      "learning_rate": 9.983396683873133e-05,
+      "loss": 0.753168523311615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6248112189859762,
+      "grad_norm": 1.2095013856887817,
+      "learning_rate": 9.983196153850493e-05,
+      "loss": 0.7378705739974976,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 1.2855647802352905,
+      "learning_rate": 9.982994422152564e-05,
+      "loss": 0.7910448908805847,
+      "step": 1450
+    },
+    {
+      "epoch": 0.626537216828479,
+      "grad_norm": 1.3895999193191528,
+      "learning_rate": 9.982791488827994e-05,
+      "loss": 0.785302996635437,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6274002157497303,
+      "grad_norm": 1.2398446798324585,
+      "learning_rate": 9.982587353925719e-05,
+      "loss": 0.7542029023170471,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6282632146709817,
+      "grad_norm": 1.2903637886047363,
+      "learning_rate": 9.982382017494965e-05,
+      "loss": 0.8081737756729126,
+      "step": 1456
+    },
+    {
+      "epoch": 0.629126213592233,
+      "grad_norm": 1.223611831665039,
+      "learning_rate": 9.98217547958525e-05,
+      "loss": 0.7886892557144165,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 1.3036459684371948,
+      "learning_rate": 9.981967740246377e-05,
+      "loss": 0.844115138053894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6308522114347357,
+      "grad_norm": 1.6421090364456177,
+      "learning_rate": 9.98175879952844e-05,
+      "loss": 0.7784068584442139,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6317152103559871,
+      "grad_norm": 1.244330644607544,
+      "learning_rate": 9.981548657481828e-05,
+      "loss": 0.746979832649231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6325782092772384,
+      "grad_norm": 1.0323492288589478,
+      "learning_rate": 9.981337314157215e-05,
+      "loss": 0.6680153608322144,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6334412081984897,
+      "grad_norm": 1.2238314151763916,
+      "learning_rate": 9.981124769605563e-05,
+      "loss": 0.7103193402290344,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 1.2647801637649536,
+      "learning_rate": 9.98091102387813e-05,
+      "loss": 0.7426751255989075,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6351672060409924,
+      "grad_norm": 1.1923259496688843,
+      "learning_rate": 9.980696077026456e-05,
+      "loss": 0.7664520144462585,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6360302049622438,
+      "grad_norm": 1.0503894090652466,
+      "learning_rate": 9.980479929102377e-05,
+      "loss": 0.7348257899284363,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6368932038834951,
+      "grad_norm": 1.2223190069198608,
+      "learning_rate": 9.980262580158014e-05,
+      "loss": 0.687703013420105,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6377562028047465,
+      "grad_norm": 1.1469025611877441,
+      "learning_rate": 9.980044030245782e-05,
+      "loss": 0.7903051972389221,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 1.2581169605255127,
+      "learning_rate": 9.979824279418381e-05,
+      "loss": 0.7920540571212769,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6394822006472491,
+      "grad_norm": 1.2616536617279053,
+      "learning_rate": 9.979603327728806e-05,
+      "loss": 0.7030006647109985,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6403451995685006,
+      "grad_norm": 1.3970606327056885,
+      "learning_rate": 9.979381175230335e-05,
+      "loss": 0.7489339709281921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6412081984897519,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 9.979157821976542e-05,
+      "loss": 0.6733010411262512,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6420711974110033,
+      "grad_norm": 1.4987449645996094,
+      "learning_rate": 9.978933268021286e-05,
+      "loss": 0.8221960663795471,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 1.2843966484069824,
+      "learning_rate": 9.978707513418719e-05,
+      "loss": 0.8829001188278198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.643797195253506,
+      "grad_norm": 1.175879955291748,
+      "learning_rate": 9.978480558223278e-05,
+      "loss": 0.7605764269828796,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6446601941747573,
+      "grad_norm": 1.1991835832595825,
+      "learning_rate": 9.978252402489692e-05,
+      "loss": 0.7618141174316406,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6455231930960086,
+      "grad_norm": 1.167635202407837,
+      "learning_rate": 9.978023046272983e-05,
+      "loss": 0.7071146368980408,
+      "step": 1496
+    },
+    {
+      "epoch": 0.64638619201726,
+      "grad_norm": 1.263371229171753,
+      "learning_rate": 9.977792489628456e-05,
+      "loss": 0.7559816837310791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 1.2108519077301025,
+      "learning_rate": 9.977560732611713e-05,
+      "loss": 0.7253944873809814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "eval_loss": 0.778505802154541,
+      "eval_runtime": 836.2317,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6481121898597627,
+      "grad_norm": 1.4794813394546509,
+      "learning_rate": 9.977327775278636e-05,
+      "loss": 0.8212620615959167,
+      "step": 1502
+    },
+    {
+      "epoch": 0.648975188781014,
+      "grad_norm": 1.2120516300201416,
+      "learning_rate": 9.977093617685404e-05,
+      "loss": 0.7220476865768433,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6498381877022654,
+      "grad_norm": 1.3129044771194458,
+      "learning_rate": 9.976858259888484e-05,
+      "loss": 0.7791266441345215,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6507011866235167,
+      "grad_norm": 1.146490454673767,
+      "learning_rate": 9.976621701944628e-05,
+      "loss": 0.7052001357078552,
+      "step": 1508
+    },
+    {
+      "epoch": 0.651564185544768,
+      "grad_norm": 1.1903722286224365,
+      "learning_rate": 9.976383943910885e-05,
+      "loss": 0.7135450839996338,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6524271844660194,
+      "grad_norm": 1.4106701612472534,
+      "learning_rate": 9.976144985844589e-05,
+      "loss": 0.8114889860153198,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6532901833872707,
+      "grad_norm": 1.245661973953247,
+      "learning_rate": 9.97590482780336e-05,
+      "loss": 0.7501562237739563,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6541531823085222,
+      "grad_norm": 1.2918543815612793,
+      "learning_rate": 9.975663469845114e-05,
+      "loss": 0.7697334289550781,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6550161812297735,
+      "grad_norm": 1.2699558734893799,
+      "learning_rate": 9.975420912028053e-05,
+      "loss": 0.7457168102264404,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6558791801510249,
+      "grad_norm": 1.3968263864517212,
+      "learning_rate": 9.975177154410668e-05,
+      "loss": 0.8368204236030579,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6567421790722762,
+      "grad_norm": 1.2161848545074463,
+      "learning_rate": 9.974932197051741e-05,
+      "loss": 0.7355788350105286,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6576051779935275,
+      "grad_norm": 1.129886507987976,
+      "learning_rate": 9.974686040010341e-05,
+      "loss": 0.8023449182510376,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6584681769147789,
+      "grad_norm": 1.1114557981491089,
+      "learning_rate": 9.974438683345829e-05,
+      "loss": 0.7965474724769592,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6593311758360302,
+      "grad_norm": 1.5741950273513794,
+      "learning_rate": 9.974190127117853e-05,
+      "loss": 0.7176838517189026,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 1.1439597606658936,
+      "learning_rate": 9.973940371386352e-05,
+      "loss": 0.7114462852478027,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6610571736785329,
+      "grad_norm": 1.273659348487854,
+      "learning_rate": 9.973689416211553e-05,
+      "loss": 0.754712700843811,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6619201725997842,
+      "grad_norm": 3.9874372482299805,
+      "learning_rate": 9.973437261653973e-05,
+      "loss": 0.6720641255378723,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6627831715210356,
+      "grad_norm": 1.2506502866744995,
+      "learning_rate": 9.973183907774417e-05,
+      "loss": 0.793043851852417,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6636461704422869,
+      "grad_norm": 1.43257737159729,
+      "learning_rate": 9.972929354633982e-05,
+      "loss": 0.7806380987167358,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6645091693635383,
+      "grad_norm": 1.4794938564300537,
+      "learning_rate": 9.97267360229405e-05,
+      "loss": 0.7176088094711304,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6653721682847896,
+      "grad_norm": 1.2398169040679932,
+      "learning_rate": 9.972416650816297e-05,
+      "loss": 0.7145742774009705,
+      "step": 1542
+    },
+    {
+      "epoch": 0.666235167206041,
+      "grad_norm": 1.1309953927993774,
+      "learning_rate": 9.972158500262683e-05,
+      "loss": 0.8025617003440857,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6670981661272923,
+      "grad_norm": 1.2479451894760132,
+      "learning_rate": 9.971899150695463e-05,
+      "loss": 0.7808589339256287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6679611650485436,
+      "grad_norm": 1.2538350820541382,
+      "learning_rate": 9.971638602177175e-05,
+      "loss": 0.7096936106681824,
+      "step": 1548
+    },
+    {
+      "epoch": 0.668824163969795,
+      "grad_norm": 1.1431525945663452,
+      "learning_rate": 9.97137685477065e-05,
+      "loss": 0.6529367566108704,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6696871628910464,
+      "grad_norm": 1.5966559648513794,
+      "learning_rate": 9.971113908539008e-05,
+      "loss": 0.7780346870422363,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6705501618122978,
+      "grad_norm": 1.186475157737732,
+      "learning_rate": 9.970849763545658e-05,
+      "loss": 0.6889166831970215,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6714131607335491,
+      "grad_norm": 1.1922476291656494,
+      "learning_rate": 9.970584419854295e-05,
+      "loss": 0.7154867649078369,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6722761596548005,
+      "grad_norm": 1.30132257938385,
+      "learning_rate": 9.970317877528906e-05,
+      "loss": 0.7419366240501404,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6731391585760518,
+      "grad_norm": 1.1539888381958008,
+      "learning_rate": 9.970050136633769e-05,
+      "loss": 0.7856602072715759,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6740021574973031,
+      "grad_norm": 1.3056161403656006,
+      "learning_rate": 9.969781197233445e-05,
+      "loss": 0.6834115386009216,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6748651564185545,
+      "grad_norm": 1.2899249792099,
+      "learning_rate": 9.96951105939279e-05,
+      "loss": 0.8490727543830872,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6757281553398058,
+      "grad_norm": 1.2182374000549316,
+      "learning_rate": 9.969239723176946e-05,
+      "loss": 0.7462526559829712,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6765911542610572,
+      "grad_norm": 1.1231273412704468,
+      "learning_rate": 9.968967188651344e-05,
+      "loss": 0.7145209908485413,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6774541531823085,
+      "grad_norm": 1.1776760816574097,
+      "learning_rate": 9.968693455881703e-05,
+      "loss": 0.7911707162857056,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6783171521035599,
+      "grad_norm": 1.3542319536209106,
+      "learning_rate": 9.968418524934035e-05,
+      "loss": 0.6946635246276855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6791801510248112,
+      "grad_norm": 1.3073060512542725,
+      "learning_rate": 9.96814239587464e-05,
+      "loss": 0.7597488164901733,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6800431499460625,
+      "grad_norm": 1.3205567598342896,
+      "learning_rate": 9.967865068770099e-05,
+      "loss": 0.6763246655464172,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6809061488673139,
+      "grad_norm": 1.1734344959259033,
+      "learning_rate": 9.967586543687295e-05,
+      "loss": 0.7560655474662781,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6817691477885652,
+      "grad_norm": 1.445770025253296,
+      "learning_rate": 9.967306820693388e-05,
+      "loss": 0.8050205707550049,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6826321467098166,
+      "grad_norm": 1.1185686588287354,
+      "learning_rate": 9.967025899855836e-05,
+      "loss": 0.6628211140632629,
+      "step": 1582
+    },
+    {
+      "epoch": 0.683495145631068,
+      "grad_norm": 1.3831934928894043,
+      "learning_rate": 9.966743781242379e-05,
+      "loss": 0.7684114575386047,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6843581445523194,
+      "grad_norm": 1.2417019605636597,
+      "learning_rate": 9.966460464921047e-05,
+      "loss": 0.7084140777587891,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6852211434735707,
+      "grad_norm": 1.2138086557388306,
+      "learning_rate": 9.966175950960163e-05,
+      "loss": 0.6817399263381958,
+      "step": 1588
+    },
+    {
+      "epoch": 0.686084142394822,
+      "grad_norm": 1.399778962135315,
+      "learning_rate": 9.965890239428339e-05,
+      "loss": 0.8422354459762573,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6869471413160734,
+      "grad_norm": 1.4348714351654053,
+      "learning_rate": 9.965603330394468e-05,
+      "loss": 0.7432681322097778,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6878101402373247,
+      "grad_norm": 1.192425012588501,
+      "learning_rate": 9.965315223927741e-05,
+      "loss": 0.7571184635162354,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6886731391585761,
+      "grad_norm": 1.2834038734436035,
+      "learning_rate": 9.96502592009763e-05,
+      "loss": 0.802443265914917,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6895361380798274,
+      "grad_norm": 1.1650444269180298,
+      "learning_rate": 9.964735418973901e-05,
+      "loss": 0.8113405704498291,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "grad_norm": 1.1714438199996948,
+      "learning_rate": 9.964443720626606e-05,
+      "loss": 0.7507118582725525,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "eval_loss": 0.7706249952316284,
+      "eval_runtime": 821.4894,
+      "eval_samples_per_second": 2.508,
+      "eval_steps_per_second": 2.508,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6912621359223301,
+      "grad_norm": 1.175459861755371,
+      "learning_rate": 9.964150825126088e-05,
+      "loss": 0.7422796487808228,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6921251348435814,
+      "grad_norm": 1.6418036222457886,
+      "learning_rate": 9.963856732542978e-05,
+      "loss": 0.7125853300094604,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6929881337648328,
+      "grad_norm": 1.3256275653839111,
+      "learning_rate": 9.963561442948193e-05,
+      "loss": 0.757859468460083,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6938511326860841,
+      "grad_norm": 1.19981050491333,
+      "learning_rate": 9.963264956412944e-05,
+      "loss": 0.6849292516708374,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6947141316073355,
+      "grad_norm": 1.0392169952392578,
+      "learning_rate": 9.962967273008723e-05,
+      "loss": 0.7636215090751648,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6955771305285868,
+      "grad_norm": 1.1106646060943604,
+      "learning_rate": 9.962668392807318e-05,
+      "loss": 0.7208617329597473,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6964401294498382,
+      "grad_norm": 1.1569621562957764,
+      "learning_rate": 9.962368315880801e-05,
+      "loss": 0.671805739402771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6973031283710895,
+      "grad_norm": 1.2157769203186035,
+      "learning_rate": 9.962067042301537e-05,
+      "loss": 0.7648762464523315,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6981661272923408,
+      "grad_norm": 1.4804017543792725,
+      "learning_rate": 9.961764572142173e-05,
+      "loss": 0.8490509390830994,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 1.0344431400299072,
+      "learning_rate": 9.961460905475651e-05,
+      "loss": 0.5950891971588135,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6998921251348436,
+      "grad_norm": 1.1109060049057007,
+      "learning_rate": 9.9611560423752e-05,
+      "loss": 0.7288329005241394,
+      "step": 1622
+    },
+    {
+      "epoch": 0.700755124056095,
+      "grad_norm": 1.096996545791626,
+      "learning_rate": 9.960849982914331e-05,
+      "loss": 0.7074542045593262,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7016181229773463,
+      "grad_norm": 1.3109341859817505,
+      "learning_rate": 9.960542727166853e-05,
+      "loss": 0.7348865270614624,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7024811218985976,
+      "grad_norm": 1.3402032852172852,
+      "learning_rate": 9.96023427520686e-05,
+      "loss": 0.7658320665359497,
+      "step": 1628
+    },
+    {
+      "epoch": 0.703344120819849,
+      "grad_norm": 1.3712561130523682,
+      "learning_rate": 9.959924627108732e-05,
+      "loss": 0.809951901435852,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7042071197411003,
+      "grad_norm": 1.2278292179107666,
+      "learning_rate": 9.959613782947139e-05,
+      "loss": 0.6834598779678345,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7050701186623517,
+      "grad_norm": 1.0510832071304321,
+      "learning_rate": 9.95930174279704e-05,
+      "loss": 0.7416030764579773,
+      "step": 1634
+    },
+    {
+      "epoch": 0.705933117583603,
+      "grad_norm": 1.130991816520691,
+      "learning_rate": 9.958988506733684e-05,
+      "loss": 0.7304092645645142,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7067961165048544,
+      "grad_norm": 1.3328672647476196,
+      "learning_rate": 9.958674074832603e-05,
+      "loss": 0.8230003118515015,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7076591154261057,
+      "grad_norm": 1.3842110633850098,
+      "learning_rate": 9.958358447169625e-05,
+      "loss": 0.8413107991218567,
+      "step": 1640
+    },
+    {
+      "epoch": 0.708522114347357,
+      "grad_norm": 1.3143812417984009,
+      "learning_rate": 9.958041623820858e-05,
+      "loss": 0.77432781457901,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7093851132686084,
+      "grad_norm": 1.4727009534835815,
+      "learning_rate": 9.957723604862704e-05,
+      "loss": 0.7563844323158264,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7102481121898597,
+      "grad_norm": 1.1892327070236206,
+      "learning_rate": 9.957404390371854e-05,
+      "loss": 0.7085853219032288,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.124180555343628,
+      "learning_rate": 9.957083980425282e-05,
+      "loss": 0.6837730407714844,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 1.2224774360656738,
+      "learning_rate": 9.956762375100255e-05,
+      "loss": 0.7255740165710449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7128371089536139,
+      "grad_norm": 1.360425591468811,
+      "learning_rate": 9.956439574474326e-05,
+      "loss": 0.7849710583686829,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7137001078748652,
+      "grad_norm": 1.2818151712417603,
+      "learning_rate": 9.956115578625338e-05,
+      "loss": 0.7669665813446045,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7145631067961165,
+      "grad_norm": 1.2415164709091187,
+      "learning_rate": 9.95579038763142e-05,
+      "loss": 0.7052890658378601,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7154261057173679,
+      "grad_norm": 1.3983269929885864,
+      "learning_rate": 9.955464001570992e-05,
+      "loss": 0.8300023674964905,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7162891046386192,
+      "grad_norm": 1.1591373682022095,
+      "learning_rate": 9.955136420522759e-05,
+      "loss": 0.7742894887924194,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7171521035598706,
+      "grad_norm": 1.2503128051757812,
+      "learning_rate": 9.954807644565715e-05,
+      "loss": 0.8458141684532166,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7180151024811219,
+      "grad_norm": 1.2954912185668945,
+      "learning_rate": 9.954477673779146e-05,
+      "loss": 0.7044742107391357,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7188781014023733,
+      "grad_norm": 1.1669788360595703,
+      "learning_rate": 9.954146508242619e-05,
+      "loss": 0.7917183041572571,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7197411003236246,
+      "grad_norm": 1.1466871500015259,
+      "learning_rate": 9.953814148035996e-05,
+      "loss": 0.7174562811851501,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7206040992448759,
+      "grad_norm": 1.367224097251892,
+      "learning_rate": 9.953480593239421e-05,
+      "loss": 0.8925880193710327,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7214670981661273,
+      "grad_norm": 1.049497365951538,
+      "learning_rate": 9.953145843933334e-05,
+      "loss": 0.70949387550354,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7223300970873786,
+      "grad_norm": 0.9995084404945374,
+      "learning_rate": 9.952809900198455e-05,
+      "loss": 0.6731935143470764,
+      "step": 1674
+    },
+    {
+      "epoch": 0.72319309600863,
+      "grad_norm": 1.100899577140808,
+      "learning_rate": 9.952472762115794e-05,
+      "loss": 0.7316275835037231,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7240560949298813,
+      "grad_norm": 1.2594135999679565,
+      "learning_rate": 9.952134429766655e-05,
+      "loss": 0.7379520535469055,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7249190938511327,
+      "grad_norm": 1.3124911785125732,
+      "learning_rate": 9.951794903232621e-05,
+      "loss": 0.7072247266769409,
+      "step": 1680
+    },
+    {
+      "epoch": 0.725782092772384,
+      "grad_norm": 1.212144374847412,
+      "learning_rate": 9.95145418259557e-05,
+      "loss": 0.745680034160614,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7266450916936353,
+      "grad_norm": 1.3913586139678955,
+      "learning_rate": 9.951112267937663e-05,
+      "loss": 0.796852707862854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7275080906148867,
+      "grad_norm": 1.3227289915084839,
+      "learning_rate": 9.950769159341353e-05,
+      "loss": 0.700892984867096,
+      "step": 1686
+    },
+    {
+      "epoch": 0.728371089536138,
+      "grad_norm": 1.4656381607055664,
+      "learning_rate": 9.950424856889378e-05,
+      "loss": 0.7973095178604126,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7292340884573895,
+      "grad_norm": 1.1698973178863525,
+      "learning_rate": 9.950079360664766e-05,
+      "loss": 0.7811378836631775,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7300970873786408,
+      "grad_norm": 1.201695203781128,
+      "learning_rate": 9.949732670750831e-05,
+      "loss": 0.753012478351593,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7309600862998922,
+      "grad_norm": 1.3465325832366943,
+      "learning_rate": 9.949384787231175e-05,
+      "loss": 0.7620642781257629,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7318230852211435,
+      "grad_norm": 1.2191137075424194,
+      "learning_rate": 9.94903571018969e-05,
+      "loss": 0.7479873299598694,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7326860841423948,
+      "grad_norm": 1.1412187814712524,
+      "learning_rate": 9.948685439710554e-05,
+      "loss": 0.7327720522880554,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "grad_norm": 1.310327410697937,
+      "learning_rate": 9.948333975878233e-05,
+      "loss": 0.761475682258606,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "eval_loss": 0.7649155259132385,
+      "eval_runtime": 828.3542,
+      "eval_samples_per_second": 2.487,
+      "eval_steps_per_second": 2.487,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7344120819848975,
+      "grad_norm": 1.3031405210494995,
+      "learning_rate": 9.947981318777482e-05,
+      "loss": 0.7564467191696167,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7352750809061489,
+      "grad_norm": 1.2672377824783325,
+      "learning_rate": 9.94762746849334e-05,
+      "loss": 0.7689576745033264,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7361380798274002,
+      "grad_norm": 1.1159595251083374,
+      "learning_rate": 9.947272425111138e-05,
+      "loss": 0.7923810482025146,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7370010787486516,
+      "grad_norm": 1.1190921068191528,
+      "learning_rate": 9.946916188716495e-05,
+      "loss": 0.6540790796279907,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 1.373832106590271,
+      "learning_rate": 9.946558759395312e-05,
+      "loss": 0.7267473936080933,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7387270765911542,
+      "grad_norm": 1.2346374988555908,
+      "learning_rate": 9.946200137233785e-05,
+      "loss": 0.7693060636520386,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7395900755124056,
+      "grad_norm": 1.1629668474197388,
+      "learning_rate": 9.945840322318391e-05,
+      "loss": 0.7751407027244568,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7404530744336569,
+      "grad_norm": 1.2268708944320679,
+      "learning_rate": 9.945479314735899e-05,
+      "loss": 0.7798545360565186,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7413160733549083,
+      "grad_norm": 0.9968738555908203,
+      "learning_rate": 9.945117114573366e-05,
+      "loss": 0.6128867268562317,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7421790722761596,
+      "grad_norm": 1.2135992050170898,
+      "learning_rate": 9.944753721918133e-05,
+      "loss": 0.7536954283714294,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7430420711974111,
+      "grad_norm": 1.1361218690872192,
+      "learning_rate": 9.94438913685783e-05,
+      "loss": 0.7034479379653931,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7439050701186624,
+      "grad_norm": 1.0728967189788818,
+      "learning_rate": 9.944023359480378e-05,
+      "loss": 0.7285382747650146,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7447680690399137,
+      "grad_norm": 1.3613945245742798,
+      "learning_rate": 9.94365638987398e-05,
+      "loss": 0.7865195274353027,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7456310679611651,
+      "grad_norm": 1.0985898971557617,
+      "learning_rate": 9.94328822812713e-05,
+      "loss": 0.7511224746704102,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7464940668824164,
+      "grad_norm": 1.1506208181381226,
+      "learning_rate": 9.942918874328606e-05,
+      "loss": 0.732453465461731,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7473570658036678,
+      "grad_norm": 1.0995100736618042,
+      "learning_rate": 9.94254832856748e-05,
+      "loss": 0.702409029006958,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7482200647249191,
+      "grad_norm": 1.0075604915618896,
+      "learning_rate": 9.942176590933106e-05,
+      "loss": 0.7515703439712524,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7490830636461704,
+      "grad_norm": 1.3360812664031982,
+      "learning_rate": 9.941803661515126e-05,
+      "loss": 0.6910237073898315,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7499460625674218,
+      "grad_norm": 1.0172319412231445,
+      "learning_rate": 9.941429540403471e-05,
+      "loss": 0.7185606956481934,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7508090614886731,
+      "grad_norm": 1.2721658945083618,
+      "learning_rate": 9.941054227688358e-05,
+      "loss": 0.7136536836624146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7516720604099245,
+      "grad_norm": 1.0775413513183594,
+      "learning_rate": 9.940677723460292e-05,
+      "loss": 0.7595619559288025,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7525350593311758,
+      "grad_norm": 1.3436479568481445,
+      "learning_rate": 9.940300027810067e-05,
+      "loss": 0.7602618336677551,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7533980582524272,
+      "grad_norm": 1.2611958980560303,
+      "learning_rate": 9.93992114082876e-05,
+      "loss": 0.7954604029655457,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7542610571736785,
+      "grad_norm": 1.2794841527938843,
+      "learning_rate": 9.93954106260774e-05,
+      "loss": 0.7661190629005432,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7551240560949298,
+      "grad_norm": 1.3042595386505127,
+      "learning_rate": 9.939159793238661e-05,
+      "loss": 0.6855312585830688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7559870550161812,
+      "grad_norm": 1.2799923419952393,
+      "learning_rate": 9.938777332813463e-05,
+      "loss": 0.7970905900001526,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7568500539374325,
+      "grad_norm": 1.169205904006958,
+      "learning_rate": 9.938393681424375e-05,
+      "loss": 0.7072843313217163,
+      "step": 1754
+    },
+    {
+      "epoch": 0.757713052858684,
+      "grad_norm": 1.3833354711532593,
+      "learning_rate": 9.938008839163913e-05,
+      "loss": 0.7134629487991333,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7585760517799353,
+      "grad_norm": 1.223452091217041,
+      "learning_rate": 9.93762280612488e-05,
+      "loss": 0.7873107194900513,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7594390507011867,
+      "grad_norm": 1.2740365266799927,
+      "learning_rate": 9.937235582400367e-05,
+      "loss": 0.7680654525756836,
+      "step": 1760
+    },
+    {
+      "epoch": 0.760302049622438,
+      "grad_norm": 1.2564697265625,
+      "learning_rate": 9.93684716808375e-05,
+      "loss": 0.734126627445221,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7611650485436893,
+      "grad_norm": 1.3858847618103027,
+      "learning_rate": 9.936457563268694e-05,
+      "loss": 0.7972320914268494,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7620280474649407,
+      "grad_norm": 1.2705003023147583,
+      "learning_rate": 9.93606676804915e-05,
+      "loss": 0.7904276847839355,
+      "step": 1766
+    },
+    {
+      "epoch": 0.762891046386192,
+      "grad_norm": 1.210766077041626,
+      "learning_rate": 9.935674782519358e-05,
+      "loss": 0.7542425990104675,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7637540453074434,
+      "grad_norm": 1.4367541074752808,
+      "learning_rate": 9.93528160677384e-05,
+      "loss": 0.7975004315376282,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7646170442286947,
+      "grad_norm": 1.0994079113006592,
+      "learning_rate": 9.934887240907414e-05,
+      "loss": 0.7537807822227478,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7654800431499461,
+      "grad_norm": 1.111660122871399,
+      "learning_rate": 9.934491685015174e-05,
+      "loss": 0.7333855628967285,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7663430420711974,
+      "grad_norm": 1.2879043817520142,
+      "learning_rate": 9.934094939192508e-05,
+      "loss": 0.7472840547561646,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7672060409924487,
+      "grad_norm": 1.3618478775024414,
+      "learning_rate": 9.933697003535094e-05,
+      "loss": 0.7644904255867004,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7680690399137001,
+      "grad_norm": 1.1235666275024414,
+      "learning_rate": 9.933297878138887e-05,
+      "loss": 0.7872052192687988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7689320388349514,
+      "grad_norm": 1.3741775751113892,
+      "learning_rate": 9.932897563100136e-05,
+      "loss": 0.7343124747276306,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7697950377562028,
+      "grad_norm": 1.2115039825439453,
+      "learning_rate": 9.932496058515375e-05,
+      "loss": 0.7486474514007568,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7706580366774541,
+      "grad_norm": 1.183457851409912,
+      "learning_rate": 9.932093364481426e-05,
+      "loss": 0.7333025932312012,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7715210355987056,
+      "grad_norm": 1.2841869592666626,
+      "learning_rate": 9.931689481095394e-05,
+      "loss": 0.7241485118865967,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7723840345199569,
+      "grad_norm": 1.2312977313995361,
+      "learning_rate": 9.931284408454678e-05,
+      "loss": 0.7510396242141724,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7732470334412082,
+      "grad_norm": 1.144895315170288,
+      "learning_rate": 9.930878146656956e-05,
+      "loss": 0.7108346819877625,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7741100323624596,
+      "grad_norm": 1.18603515625,
+      "learning_rate": 9.9304706958002e-05,
+      "loss": 0.7687607407569885,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7749730312837109,
+      "grad_norm": 1.2827935218811035,
+      "learning_rate": 9.93006205598266e-05,
+      "loss": 0.7826682329177856,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7758360302049623,
+      "grad_norm": 1.164852261543274,
+      "learning_rate": 9.929652227302881e-05,
+      "loss": 0.6970428824424744,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 1.0439956188201904,
+      "learning_rate": 9.92924120985969e-05,
+      "loss": 0.7238826751708984,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "eval_loss": 0.7598537802696228,
+      "eval_runtime": 838.9645,
+      "eval_samples_per_second": 2.455,
+      "eval_steps_per_second": 2.455,
+      "step": 1800
+    },
+    {
+      "epoch": 0.777562028047465,
+      "grad_norm": 1.2464938163757324,
+      "learning_rate": 9.928829003752202e-05,
+      "loss": 0.7723641395568848,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7784250269687163,
+      "grad_norm": 1.2445050477981567,
+      "learning_rate": 9.92841560907982e-05,
+      "loss": 0.7154937982559204,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7792880258899676,
+      "grad_norm": 1.2235661745071411,
+      "learning_rate": 9.928001025942231e-05,
+      "loss": 0.8325800895690918,
+      "step": 1806
+    },
+    {
+      "epoch": 0.780151024811219,
+      "grad_norm": 1.268066644668579,
+      "learning_rate": 9.927585254439409e-05,
+      "loss": 0.7093122601509094,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7810140237324703,
+      "grad_norm": 1.1731390953063965,
+      "learning_rate": 9.927168294671618e-05,
+      "loss": 0.7601775527000427,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7818770226537217,
+      "grad_norm": 1.0685608386993408,
+      "learning_rate": 9.926750146739403e-05,
+      "loss": 0.6889236569404602,
+      "step": 1812
+    },
+    {
+      "epoch": 0.782740021574973,
+      "grad_norm": 1.1098639965057373,
+      "learning_rate": 9.926330810743602e-05,
+      "loss": 0.7397196888923645,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7836030204962244,
+      "grad_norm": 1.3606436252593994,
+      "learning_rate": 9.925910286785332e-05,
+      "loss": 0.7180029153823853,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7844660194174757,
+      "grad_norm": 1.143418550491333,
+      "learning_rate": 9.925488574966004e-05,
+      "loss": 0.7039036750793457,
+      "step": 1818
+    },
+    {
+      "epoch": 0.785329018338727,
+      "grad_norm": 1.180968999862671,
+      "learning_rate": 9.92506567538731e-05,
+      "loss": 0.741872251033783,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7861920172599784,
+      "grad_norm": 1.1953548192977905,
+      "learning_rate": 9.924641588151231e-05,
+      "loss": 0.7580198049545288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7870550161812297,
+      "grad_norm": 1.3049067258834839,
+      "learning_rate": 9.924216313360033e-05,
+      "loss": 0.7479216456413269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7879180151024812,
+      "grad_norm": 1.3962669372558594,
+      "learning_rate": 9.92378985111627e-05,
+      "loss": 0.7382357716560364,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7887810140237325,
+      "grad_norm": 1.1519217491149902,
+      "learning_rate": 9.92336220152278e-05,
+      "loss": 0.6356467604637146,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7896440129449838,
+      "grad_norm": 1.2544089555740356,
+      "learning_rate": 9.922933364682689e-05,
+      "loss": 0.7674193978309631,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7905070118662352,
+      "grad_norm": 1.1358295679092407,
+      "learning_rate": 9.922503340699412e-05,
+      "loss": 0.7083898782730103,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7913700107874865,
+      "grad_norm": 1.1744481325149536,
+      "learning_rate": 9.922072129676644e-05,
+      "loss": 0.6273066401481628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7922330097087379,
+      "grad_norm": 1.20509672164917,
+      "learning_rate": 9.92163973171837e-05,
+      "loss": 0.808181643486023,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7930960086299892,
+      "grad_norm": 1.2035597562789917,
+      "learning_rate": 9.921206146928863e-05,
+      "loss": 0.6631553769111633,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7939590075512406,
+      "grad_norm": 1.058695912361145,
+      "learning_rate": 9.920771375412678e-05,
+      "loss": 0.72016841173172,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7948220064724919,
+      "grad_norm": 1.035791039466858,
+      "learning_rate": 9.92033541727466e-05,
+      "loss": 0.7240060567855835,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7956850053937432,
+      "grad_norm": 1.1845753192901611,
+      "learning_rate": 9.919898272619935e-05,
+      "loss": 0.743224561214447,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7965480043149946,
+      "grad_norm": 1.2942636013031006,
+      "learning_rate": 9.919459941553922e-05,
+      "loss": 0.7064419388771057,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7974110032362459,
+      "grad_norm": 1.2736715078353882,
+      "learning_rate": 9.91902042418232e-05,
+      "loss": 0.805932343006134,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7982740021574973,
+      "grad_norm": 1.2803614139556885,
+      "learning_rate": 9.918579720611119e-05,
+      "loss": 0.8050802946090698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7991370010787486,
+      "grad_norm": 1.2225072383880615,
+      "learning_rate": 9.91813783094659e-05,
+      "loss": 0.7013180255889893,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3375571966171265,
+      "learning_rate": 9.917694755295294e-05,
+      "loss": 0.7415283918380737,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8008629989212513,
+      "grad_norm": 1.2209177017211914,
+      "learning_rate": 9.917250493764079e-05,
+      "loss": 0.7264692783355713,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8017259978425026,
+      "grad_norm": 1.1158864498138428,
+      "learning_rate": 9.916805046460074e-05,
+      "loss": 0.7083719372749329,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8025889967637541,
+      "grad_norm": 1.171000361442566,
+      "learning_rate": 9.916358413490696e-05,
+      "loss": 0.820396900177002,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8034519956850054,
+      "grad_norm": 0.9942941069602966,
+      "learning_rate": 9.91591059496365e-05,
+      "loss": 0.6754014492034912,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8043149946062568,
+      "grad_norm": 1.239795207977295,
+      "learning_rate": 9.915461590986926e-05,
+      "loss": 0.813154399394989,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8051779935275081,
+      "grad_norm": 1.076019525527954,
+      "learning_rate": 9.915011401668796e-05,
+      "loss": 0.745580792427063,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8060409924487595,
+      "grad_norm": 1.2134130001068115,
+      "learning_rate": 9.914560027117826e-05,
+      "loss": 0.7564949989318848,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8069039913700108,
+      "grad_norm": 1.1774994134902954,
+      "learning_rate": 9.914107467442859e-05,
+      "loss": 0.665183424949646,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8077669902912621,
+      "grad_norm": 1.0606465339660645,
+      "learning_rate": 9.91365372275303e-05,
+      "loss": 0.7597744464874268,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8086299892125135,
+      "grad_norm": 1.397161841392517,
+      "learning_rate": 9.913198793157757e-05,
+      "loss": 0.7453591227531433,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8094929881337648,
+      "grad_norm": 1.250003695487976,
+      "learning_rate": 9.912742678766742e-05,
+      "loss": 0.715792179107666,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8103559870550162,
+      "grad_norm": 1.2343473434448242,
+      "learning_rate": 9.912285379689979e-05,
+      "loss": 0.7094268798828125,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8112189859762675,
+      "grad_norm": 1.453764796257019,
+      "learning_rate": 9.911826896037739e-05,
+      "loss": 0.6652816534042358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8120819848975189,
+      "grad_norm": 1.030903935432434,
+      "learning_rate": 9.911367227920587e-05,
+      "loss": 0.6764529347419739,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8129449838187702,
+      "grad_norm": 1.2725969552993774,
+      "learning_rate": 9.91090637544937e-05,
+      "loss": 0.7043389678001404,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8138079827400215,
+      "grad_norm": 1.1610769033432007,
+      "learning_rate": 9.910444338735216e-05,
+      "loss": 0.6960558295249939,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8146709816612729,
+      "grad_norm": 1.2049180269241333,
+      "learning_rate": 9.909981117889548e-05,
+      "loss": 0.8029319643974304,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 1.1671873331069946,
+      "learning_rate": 9.909516713024068e-05,
+      "loss": 0.683114767074585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8163969795037757,
+      "grad_norm": 1.247768759727478,
+      "learning_rate": 9.909051124250764e-05,
+      "loss": 0.7193703055381775,
+      "step": 1892
+    },
+    {
+      "epoch": 0.817259978425027,
+      "grad_norm": 1.1376357078552246,
+      "learning_rate": 9.908584351681911e-05,
+      "loss": 0.7957084774971008,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8181229773462784,
+      "grad_norm": 1.1918323040008545,
+      "learning_rate": 9.90811639543007e-05,
+      "loss": 0.7168112993240356,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8189859762675297,
+      "grad_norm": 1.1249650716781616,
+      "learning_rate": 9.907647255608087e-05,
+      "loss": 0.7519245147705078,
+      "step": 1898
+    },
+    {
+      "epoch": 0.819848975188781,
+      "grad_norm": 1.0688576698303223,
+      "learning_rate": 9.907176932329092e-05,
+      "loss": 0.7496176362037659,
+      "step": 1900
+    },
+    {
+      "epoch": 0.819848975188781,
+      "eval_loss": 0.7533769607543945,
+      "eval_runtime": 830.2894,
+      "eval_samples_per_second": 2.481,
+      "eval_steps_per_second": 2.481,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8207119741100324,
+      "grad_norm": 0.9264270663261414,
+      "learning_rate": 9.906705425706499e-05,
+      "loss": 0.639653205871582,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8215749730312837,
+      "grad_norm": 1.2231628894805908,
+      "learning_rate": 9.906232735854016e-05,
+      "loss": 0.7555551528930664,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8224379719525351,
+      "grad_norm": 1.3369156122207642,
+      "learning_rate": 9.905758862885624e-05,
+      "loss": 0.7113592028617859,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8233009708737864,
+      "grad_norm": 1.217238187789917,
+      "learning_rate": 9.905283806915597e-05,
+      "loss": 0.6823481321334839,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8241639697950378,
+      "grad_norm": 1.2462831735610962,
+      "learning_rate": 9.904807568058495e-05,
+      "loss": 0.7347100377082825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8250269687162891,
+      "grad_norm": 1.2507221698760986,
+      "learning_rate": 9.904330146429157e-05,
+      "loss": 0.7981513738632202,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8258899676375404,
+      "grad_norm": 1.2246814966201782,
+      "learning_rate": 9.903851542142715e-05,
+      "loss": 0.759779155254364,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8267529665587918,
+      "grad_norm": 1.1746087074279785,
+      "learning_rate": 9.903371755314579e-05,
+      "loss": 0.741083562374115,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8276159654800431,
+      "grad_norm": 1.1913539171218872,
+      "learning_rate": 9.902890786060448e-05,
+      "loss": 0.7491101622581482,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8284789644012945,
+      "grad_norm": 1.2342501878738403,
+      "learning_rate": 9.902408634496307e-05,
+      "loss": 0.7655978202819824,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8293419633225458,
+      "grad_norm": 1.075892686843872,
+      "learning_rate": 9.901925300738424e-05,
+      "loss": 0.7199848890304565,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8302049622437971,
+      "grad_norm": 1.1517611742019653,
+      "learning_rate": 9.901440784903353e-05,
+      "loss": 0.8325693011283875,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8310679611650486,
+      "grad_norm": 1.2084628343582153,
+      "learning_rate": 9.900955087107933e-05,
+      "loss": 0.7674157023429871,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8319309600862999,
+      "grad_norm": 1.334679365158081,
+      "learning_rate": 9.900468207469287e-05,
+      "loss": 0.7415181398391724,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8327939590075513,
+      "grad_norm": 1.0505887269973755,
+      "learning_rate": 9.899980146104824e-05,
+      "loss": 0.672336220741272,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8336569579288026,
+      "grad_norm": 1.415797233581543,
+      "learning_rate": 9.899490903132238e-05,
+      "loss": 0.8077969551086426,
+      "step": 1932
+    },
+    {
+      "epoch": 0.834519956850054,
+      "grad_norm": 1.1793439388275146,
+      "learning_rate": 9.899000478669507e-05,
+      "loss": 0.8053045272827148,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8353829557713053,
+      "grad_norm": 1.196115493774414,
+      "learning_rate": 9.898508872834896e-05,
+      "loss": 0.7585914731025696,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8362459546925566,
+      "grad_norm": 1.2181763648986816,
+      "learning_rate": 9.898016085746953e-05,
+      "loss": 0.691072940826416,
+      "step": 1938
+    },
+    {
+      "epoch": 0.837108953613808,
+      "grad_norm": 1.2929240465164185,
+      "learning_rate": 9.897522117524512e-05,
+      "loss": 0.7723121643066406,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8379719525350593,
+      "grad_norm": 1.2859764099121094,
+      "learning_rate": 9.89702696828669e-05,
+      "loss": 0.8236626386642456,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8388349514563107,
+      "grad_norm": 1.2043784856796265,
+      "learning_rate": 9.89653063815289e-05,
+      "loss": 0.7315511703491211,
+      "step": 1944
+    },
+    {
+      "epoch": 0.839697950377562,
+      "grad_norm": 1.2063398361206055,
+      "learning_rate": 9.896033127242802e-05,
+      "loss": 0.6707082986831665,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8405609492988134,
+      "grad_norm": 1.1906715631484985,
+      "learning_rate": 9.895534435676393e-05,
+      "loss": 0.6920633912086487,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 1.122567892074585,
+      "learning_rate": 9.895034563573926e-05,
+      "loss": 0.7134768962860107,
+      "step": 1950
+    },
+    {
+      "epoch": 0.842286947141316,
+      "grad_norm": 1.0681787729263306,
+      "learning_rate": 9.894533511055943e-05,
+      "loss": 0.6333967447280884,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8431499460625674,
+      "grad_norm": 1.5216220617294312,
+      "learning_rate": 9.894031278243266e-05,
+      "loss": 0.7744026780128479,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8440129449838187,
+      "grad_norm": 1.0934535264968872,
+      "learning_rate": 9.89352786525701e-05,
+      "loss": 0.7678086757659912,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8448759439050701,
+      "grad_norm": 1.223285436630249,
+      "learning_rate": 9.89302327221857e-05,
+      "loss": 0.7344630360603333,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8457389428263214,
+      "grad_norm": 1.1123322248458862,
+      "learning_rate": 9.892517499249623e-05,
+      "loss": 0.76033616065979,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8466019417475729,
+      "grad_norm": 1.0472193956375122,
+      "learning_rate": 9.89201054647214e-05,
+      "loss": 0.7176017761230469,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8474649406688242,
+      "grad_norm": 1.8538368940353394,
+      "learning_rate": 9.891502414008365e-05,
+      "loss": 0.7220107316970825,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8483279395900755,
+      "grad_norm": 1.2280668020248413,
+      "learning_rate": 9.890993101980834e-05,
+      "loss": 0.7399556636810303,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8491909385113269,
+      "grad_norm": 1.1482552289962769,
+      "learning_rate": 9.890482610512364e-05,
+      "loss": 0.7700466513633728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8500539374325782,
+      "grad_norm": 1.1877986192703247,
+      "learning_rate": 9.889970939726061e-05,
+      "loss": 0.6666419506072998,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8509169363538296,
+      "grad_norm": 1.407792329788208,
+      "learning_rate": 9.88945808974531e-05,
+      "loss": 0.7883278131484985,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8517799352750809,
+      "grad_norm": 1.1426323652267456,
+      "learning_rate": 9.888944060693781e-05,
+      "loss": 0.7441611289978027,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8526429341963323,
+      "grad_norm": 1.2660622596740723,
+      "learning_rate": 9.888428852695432e-05,
+      "loss": 0.6464171409606934,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8535059331175836,
+      "grad_norm": 1.1639022827148438,
+      "learning_rate": 9.887912465874502e-05,
+      "loss": 0.7242478728294373,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 1.2546422481536865,
+      "learning_rate": 9.887394900355516e-05,
+      "loss": 0.7431875467300415,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8552319309600863,
+      "grad_norm": 1.1057130098342896,
+      "learning_rate": 9.886876156263282e-05,
+      "loss": 0.7662370204925537,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8560949298813376,
+      "grad_norm": 1.1017125844955444,
+      "learning_rate": 9.886356233722894e-05,
+      "loss": 0.7120938897132874,
+      "step": 1984
+    },
+    {
+      "epoch": 0.856957928802589,
+      "grad_norm": 1.3548704385757446,
+      "learning_rate": 9.885835132859727e-05,
+      "loss": 0.789219856262207,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8578209277238403,
+      "grad_norm": 1.1920398473739624,
+      "learning_rate": 9.885312853799443e-05,
+      "loss": 0.7599530816078186,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8586839266450917,
+      "grad_norm": 1.170946478843689,
+      "learning_rate": 9.884789396667988e-05,
+      "loss": 0.8315809369087219,
+      "step": 1990
+    },
+    {
+      "epoch": 0.859546925566343,
+      "grad_norm": 1.2530180215835571,
+      "learning_rate": 9.88426476159159e-05,
+      "loss": 0.713994026184082,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8604099244875943,
+      "grad_norm": 1.1056885719299316,
+      "learning_rate": 9.883738948696765e-05,
+      "loss": 0.6656507849693298,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8612729234088458,
+      "grad_norm": 1.169382929801941,
+      "learning_rate": 9.883211958110308e-05,
+      "loss": 0.7164257764816284,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8621359223300971,
+      "grad_norm": 1.1970325708389282,
+      "learning_rate": 9.882683789959299e-05,
+      "loss": 0.6885558366775513,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "grad_norm": 1.1700761318206787,
+      "learning_rate": 9.882154444371107e-05,
+      "loss": 0.7265812754631042,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "eval_loss": 0.7487027645111084,
+      "eval_runtime": 825.141,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8638619201725998,
+      "grad_norm": 1.3120243549346924,
+      "learning_rate": 9.88162392147338e-05,
+      "loss": 0.7371920347213745,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8647249190938512,
+      "grad_norm": 1.098697304725647,
+      "learning_rate": 9.881092221394049e-05,
+      "loss": 0.7453275918960571,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8655879180151025,
+      "grad_norm": 1.1961677074432373,
+      "learning_rate": 9.880559344261334e-05,
+      "loss": 0.7562432289123535,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8664509169363538,
+      "grad_norm": 1.1655291318893433,
+      "learning_rate": 9.880025290203735e-05,
+      "loss": 0.7654905319213867,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8673139158576052,
+      "grad_norm": 1.0979351997375488,
+      "learning_rate": 9.879490059350036e-05,
+      "loss": 0.6807888746261597,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8681769147788565,
+      "grad_norm": 1.0378475189208984,
+      "learning_rate": 9.878953651829306e-05,
+      "loss": 0.6887364387512207,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8690399137001079,
+      "grad_norm": 2.0228428840637207,
+      "learning_rate": 9.878416067770898e-05,
+      "loss": 0.718578577041626,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8699029126213592,
+      "grad_norm": 1.1610580682754517,
+      "learning_rate": 9.877877307304446e-05,
+      "loss": 0.7424600720405579,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8707659115426106,
+      "grad_norm": 1.1415833234786987,
+      "learning_rate": 9.877337370559873e-05,
+      "loss": 0.7450630068778992,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8716289104638619,
+      "grad_norm": 1.1008672714233398,
+      "learning_rate": 9.876796257667379e-05,
+      "loss": 0.7398725748062134,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8724919093851132,
+      "grad_norm": 1.1662763357162476,
+      "learning_rate": 9.876253968757452e-05,
+      "loss": 0.7122271060943604,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8733549083063646,
+      "grad_norm": 1.1835166215896606,
+      "learning_rate": 9.875710503960862e-05,
+      "loss": 0.7026485800743103,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8742179072276159,
+      "grad_norm": 1.1602214574813843,
+      "learning_rate": 9.875165863408666e-05,
+      "loss": 0.7211011052131653,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8750809061488674,
+      "grad_norm": 1.222589373588562,
+      "learning_rate": 9.874620047232198e-05,
+      "loss": 0.7565322518348694,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8759439050701187,
+      "grad_norm": 1.203661561012268,
+      "learning_rate": 9.87407305556308e-05,
+      "loss": 0.7173075675964355,
+      "step": 2030
+    },
+    {
+      "epoch": 0.87680690399137,
+      "grad_norm": 1.3316307067871094,
+      "learning_rate": 9.87352488853322e-05,
+      "loss": 0.820809006690979,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8776699029126214,
+      "grad_norm": 1.3071057796478271,
+      "learning_rate": 9.872975546274799e-05,
+      "loss": 0.8689352869987488,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8785329018338727,
+      "grad_norm": 1.3557701110839844,
+      "learning_rate": 9.872425028920294e-05,
+      "loss": 0.7090175151824951,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8793959007551241,
+      "grad_norm": 1.2234264612197876,
+      "learning_rate": 9.871873336602459e-05,
+      "loss": 0.6883202195167542,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8802588996763754,
+      "grad_norm": 1.3064714670181274,
+      "learning_rate": 9.87132046945433e-05,
+      "loss": 0.7134175896644592,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8811218985976268,
+      "grad_norm": 1.2502511739730835,
+      "learning_rate": 9.870766427609232e-05,
+      "loss": 0.7127227783203125,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8819848975188781,
+      "grad_norm": 1.1354259252548218,
+      "learning_rate": 9.870211211200766e-05,
+      "loss": 0.719437301158905,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8828478964401294,
+      "grad_norm": 1.1706442832946777,
+      "learning_rate": 9.869654820362821e-05,
+      "loss": 0.6933091282844543,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8837108953613808,
+      "grad_norm": 1.0227864980697632,
+      "learning_rate": 9.86909725522957e-05,
+      "loss": 0.6825079321861267,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8845738942826321,
+      "grad_norm": 1.103418231010437,
+      "learning_rate": 9.868538515935465e-05,
+      "loss": 0.7306133508682251,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8854368932038835,
+      "grad_norm": 1.0938395261764526,
+      "learning_rate": 9.867978602615245e-05,
+      "loss": 0.6956495046615601,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8862998921251348,
+      "grad_norm": 1.115692377090454,
+      "learning_rate": 9.86741751540393e-05,
+      "loss": 0.7740456461906433,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8871628910463862,
+      "grad_norm": 1.0264382362365723,
+      "learning_rate": 9.866855254436824e-05,
+      "loss": 0.6996002197265625,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8880258899676375,
+      "grad_norm": 1.2523163557052612,
+      "learning_rate": 9.866291819849511e-05,
+      "loss": 0.760357677936554,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.0595446825027466,
+      "learning_rate": 9.865727211777864e-05,
+      "loss": 0.6389813423156738,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8897518878101403,
+      "grad_norm": 1.198530912399292,
+      "learning_rate": 9.865161430358036e-05,
+      "loss": 0.7859060168266296,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8906148867313916,
+      "grad_norm": 1.0271193981170654,
+      "learning_rate": 9.864594475726461e-05,
+      "loss": 0.7120063900947571,
+      "step": 2064
+    },
+    {
+      "epoch": 0.891477885652643,
+      "grad_norm": 1.1619046926498413,
+      "learning_rate": 9.864026348019857e-05,
+      "loss": 0.6751878261566162,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8923408845738943,
+      "grad_norm": 1.2812464237213135,
+      "learning_rate": 9.863457047375227e-05,
+      "loss": 0.610360860824585,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 1.112628698348999,
+      "learning_rate": 9.862886573929855e-05,
+      "loss": 0.6540104746818542,
+      "step": 2070
+    },
+    {
+      "epoch": 0.894066882416397,
+      "grad_norm": 1.6257048845291138,
+      "learning_rate": 9.862314927821306e-05,
+      "loss": 0.726277232170105,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8949298813376483,
+      "grad_norm": 1.227334976196289,
+      "learning_rate": 9.861742109187432e-05,
+      "loss": 0.7798765301704407,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8957928802588997,
+      "grad_norm": 1.0728148221969604,
+      "learning_rate": 9.861168118166366e-05,
+      "loss": 0.637127161026001,
+      "step": 2076
+    },
+    {
+      "epoch": 0.896655879180151,
+      "grad_norm": 1.220282793045044,
+      "learning_rate": 9.860592954896523e-05,
+      "loss": 0.7332831025123596,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8975188781014024,
+      "grad_norm": 1.2287596464157104,
+      "learning_rate": 9.8600166195166e-05,
+      "loss": 0.7337175607681274,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8983818770226537,
+      "grad_norm": 1.0701712369918823,
+      "learning_rate": 9.859439112165579e-05,
+      "loss": 0.7422571182250977,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8992448759439051,
+      "grad_norm": 1.1583118438720703,
+      "learning_rate": 9.858860432982721e-05,
+      "loss": 0.7010003328323364,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9001078748651564,
+      "grad_norm": 0.9644797444343567,
+      "learning_rate": 9.858280582107575e-05,
+      "loss": 0.6211312413215637,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9009708737864077,
+      "grad_norm": 1.2448554039001465,
+      "learning_rate": 9.857699559679968e-05,
+      "loss": 0.7299860715866089,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9018338727076591,
+      "grad_norm": 1.2735580205917358,
+      "learning_rate": 9.85711736584001e-05,
+      "loss": 0.7461655735969543,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9026968716289104,
+      "grad_norm": 1.36841881275177,
+      "learning_rate": 9.856534000728093e-05,
+      "loss": 0.7218145132064819,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9035598705501618,
+      "grad_norm": 1.1471478939056396,
+      "learning_rate": 9.855949464484898e-05,
+      "loss": 0.7045709490776062,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9044228694714131,
+      "grad_norm": 1.058254361152649,
+      "learning_rate": 9.85536375725138e-05,
+      "loss": 0.7189744710922241,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9052858683926646,
+      "grad_norm": 1.4070709943771362,
+      "learning_rate": 9.854776879168779e-05,
+      "loss": 0.740775465965271,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 1.2182649374008179,
+      "learning_rate": 9.854188830378622e-05,
+      "loss": 0.8191262483596802,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "eval_loss": 0.7449111342430115,
+      "eval_runtime": 832.5135,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9070118662351672,
+      "grad_norm": 1.139978289604187,
+      "learning_rate": 9.853599611022708e-05,
+      "loss": 0.7295873761177063,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9078748651564186,
+      "grad_norm": 1.1802685260772705,
+      "learning_rate": 9.853009221243127e-05,
+      "loss": 0.7141753435134888,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9087378640776699,
+      "grad_norm": 1.076759934425354,
+      "learning_rate": 9.852417661182253e-05,
+      "loss": 0.7557607293128967,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9096008629989213,
+      "grad_norm": 1.1000943183898926,
+      "learning_rate": 9.851824930982733e-05,
+      "loss": 0.7025325298309326,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9104638619201726,
+      "grad_norm": 1.0235389471054077,
+      "learning_rate": 9.851231030787504e-05,
+      "loss": 0.6967638731002808,
+      "step": 2110
+    },
+    {
+      "epoch": 0.911326860841424,
+      "grad_norm": 1.2311428785324097,
+      "learning_rate": 9.850635960739782e-05,
+      "loss": 0.7890311479568481,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9121898597626753,
+      "grad_norm": 1.173770785331726,
+      "learning_rate": 9.850039720983064e-05,
+      "loss": 0.6824729442596436,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9130528586839266,
+      "grad_norm": 1.2165251970291138,
+      "learning_rate": 9.849442311661131e-05,
+      "loss": 0.7157722115516663,
+      "step": 2116
+    },
+    {
+      "epoch": 0.913915857605178,
+      "grad_norm": 1.251520037651062,
+      "learning_rate": 9.848843732918047e-05,
+      "loss": 0.7164111733436584,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9147788565264293,
+      "grad_norm": 1.1330616474151611,
+      "learning_rate": 9.848243984898153e-05,
+      "loss": 0.6830796003341675,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9156418554476807,
+      "grad_norm": 1.1052451133728027,
+      "learning_rate": 9.847643067746081e-05,
+      "loss": 0.7149171829223633,
+      "step": 2122
+    },
+    {
+      "epoch": 0.916504854368932,
+      "grad_norm": 1.0389703512191772,
+      "learning_rate": 9.847040981606735e-05,
+      "loss": 0.7213287949562073,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9173678532901833,
+      "grad_norm": 1.151895523071289,
+      "learning_rate": 9.846437726625308e-05,
+      "loss": 0.703312873840332,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9182308522114347,
+      "grad_norm": 1.1624029874801636,
+      "learning_rate": 9.845833302947269e-05,
+      "loss": 0.8059951066970825,
+      "step": 2128
+    },
+    {
+      "epoch": 0.919093851132686,
+      "grad_norm": 1.2247511148452759,
+      "learning_rate": 9.845227710718377e-05,
+      "loss": 0.735476553440094,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9199568500539375,
+      "grad_norm": 1.1649967432022095,
+      "learning_rate": 9.844620950084661e-05,
+      "loss": 0.8130571842193604,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9208198489751888,
+      "grad_norm": 1.1915717124938965,
+      "learning_rate": 9.844013021192447e-05,
+      "loss": 0.6908681392669678,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9216828478964402,
+      "grad_norm": 1.211199164390564,
+      "learning_rate": 9.843403924188328e-05,
+      "loss": 0.652173638343811,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9225458468176915,
+      "grad_norm": 1.267242193222046,
+      "learning_rate": 9.842793659219186e-05,
+      "loss": 0.6672233939170837,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9234088457389428,
+      "grad_norm": 1.7038761377334595,
+      "learning_rate": 9.842182226432186e-05,
+      "loss": 0.7274143695831299,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9242718446601942,
+      "grad_norm": 1.151426076889038,
+      "learning_rate": 9.841569625974769e-05,
+      "loss": 0.707998514175415,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9251348435814455,
+      "grad_norm": 1.0414749383926392,
+      "learning_rate": 9.840955857994663e-05,
+      "loss": 0.6639942526817322,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9259978425026969,
+      "grad_norm": 1.1045551300048828,
+      "learning_rate": 9.840340922639875e-05,
+      "loss": 0.7096070647239685,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9268608414239482,
+      "grad_norm": 1.1193996667861938,
+      "learning_rate": 9.839724820058695e-05,
+      "loss": 0.71246737241745,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9277238403451996,
+      "grad_norm": 1.2609117031097412,
+      "learning_rate": 9.839107550399692e-05,
+      "loss": 0.7351052761077881,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9285868392664509,
+      "grad_norm": 1.1127749681472778,
+      "learning_rate": 9.838489113811717e-05,
+      "loss": 0.7125008702278137,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9294498381877022,
+      "grad_norm": 1.4771665334701538,
+      "learning_rate": 9.837869510443905e-05,
+      "loss": 0.7891085743904114,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9303128371089536,
+      "grad_norm": 1.0701470375061035,
+      "learning_rate": 9.83724874044567e-05,
+      "loss": 0.7370918393135071,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9311758360302049,
+      "grad_norm": 1.2045820951461792,
+      "learning_rate": 9.836626803966708e-05,
+      "loss": 0.7543193697929382,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.1944687366485596,
+      "learning_rate": 9.836003701156996e-05,
+      "loss": 0.6903626322746277,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9329018338727076,
+      "grad_norm": 1.1035394668579102,
+      "learning_rate": 9.83537943216679e-05,
+      "loss": 0.6923050880432129,
+      "step": 2162
+    },
+    {
+      "epoch": 0.933764832793959,
+      "grad_norm": 1.3766103982925415,
+      "learning_rate": 9.834753997146633e-05,
+      "loss": 0.8099993467330933,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9346278317152104,
+      "grad_norm": 1.0874576568603516,
+      "learning_rate": 9.834127396247346e-05,
+      "loss": 0.7896219491958618,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9354908306364617,
+      "grad_norm": 1.1633459329605103,
+      "learning_rate": 9.833499629620028e-05,
+      "loss": 0.7659367918968201,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9363538295577131,
+      "grad_norm": 1.405423641204834,
+      "learning_rate": 9.832870697416066e-05,
+      "loss": 0.7474715709686279,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9372168284789644,
+      "grad_norm": 1.1794270277023315,
+      "learning_rate": 9.83224059978712e-05,
+      "loss": 0.7018750905990601,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9380798274002158,
+      "grad_norm": 1.082466959953308,
+      "learning_rate": 9.831609336885139e-05,
+      "loss": 0.6477128267288208,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9389428263214671,
+      "grad_norm": 0.9830297827720642,
+      "learning_rate": 9.830976908862346e-05,
+      "loss": 0.6516029238700867,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9398058252427185,
+      "grad_norm": 1.112289309501648,
+      "learning_rate": 9.83034331587125e-05,
+      "loss": 0.6793080568313599,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9406688241639698,
+      "grad_norm": 1.1212154626846313,
+      "learning_rate": 9.829708558064638e-05,
+      "loss": 0.7220560312271118,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9415318230852211,
+      "grad_norm": 1.246471643447876,
+      "learning_rate": 9.829072635595582e-05,
+      "loss": 0.7621577978134155,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9423948220064725,
+      "grad_norm": 1.1732288599014282,
+      "learning_rate": 9.828435548617428e-05,
+      "loss": 0.6988050937652588,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9432578209277238,
+      "grad_norm": 1.1767359972000122,
+      "learning_rate": 9.827797297283808e-05,
+      "loss": 0.7909345626831055,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9441208198489752,
+      "grad_norm": 1.007215976715088,
+      "learning_rate": 9.827157881748633e-05,
+      "loss": 0.6651493310928345,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9449838187702265,
+      "grad_norm": 1.184118628501892,
+      "learning_rate": 9.826517302166096e-05,
+      "loss": 0.7742198705673218,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9458468176914779,
+      "grad_norm": 1.1227030754089355,
+      "learning_rate": 9.82587555869067e-05,
+      "loss": 0.7614272236824036,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9467098166127292,
+      "grad_norm": 1.1030738353729248,
+      "learning_rate": 9.825232651477109e-05,
+      "loss": 0.7302108407020569,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9475728155339805,
+      "grad_norm": 1.5209563970565796,
+      "learning_rate": 9.824588580680445e-05,
+      "loss": 0.7408936023712158,
+      "step": 2196
+    },
+    {
+      "epoch": 0.948435814455232,
+      "grad_norm": 1.1853431463241577,
+      "learning_rate": 9.823943346455995e-05,
+      "loss": 0.8050196766853333,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "grad_norm": 1.285551905632019,
+      "learning_rate": 9.823296948959354e-05,
+      "loss": 0.726731538772583,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "eval_loss": 0.7400083541870117,
+      "eval_runtime": 838.7633,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9501618122977347,
+      "grad_norm": 1.2159804105758667,
+      "learning_rate": 9.822649388346396e-05,
+      "loss": 0.7061861753463745,
+      "step": 2202
+    },
+    {
+      "epoch": 0.951024811218986,
+      "grad_norm": 1.1189815998077393,
+      "learning_rate": 9.822000664773278e-05,
+      "loss": 0.824293851852417,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9518878101402374,
+      "grad_norm": 0.9968506097793579,
+      "learning_rate": 9.821350778396439e-05,
+      "loss": 0.7230689525604248,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9527508090614887,
+      "grad_norm": 1.101393222808838,
+      "learning_rate": 9.820699729372593e-05,
+      "loss": 0.6633790135383606,
+      "step": 2208
+    },
+    {
+      "epoch": 0.95361380798274,
+      "grad_norm": 1.084307312965393,
+      "learning_rate": 9.820047517858741e-05,
+      "loss": 0.6548227071762085,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9544768069039914,
+      "grad_norm": 1.1291496753692627,
+      "learning_rate": 9.819394144012158e-05,
+      "loss": 0.6434295177459717,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9553398058252427,
+      "grad_norm": 2.303483724594116,
+      "learning_rate": 9.818739607990404e-05,
+      "loss": 0.8122013211250305,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9562028047464941,
+      "grad_norm": 1.244541883468628,
+      "learning_rate": 9.818083909951315e-05,
+      "loss": 0.6120955348014832,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9570658036677454,
+      "grad_norm": 1.1051502227783203,
+      "learning_rate": 9.817427050053012e-05,
+      "loss": 0.7569628953933716,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9579288025889967,
+      "grad_norm": 1.3161911964416504,
+      "learning_rate": 9.816769028453893e-05,
+      "loss": 0.7588980197906494,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9587918015102481,
+      "grad_norm": 1.2526928186416626,
+      "learning_rate": 9.816109845312636e-05,
+      "loss": 0.743370771408081,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9596548004314994,
+      "grad_norm": 1.196215033531189,
+      "learning_rate": 9.815449500788202e-05,
+      "loss": 0.80516517162323,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9605177993527508,
+      "grad_norm": 1.3204832077026367,
+      "learning_rate": 9.81478799503983e-05,
+      "loss": 0.7331632971763611,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9613807982740021,
+      "grad_norm": 1.2442790269851685,
+      "learning_rate": 9.814125328227035e-05,
+      "loss": 0.6810339093208313,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9622437971952535,
+      "grad_norm": 1.1739158630371094,
+      "learning_rate": 9.813461500509623e-05,
+      "loss": 0.7269608378410339,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9631067961165048,
+      "grad_norm": 1.1016241312026978,
+      "learning_rate": 9.812796512047666e-05,
+      "loss": 0.7159227132797241,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9639697950377561,
+      "grad_norm": 1.1654549837112427,
+      "learning_rate": 9.812130363001529e-05,
+      "loss": 0.7145758867263794,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9648327939590076,
+      "grad_norm": 1.0521938800811768,
+      "learning_rate": 9.811463053531846e-05,
+      "loss": 0.6873835325241089,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9656957928802589,
+      "grad_norm": 1.2285776138305664,
+      "learning_rate": 9.810794583799538e-05,
+      "loss": 0.6683027148246765,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9665587918015103,
+      "grad_norm": 0.9826368689537048,
+      "learning_rate": 9.810124953965805e-05,
+      "loss": 0.663141131401062,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9674217907227616,
+      "grad_norm": 1.0102336406707764,
+      "learning_rate": 9.809454164192123e-05,
+      "loss": 0.6688778400421143,
+      "step": 2242
+    },
+    {
+      "epoch": 0.968284789644013,
+      "grad_norm": 0.9781575202941895,
+      "learning_rate": 9.808782214640251e-05,
+      "loss": 0.671240508556366,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9691477885652643,
+      "grad_norm": 1.137385368347168,
+      "learning_rate": 9.808109105472225e-05,
+      "loss": 0.7673252820968628,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9700107874865156,
+      "grad_norm": 1.3359450101852417,
+      "learning_rate": 9.807434836850362e-05,
+      "loss": 0.8085554838180542,
+      "step": 2248
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 1.150694727897644,
+      "learning_rate": 9.806759408937262e-05,
+      "loss": 0.6750202178955078,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9717367853290183,
+      "grad_norm": 1.080316424369812,
+      "learning_rate": 9.8060828218958e-05,
+      "loss": 0.5928418040275574,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9725997842502697,
+      "grad_norm": 1.0489610433578491,
+      "learning_rate": 9.805405075889129e-05,
+      "loss": 0.6947471499443054,
+      "step": 2254
+    },
+    {
+      "epoch": 0.973462783171521,
+      "grad_norm": 1.199119210243225,
+      "learning_rate": 9.80472617108069e-05,
+      "loss": 0.7273924350738525,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9743257820927724,
+      "grad_norm": 0.9830896258354187,
+      "learning_rate": 9.804046107634193e-05,
+      "loss": 0.668613612651825,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9751887810140237,
+      "grad_norm": 1.1655877828598022,
+      "learning_rate": 9.803364885713633e-05,
+      "loss": 0.7255772352218628,
+      "step": 2260
+    },
+    {
+      "epoch": 0.976051779935275,
+      "grad_norm": 1.0289462804794312,
+      "learning_rate": 9.802682505483284e-05,
+      "loss": 0.6865858435630798,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9769147788565264,
+      "grad_norm": 1.1020126342773438,
+      "learning_rate": 9.801998967107702e-05,
+      "loss": 0.6680663228034973,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 1.0846027135849,
+      "learning_rate": 9.801314270751716e-05,
+      "loss": 0.7476468086242676,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9786407766990292,
+      "grad_norm": 1.1834278106689453,
+      "learning_rate": 9.800628416580437e-05,
+      "loss": 0.7173228859901428,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9795037756202805,
+      "grad_norm": 1.0082321166992188,
+      "learning_rate": 9.799941404759257e-05,
+      "loss": 0.6768618226051331,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9803667745415319,
+      "grad_norm": 1.1573102474212646,
+      "learning_rate": 9.799253235453846e-05,
+      "loss": 0.7877311110496521,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9812297734627832,
+      "grad_norm": 1.2030092477798462,
+      "learning_rate": 9.798563908830152e-05,
+      "loss": 0.6843966841697693,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9820927723840345,
+      "grad_norm": 1.080560564994812,
+      "learning_rate": 9.797873425054404e-05,
+      "loss": 0.6084349155426025,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9829557713052859,
+      "grad_norm": 1.1488033533096313,
+      "learning_rate": 9.797181784293107e-05,
+      "loss": 0.6784868836402893,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9838187702265372,
+      "grad_norm": 1.144773006439209,
+      "learning_rate": 9.796488986713049e-05,
+      "loss": 0.6111500263214111,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9846817691477886,
+      "grad_norm": 1.1356254816055298,
+      "learning_rate": 9.795795032481297e-05,
+      "loss": 0.7634603977203369,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9855447680690399,
+      "grad_norm": 1.139446496963501,
+      "learning_rate": 9.79509992176519e-05,
+      "loss": 0.6470903158187866,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9864077669902913,
+      "grad_norm": 1.1564133167266846,
+      "learning_rate": 9.794403654732354e-05,
+      "loss": 0.71474289894104,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9872707659115426,
+      "grad_norm": 1.0605655908584595,
+      "learning_rate": 9.793706231550689e-05,
+      "loss": 0.6413968801498413,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9881337648327939,
+      "grad_norm": 1.1660863161087036,
+      "learning_rate": 9.793007652388378e-05,
+      "loss": 0.6595705151557922,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9889967637540453,
+      "grad_norm": 1.1661914587020874,
+      "learning_rate": 9.792307917413877e-05,
+      "loss": 0.7144066095352173,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9898597626752966,
+      "grad_norm": 1.2250727415084839,
+      "learning_rate": 9.791607026795926e-05,
+      "loss": 0.7956018447875977,
+      "step": 2294
+    },
+    {
+      "epoch": 0.990722761596548,
+      "grad_norm": 0.9593650698661804,
+      "learning_rate": 9.790904980703541e-05,
+      "loss": 0.5786031484603882,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9915857605177993,
+      "grad_norm": 1.0697777271270752,
+      "learning_rate": 9.790201779306018e-05,
+      "loss": 0.6817460656166077,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "grad_norm": 1.2658408880233765,
+      "learning_rate": 9.789497422772931e-05,
+      "loss": 0.6845148801803589,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "eval_loss": 0.7360134720802307,
+      "eval_runtime": 825.8559,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2300
+    },
+    {
+      "epoch": 0.993311758360302,
+      "grad_norm": 1.0869544744491577,
+      "learning_rate": 9.788791911274131e-05,
+      "loss": 0.6689707040786743,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9941747572815534,
+      "grad_norm": 1.0049875974655151,
+      "learning_rate": 9.788085244979748e-05,
+      "loss": 0.6470839381217957,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9950377562028048,
+      "grad_norm": 1.4471046924591064,
+      "learning_rate": 9.787377424060196e-05,
+      "loss": 0.8018016815185547,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9959007551240561,
+      "grad_norm": 1.3750648498535156,
+      "learning_rate": 9.786668448686159e-05,
+      "loss": 0.8078322410583496,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9967637540453075,
+      "grad_norm": 1.098283290863037,
+      "learning_rate": 9.7859583190286e-05,
+      "loss": 0.6740908622741699,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9976267529665588,
+      "grad_norm": 1.0312142372131348,
+      "learning_rate": 9.785247035258773e-05,
+      "loss": 0.6844386458396912,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9984897518878102,
+      "grad_norm": 1.4482319355010986,
+      "learning_rate": 9.784534597548193e-05,
+      "loss": 0.7547479271888733,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9993527508090615,
+      "grad_norm": 1.0897735357284546,
+      "learning_rate": 9.783821006068663e-05,
+      "loss": 0.7713987231254578,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7887805700302124,
+      "learning_rate": 9.783106260992264e-05,
+      "loss": 0.6868399381637573,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0008629989212514,
+      "grad_norm": 1.0021663904190063,
+      "learning_rate": 9.782390362491353e-05,
+      "loss": 0.696355938911438,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0017259978425026,
+      "grad_norm": 1.3650091886520386,
+      "learning_rate": 9.781673310738564e-05,
+      "loss": 0.7315902709960938,
+      "step": 2322
+    },
+    {
+      "epoch": 1.002588996763754,
+      "grad_norm": 1.2247875928878784,
+      "learning_rate": 9.780955105906813e-05,
+      "loss": 0.7667547464370728,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0034519956850054,
+      "grad_norm": 1.1917641162872314,
+      "learning_rate": 9.780235748169291e-05,
+      "loss": 0.7380582094192505,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0043149946062568,
+      "grad_norm": 1.146159052848816,
+      "learning_rate": 9.779515237699466e-05,
+      "loss": 0.6780733466148376,
+      "step": 2328
+    },
+    {
+      "epoch": 1.005177993527508,
+      "grad_norm": 1.4877620935440063,
+      "learning_rate": 9.778793574671089e-05,
+      "loss": 0.8116409182548523,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0060409924487594,
+      "grad_norm": 1.3488554954528809,
+      "learning_rate": 9.778070759258185e-05,
+      "loss": 0.65245521068573,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0069039913700109,
+      "grad_norm": 1.096686840057373,
+      "learning_rate": 9.777346791635056e-05,
+      "loss": 0.584334135055542,
+      "step": 2334
+    },
+    {
+      "epoch": 1.007766990291262,
+      "grad_norm": 1.063981056213379,
+      "learning_rate": 9.776621671976285e-05,
+      "loss": 0.6594714522361755,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0086299892125135,
+      "grad_norm": 1.2811797857284546,
+      "learning_rate": 9.775895400456729e-05,
+      "loss": 0.6963099837303162,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0094929881337649,
+      "grad_norm": 0.9491971731185913,
+      "learning_rate": 9.775167977251529e-05,
+      "loss": 0.6246375441551208,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0103559870550163,
+      "grad_norm": 1.198683500289917,
+      "learning_rate": 9.774439402536098e-05,
+      "loss": 0.6352161169052124,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0112189859762675,
+      "grad_norm": 1.2178813219070435,
+      "learning_rate": 9.773709676486129e-05,
+      "loss": 0.6884894967079163,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0120819848975189,
+      "grad_norm": 1.1905319690704346,
+      "learning_rate": 9.77297879927759e-05,
+      "loss": 0.5945048928260803,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0129449838187703,
+      "grad_norm": 1.210302710533142,
+      "learning_rate": 9.77224677108673e-05,
+      "loss": 0.6533923149108887,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0138079827400215,
+      "grad_norm": 1.15508234500885,
+      "learning_rate": 9.771513592090076e-05,
+      "loss": 0.6619405746459961,
+      "step": 2350
+    },
+    {
+      "epoch": 1.014670981661273,
+      "grad_norm": 1.0833762884140015,
+      "learning_rate": 9.77077926246443e-05,
+      "loss": 0.6581310033798218,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0155339805825243,
+      "grad_norm": 1.0330514907836914,
+      "learning_rate": 9.770043782386869e-05,
+      "loss": 0.6670945882797241,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0163969795037757,
+      "grad_norm": 1.1472195386886597,
+      "learning_rate": 9.769307152034755e-05,
+      "loss": 0.6801814436912537,
+      "step": 2356
+    },
+    {
+      "epoch": 1.017259978425027,
+      "grad_norm": 1.2204132080078125,
+      "learning_rate": 9.76856937158572e-05,
+      "loss": 0.7206465005874634,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0181229773462783,
+      "grad_norm": 1.2335962057113647,
+      "learning_rate": 9.76783044121768e-05,
+      "loss": 0.6884574294090271,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0189859762675297,
+      "grad_norm": 1.0858685970306396,
+      "learning_rate": 9.767090361108822e-05,
+      "loss": 0.655097484588623,
+      "step": 2362
+    },
+    {
+      "epoch": 1.019848975188781,
+      "grad_norm": 1.117907166481018,
+      "learning_rate": 9.766349131437614e-05,
+      "loss": 0.6478561758995056,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0207119741100323,
+      "grad_norm": 1.1995056867599487,
+      "learning_rate": 9.765606752382797e-05,
+      "loss": 0.6310951113700867,
+      "step": 2366
+    },
+    {
+      "epoch": 1.0215749730312838,
+      "grad_norm": 1.354819655418396,
+      "learning_rate": 9.764863224123397e-05,
+      "loss": 0.6373490691184998,
+      "step": 2368
+    },
+    {
+      "epoch": 1.0224379719525352,
+      "grad_norm": 1.2194819450378418,
+      "learning_rate": 9.764118546838712e-05,
+      "loss": 0.6595761179924011,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0233009708737864,
+      "grad_norm": 1.1485158205032349,
+      "learning_rate": 9.763372720708314e-05,
+      "loss": 0.6299480199813843,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0241639697950378,
+      "grad_norm": 1.1153662204742432,
+      "learning_rate": 9.76262574591206e-05,
+      "loss": 0.6788742542266846,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0250269687162892,
+      "grad_norm": 1.0690817832946777,
+      "learning_rate": 9.761877622630075e-05,
+      "loss": 0.559056282043457,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0258899676375404,
+      "grad_norm": 1.1896858215332031,
+      "learning_rate": 9.76112835104277e-05,
+      "loss": 0.7164555191993713,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0267529665587918,
+      "grad_norm": 1.1792134046554565,
+      "learning_rate": 9.760377931330824e-05,
+      "loss": 0.6791456937789917,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0276159654800432,
+      "grad_norm": 1.361763834953308,
+      "learning_rate": 9.759626363675202e-05,
+      "loss": 0.7421526312828064,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0284789644012946,
+      "grad_norm": 1.1960662603378296,
+      "learning_rate": 9.758873648257138e-05,
+      "loss": 0.7032517194747925,
+      "step": 2384
+    },
+    {
+      "epoch": 1.0293419633225458,
+      "grad_norm": 1.323255181312561,
+      "learning_rate": 9.758119785258146e-05,
+      "loss": 0.6953344345092773,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0302049622437972,
+      "grad_norm": 1.2343937158584595,
+      "learning_rate": 9.75736477486002e-05,
+      "loss": 0.7287829518318176,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0310679611650486,
+      "grad_norm": 1.0712062120437622,
+      "learning_rate": 9.756608617244823e-05,
+      "loss": 0.6812928915023804,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0319309600862998,
+      "grad_norm": 1.264726996421814,
+      "learning_rate": 9.755851312594902e-05,
+      "loss": 0.7173963785171509,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0327939590075512,
+      "grad_norm": 1.1563389301300049,
+      "learning_rate": 9.755092861092874e-05,
+      "loss": 0.6509836316108704,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0336569579288026,
+      "grad_norm": 1.326188087463379,
+      "learning_rate": 9.75433326292164e-05,
+      "loss": 0.718437135219574,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0345199568500538,
+      "grad_norm": 1.2263392210006714,
+      "learning_rate": 9.753572518264373e-05,
+      "loss": 0.664516031742096,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "grad_norm": 1.2874478101730347,
+      "learning_rate": 9.752810627304523e-05,
+      "loss": 0.6754546165466309,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "eval_loss": 0.7346783876419067,
+      "eval_runtime": 825.8356,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0362459546925566,
+      "grad_norm": 1.304518461227417,
+      "learning_rate": 9.752047590225814e-05,
+      "loss": 0.708167314529419,
+      "step": 2402
+    },
+    {
+      "epoch": 1.037108953613808,
+      "grad_norm": 1.21835196018219,
+      "learning_rate": 9.751283407212252e-05,
+      "loss": 0.7213018536567688,
+      "step": 2404
+    },
+    {
+      "epoch": 1.0379719525350593,
+      "grad_norm": 1.2947579622268677,
+      "learning_rate": 9.750518078448116e-05,
+      "loss": 0.6739231944084167,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 1.138368010520935,
+      "learning_rate": 9.749751604117959e-05,
+      "loss": 0.7671371102333069,
+      "step": 2408
+    },
+    {
+      "epoch": 1.039697950377562,
+      "grad_norm": 1.1255316734313965,
+      "learning_rate": 9.748983984406616e-05,
+      "loss": 0.6593353748321533,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0405609492988135,
+      "grad_norm": 1.2362794876098633,
+      "learning_rate": 9.748215219499193e-05,
+      "loss": 0.713056743144989,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0414239482200647,
+      "grad_norm": 1.2021304368972778,
+      "learning_rate": 9.747445309581076e-05,
+      "loss": 0.6989394426345825,
+      "step": 2414
+    },
+    {
+      "epoch": 1.042286947141316,
+      "grad_norm": 1.127341389656067,
+      "learning_rate": 9.746674254837925e-05,
+      "loss": 0.623035192489624,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0431499460625675,
+      "grad_norm": 1.1254255771636963,
+      "learning_rate": 9.745902055455672e-05,
+      "loss": 0.6490310430526733,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0440129449838187,
+      "grad_norm": 1.2142512798309326,
+      "learning_rate": 9.745128711620535e-05,
+      "loss": 0.6210505366325378,
+      "step": 2420
+    },
+    {
+      "epoch": 1.04487594390507,
+      "grad_norm": 1.1306709051132202,
+      "learning_rate": 9.744354223519001e-05,
+      "loss": 0.6243326663970947,
+      "step": 2422
+    },
+    {
+      "epoch": 1.0457389428263215,
+      "grad_norm": 1.202621340751648,
+      "learning_rate": 9.743578591337833e-05,
+      "loss": 0.6312575340270996,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0466019417475727,
+      "grad_norm": 1.0556137561798096,
+      "learning_rate": 9.742801815264072e-05,
+      "loss": 0.6190714836120605,
+      "step": 2426
+    },
+    {
+      "epoch": 1.0474649406688241,
+      "grad_norm": 1.5071814060211182,
+      "learning_rate": 9.742023895485033e-05,
+      "loss": 0.7496671080589294,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0483279395900755,
+      "grad_norm": 1.1448185443878174,
+      "learning_rate": 9.74124483218831e-05,
+      "loss": 0.6288847327232361,
+      "step": 2430
+    },
+    {
+      "epoch": 1.049190938511327,
+      "grad_norm": 1.0072964429855347,
+      "learning_rate": 9.74046462556177e-05,
+      "loss": 0.6938424706459045,
+      "step": 2432
+    },
+    {
+      "epoch": 1.0500539374325781,
+      "grad_norm": 1.198836326599121,
+      "learning_rate": 9.739683275793554e-05,
+      "loss": 0.7529639601707458,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0509169363538295,
+      "grad_norm": 1.2170380353927612,
+      "learning_rate": 9.738900783072083e-05,
+      "loss": 0.7193960547447205,
+      "step": 2436
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 1.0384336709976196,
+      "learning_rate": 9.738117147586051e-05,
+      "loss": 0.6488150954246521,
+      "step": 2438
+    },
+    {
+      "epoch": 1.0526429341963321,
+      "grad_norm": 1.1562492847442627,
+      "learning_rate": 9.737332369524429e-05,
+      "loss": 0.6672247052192688,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0535059331175836,
+      "grad_norm": 1.1553802490234375,
+      "learning_rate": 9.736546449076462e-05,
+      "loss": 0.651012122631073,
+      "step": 2442
+    },
+    {
+      "epoch": 1.054368932038835,
+      "grad_norm": 1.136838674545288,
+      "learning_rate": 9.735759386431671e-05,
+      "loss": 0.681246817111969,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0552319309600864,
+      "grad_norm": 1.177869200706482,
+      "learning_rate": 9.734971181779854e-05,
+      "loss": 0.7331019639968872,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0560949298813376,
+      "grad_norm": 1.2983415126800537,
+      "learning_rate": 9.73418183531108e-05,
+      "loss": 0.7316935658454895,
+      "step": 2448
+    },
+    {
+      "epoch": 1.056957928802589,
+      "grad_norm": 1.1189959049224854,
+      "learning_rate": 9.733391347215699e-05,
+      "loss": 0.6513608694076538,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0578209277238404,
+      "grad_norm": 1.3256937265396118,
+      "learning_rate": 9.732599717684334e-05,
+      "loss": 0.6469123363494873,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0586839266450916,
+      "grad_norm": 0.9766954779624939,
+      "learning_rate": 9.731806946907881e-05,
+      "loss": 0.5860153436660767,
+      "step": 2454
+    },
+    {
+      "epoch": 1.059546925566343,
+      "grad_norm": 1.1360441446304321,
+      "learning_rate": 9.731013035077514e-05,
+      "loss": 0.6783872842788696,
+      "step": 2456
+    },
+    {
+      "epoch": 1.0604099244875944,
+      "grad_norm": 1.3483080863952637,
+      "learning_rate": 9.73021798238468e-05,
+      "loss": 0.7894490361213684,
+      "step": 2458
+    },
+    {
+      "epoch": 1.0612729234088458,
+      "grad_norm": 1.1412159204483032,
+      "learning_rate": 9.729421789021105e-05,
+      "loss": 0.7098361849784851,
+      "step": 2460
+    },
+    {
+      "epoch": 1.062135922330097,
+      "grad_norm": 1.2670512199401855,
+      "learning_rate": 9.728624455178785e-05,
+      "loss": 0.7138097286224365,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0629989212513484,
+      "grad_norm": 1.2577205896377563,
+      "learning_rate": 9.727825981049994e-05,
+      "loss": 0.7136340141296387,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0638619201725998,
+      "grad_norm": 1.098997712135315,
+      "learning_rate": 9.727026366827283e-05,
+      "loss": 0.6273087859153748,
+      "step": 2466
+    },
+    {
+      "epoch": 1.064724919093851,
+      "grad_norm": 1.1550003290176392,
+      "learning_rate": 9.72622561270347e-05,
+      "loss": 0.7076449394226074,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0655879180151024,
+      "grad_norm": 1.0861129760742188,
+      "learning_rate": 9.725423718871659e-05,
+      "loss": 0.6856985092163086,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0664509169363539,
+      "grad_norm": 1.2742570638656616,
+      "learning_rate": 9.724620685525219e-05,
+      "loss": 0.7311446070671082,
+      "step": 2472
+    },
+    {
+      "epoch": 1.0673139158576053,
+      "grad_norm": 1.504339337348938,
+      "learning_rate": 9.7238165128578e-05,
+      "loss": 0.7728472948074341,
+      "step": 2474
+    },
+    {
+      "epoch": 1.0681769147788565,
+      "grad_norm": 1.2232030630111694,
+      "learning_rate": 9.723011201063323e-05,
+      "loss": 0.6259000301361084,
+      "step": 2476
+    },
+    {
+      "epoch": 1.0690399137001079,
+      "grad_norm": 1.3238953351974487,
+      "learning_rate": 9.722204750335989e-05,
+      "loss": 0.7132798433303833,
+      "step": 2478
+    },
+    {
+      "epoch": 1.0699029126213593,
+      "grad_norm": 1.2286484241485596,
+      "learning_rate": 9.721397160870265e-05,
+      "loss": 0.6067986488342285,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0707659115426105,
+      "grad_norm": 1.3101580142974854,
+      "learning_rate": 9.7205884328609e-05,
+      "loss": 0.6884648203849792,
+      "step": 2482
+    },
+    {
+      "epoch": 1.0716289104638619,
+      "grad_norm": 1.437833309173584,
+      "learning_rate": 9.719778566502915e-05,
+      "loss": 0.7164831757545471,
+      "step": 2484
+    },
+    {
+      "epoch": 1.0724919093851133,
+      "grad_norm": 1.0980390310287476,
+      "learning_rate": 9.718967561991606e-05,
+      "loss": 0.6326627731323242,
+      "step": 2486
+    },
+    {
+      "epoch": 1.0733549083063647,
+      "grad_norm": 0.9919570088386536,
+      "learning_rate": 9.718155419522543e-05,
+      "loss": 0.5712647438049316,
+      "step": 2488
+    },
+    {
+      "epoch": 1.074217907227616,
+      "grad_norm": 1.0773793458938599,
+      "learning_rate": 9.717342139291569e-05,
+      "loss": 0.6923331022262573,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0750809061488673,
+      "grad_norm": 1.1553318500518799,
+      "learning_rate": 9.716527721494805e-05,
+      "loss": 0.6737485527992249,
+      "step": 2492
+    },
+    {
+      "epoch": 1.0759439050701187,
+      "grad_norm": 1.112865924835205,
+      "learning_rate": 9.715712166328644e-05,
+      "loss": 0.6422063708305359,
+      "step": 2494
+    },
+    {
+      "epoch": 1.07680690399137,
+      "grad_norm": 1.0651512145996094,
+      "learning_rate": 9.71489547398975e-05,
+      "loss": 0.6399799585342407,
+      "step": 2496
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 1.1394895315170288,
+      "learning_rate": 9.71407764467507e-05,
+      "loss": 0.6218312382698059,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "grad_norm": 1.3323278427124023,
+      "learning_rate": 9.713258678581815e-05,
+      "loss": 0.65947425365448,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "eval_loss": 0.7294915914535522,
+      "eval_runtime": 836.4643,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0793959007551241,
+      "grad_norm": 1.3122533559799194,
+      "learning_rate": 9.712438575907479e-05,
+      "loss": 0.66681307554245,
+      "step": 2502
+    },
+    {
+      "epoch": 1.0802588996763753,
+      "grad_norm": 1.3631974458694458,
+      "learning_rate": 9.711617336849824e-05,
+      "loss": 0.6926553249359131,
+      "step": 2504
+    },
+    {
+      "epoch": 1.0811218985976268,
+      "grad_norm": 1.2383692264556885,
+      "learning_rate": 9.710794961606886e-05,
+      "loss": 0.6706596612930298,
+      "step": 2506
+    },
+    {
+      "epoch": 1.0819848975188782,
+      "grad_norm": 1.1660504341125488,
+      "learning_rate": 9.709971450376982e-05,
+      "loss": 0.6011359095573425,
+      "step": 2508
+    },
+    {
+      "epoch": 1.0828478964401294,
+      "grad_norm": 1.3021246194839478,
+      "learning_rate": 9.709146803358693e-05,
+      "loss": 0.689009428024292,
+      "step": 2510
+    },
+    {
+      "epoch": 1.0837108953613808,
+      "grad_norm": 1.0553252696990967,
+      "learning_rate": 9.708321020750883e-05,
+      "loss": 0.6216740012168884,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0845738942826322,
+      "grad_norm": 1.1862177848815918,
+      "learning_rate": 9.707494102752683e-05,
+      "loss": 0.6439616680145264,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0854368932038836,
+      "grad_norm": 1.1166558265686035,
+      "learning_rate": 9.706666049563504e-05,
+      "loss": 0.6825361251831055,
+      "step": 2516
+    },
+    {
+      "epoch": 1.0862998921251348,
+      "grad_norm": 1.128812551498413,
+      "learning_rate": 9.705836861383022e-05,
+      "loss": 0.6471308469772339,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0871628910463862,
+      "grad_norm": 1.324802279472351,
+      "learning_rate": 9.705006538411197e-05,
+      "loss": 0.6535511016845703,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0880258899676376,
+      "grad_norm": 1.3586591482162476,
+      "learning_rate": 9.704175080848255e-05,
+      "loss": 0.678022563457489,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 1.1334232091903687,
+      "learning_rate": 9.703342488894697e-05,
+      "loss": 0.5904073119163513,
+      "step": 2524
+    },
+    {
+      "epoch": 1.0897518878101402,
+      "grad_norm": 1.182215929031372,
+      "learning_rate": 9.702508762751303e-05,
+      "loss": 0.6072601079940796,
+      "step": 2526
+    },
+    {
+      "epoch": 1.0906148867313916,
+      "grad_norm": 1.2525769472122192,
+      "learning_rate": 9.701673902619119e-05,
+      "loss": 0.6752429604530334,
+      "step": 2528
+    },
+    {
+      "epoch": 1.091477885652643,
+      "grad_norm": 1.2811026573181152,
+      "learning_rate": 9.70083790869947e-05,
+      "loss": 0.7192611694335938,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0923408845738942,
+      "grad_norm": 1.1434061527252197,
+      "learning_rate": 9.700000781193949e-05,
+      "loss": 0.6991751194000244,
+      "step": 2532
+    },
+    {
+      "epoch": 1.0932038834951456,
+      "grad_norm": 1.26658296585083,
+      "learning_rate": 9.699162520304428e-05,
+      "loss": 0.7010498046875,
+      "step": 2534
+    },
+    {
+      "epoch": 1.094066882416397,
+      "grad_norm": 1.2083529233932495,
+      "learning_rate": 9.698323126233048e-05,
+      "loss": 0.6945615410804749,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0949298813376482,
+      "grad_norm": 1.2064114809036255,
+      "learning_rate": 9.697482599182227e-05,
+      "loss": 0.689496636390686,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0957928802588996,
+      "grad_norm": 1.17113196849823,
+      "learning_rate": 9.696640939354655e-05,
+      "loss": 0.6966884136199951,
+      "step": 2540
+    },
+    {
+      "epoch": 1.096655879180151,
+      "grad_norm": 1.1551356315612793,
+      "learning_rate": 9.695798146953291e-05,
+      "loss": 0.6524666547775269,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0975188781014025,
+      "grad_norm": 1.0966521501541138,
+      "learning_rate": 9.694954222181374e-05,
+      "loss": 0.5769726037979126,
+      "step": 2544
+    },
+    {
+      "epoch": 1.0983818770226537,
+      "grad_norm": 1.320669174194336,
+      "learning_rate": 9.694109165242411e-05,
+      "loss": 0.6986333727836609,
+      "step": 2546
+    },
+    {
+      "epoch": 1.099244875943905,
+      "grad_norm": 1.1678471565246582,
+      "learning_rate": 9.693262976340183e-05,
+      "loss": 0.7014455199241638,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1001078748651565,
+      "grad_norm": 1.1527155637741089,
+      "learning_rate": 9.692415655678745e-05,
+      "loss": 0.5786730647087097,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1009708737864077,
+      "grad_norm": 1.2511217594146729,
+      "learning_rate": 9.691567203462428e-05,
+      "loss": 0.755945086479187,
+      "step": 2552
+    },
+    {
+      "epoch": 1.101833872707659,
+      "grad_norm": 1.0910221338272095,
+      "learning_rate": 9.690717619895828e-05,
+      "loss": 0.6487610936164856,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1026968716289105,
+      "grad_norm": 1.3682197332382202,
+      "learning_rate": 9.68986690518382e-05,
+      "loss": 0.6346665024757385,
+      "step": 2556
+    },
+    {
+      "epoch": 1.103559870550162,
+      "grad_norm": 1.2161564826965332,
+      "learning_rate": 9.68901505953155e-05,
+      "loss": 0.647144615650177,
+      "step": 2558
+    },
+    {
+      "epoch": 1.104422869471413,
+      "grad_norm": 1.2559088468551636,
+      "learning_rate": 9.688162083144436e-05,
+      "loss": 0.6461755633354187,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1052858683926645,
+      "grad_norm": 1.1260225772857666,
+      "learning_rate": 9.68730797622817e-05,
+      "loss": 0.6888763904571533,
+      "step": 2562
+    },
+    {
+      "epoch": 1.106148867313916,
+      "grad_norm": 1.1499868631362915,
+      "learning_rate": 9.686452738988716e-05,
+      "loss": 0.6318327784538269,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1070118662351671,
+      "grad_norm": 1.1775025129318237,
+      "learning_rate": 9.68559637163231e-05,
+      "loss": 0.6538340449333191,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1078748651564185,
+      "grad_norm": 1.264656901359558,
+      "learning_rate": 9.684738874365465e-05,
+      "loss": 0.6706726551055908,
+      "step": 2568
+    },
+    {
+      "epoch": 1.10873786407767,
+      "grad_norm": 1.1130270957946777,
+      "learning_rate": 9.683880247394956e-05,
+      "loss": 0.7573289275169373,
+      "step": 2570
+    },
+    {
+      "epoch": 1.1096008629989211,
+      "grad_norm": 1.2781217098236084,
+      "learning_rate": 9.683020490927841e-05,
+      "loss": 0.6603306531906128,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1104638619201725,
+      "grad_norm": 1.251062035560608,
+      "learning_rate": 9.682159605171446e-05,
+      "loss": 0.6523779630661011,
+      "step": 2574
+    },
+    {
+      "epoch": 1.111326860841424,
+      "grad_norm": 1.2631903886795044,
+      "learning_rate": 9.68129759033337e-05,
+      "loss": 0.677550733089447,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1121898597626754,
+      "grad_norm": 1.0212602615356445,
+      "learning_rate": 9.680434446621483e-05,
+      "loss": 0.666540801525116,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1130528586839266,
+      "grad_norm": 1.221718430519104,
+      "learning_rate": 9.679570174243929e-05,
+      "loss": 0.7467086315155029,
+      "step": 2580
+    },
+    {
+      "epoch": 1.113915857605178,
+      "grad_norm": 1.1175146102905273,
+      "learning_rate": 9.678704773409124e-05,
+      "loss": 0.673492968082428,
+      "step": 2582
+    },
+    {
+      "epoch": 1.1147788565264294,
+      "grad_norm": 1.1274988651275635,
+      "learning_rate": 9.677838244325754e-05,
+      "loss": 0.5724847912788391,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1156418554476808,
+      "grad_norm": 1.297654390335083,
+      "learning_rate": 9.676970587202779e-05,
+      "loss": 0.634985625743866,
+      "step": 2586
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 1.1908061504364014,
+      "learning_rate": 9.67610180224943e-05,
+      "loss": 0.6813890933990479,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1173678532901834,
+      "grad_norm": 1.1787118911743164,
+      "learning_rate": 9.675231889675212e-05,
+      "loss": 0.6738101243972778,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1182308522114348,
+      "grad_norm": 1.1327152252197266,
+      "learning_rate": 9.674360849689899e-05,
+      "loss": 0.6662938594818115,
+      "step": 2592
+    },
+    {
+      "epoch": 1.119093851132686,
+      "grad_norm": 1.1246299743652344,
+      "learning_rate": 9.67348868250354e-05,
+      "loss": 0.677423357963562,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1199568500539374,
+      "grad_norm": 1.1411702632904053,
+      "learning_rate": 9.672615388326454e-05,
+      "loss": 0.6257020235061646,
+      "step": 2596
+    },
+    {
+      "epoch": 1.1208198489751888,
+      "grad_norm": 1.1607346534729004,
+      "learning_rate": 9.671740967369231e-05,
+      "loss": 0.7029470205307007,
+      "step": 2598
+    },
+    {
+      "epoch": 1.12168284789644,
+      "grad_norm": 1.047934651374817,
+      "learning_rate": 9.670865419842735e-05,
+      "loss": 0.6040990352630615,
+      "step": 2600
+    },
+    {
+      "epoch": 1.12168284789644,
+      "eval_loss": 0.7287447452545166,
+      "eval_runtime": 831.8138,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2600
+    },
+    {
+      "epoch": 1.1225458468176914,
+      "grad_norm": 1.0571783781051636,
+      "learning_rate": 9.669988745958098e-05,
+      "loss": 0.6872879266738892,
+      "step": 2602
+    },
+    {
+      "epoch": 1.1234088457389428,
+      "grad_norm": 1.0940403938293457,
+      "learning_rate": 9.669110945926727e-05,
+      "loss": 0.6173275113105774,
+      "step": 2604
+    },
+    {
+      "epoch": 1.1242718446601943,
+      "grad_norm": 1.1180222034454346,
+      "learning_rate": 9.6682320199603e-05,
+      "loss": 0.6142664551734924,
+      "step": 2606
+    },
+    {
+      "epoch": 1.1251348435814454,
+      "grad_norm": 1.251098871231079,
+      "learning_rate": 9.667351968270768e-05,
+      "loss": 0.7225713133811951,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1259978425026969,
+      "grad_norm": 1.2333059310913086,
+      "learning_rate": 9.666470791070347e-05,
+      "loss": 0.6830481290817261,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1268608414239483,
+      "grad_norm": 1.2904123067855835,
+      "learning_rate": 9.665588488571531e-05,
+      "loss": 0.6835238337516785,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1277238403451997,
+      "grad_norm": 1.4779443740844727,
+      "learning_rate": 9.664705060987085e-05,
+      "loss": 0.6993544697761536,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1285868392664509,
+      "grad_norm": 1.267021656036377,
+      "learning_rate": 9.663820508530042e-05,
+      "loss": 0.6819394826889038,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1294498381877023,
+      "grad_norm": 1.0755046606063843,
+      "learning_rate": 9.662934831413708e-05,
+      "loss": 0.6200308799743652,
+      "step": 2618
+    },
+    {
+      "epoch": 1.1303128371089537,
+      "grad_norm": 1.2043033838272095,
+      "learning_rate": 9.66204802985166e-05,
+      "loss": 0.7138658165931702,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1311758360302049,
+      "grad_norm": 1.2641128301620483,
+      "learning_rate": 9.661160104057744e-05,
+      "loss": 0.7238332033157349,
+      "step": 2622
+    },
+    {
+      "epoch": 1.1320388349514563,
+      "grad_norm": 1.0422990322113037,
+      "learning_rate": 9.660271054246085e-05,
+      "loss": 0.6401007175445557,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1329018338727077,
+      "grad_norm": 1.2298130989074707,
+      "learning_rate": 9.659380880631068e-05,
+      "loss": 0.6305879950523376,
+      "step": 2626
+    },
+    {
+      "epoch": 1.133764832793959,
+      "grad_norm": 1.254573106765747,
+      "learning_rate": 9.658489583427357e-05,
+      "loss": 0.5751872062683105,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1346278317152103,
+      "grad_norm": 1.15679132938385,
+      "learning_rate": 9.657597162849883e-05,
+      "loss": 0.6629595756530762,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1354908306364617,
+      "grad_norm": 1.193629264831543,
+      "learning_rate": 9.656703619113849e-05,
+      "loss": 0.6832258105278015,
+      "step": 2632
+    },
+    {
+      "epoch": 1.1363538295577131,
+      "grad_norm": 1.2330398559570312,
+      "learning_rate": 9.655808952434732e-05,
+      "loss": 0.6823931932449341,
+      "step": 2634
+    },
+    {
+      "epoch": 1.1372168284789643,
+      "grad_norm": 1.3926069736480713,
+      "learning_rate": 9.654913163028272e-05,
+      "loss": 0.6671395897865295,
+      "step": 2636
+    },
+    {
+      "epoch": 1.1380798274002157,
+      "grad_norm": 1.2995383739471436,
+      "learning_rate": 9.65401625111049e-05,
+      "loss": 0.662219226360321,
+      "step": 2638
+    },
+    {
+      "epoch": 1.1389428263214671,
+      "grad_norm": 1.1584415435791016,
+      "learning_rate": 9.653118216897667e-05,
+      "loss": 0.6419461965560913,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1398058252427186,
+      "grad_norm": 1.3584444522857666,
+      "learning_rate": 9.652219060606364e-05,
+      "loss": 0.6844428181648254,
+      "step": 2642
+    },
+    {
+      "epoch": 1.1406688241639698,
+      "grad_norm": 1.2619858980178833,
+      "learning_rate": 9.651318782453407e-05,
+      "loss": 0.6761999130249023,
+      "step": 2644
+    },
+    {
+      "epoch": 1.1415318230852212,
+      "grad_norm": 1.2202177047729492,
+      "learning_rate": 9.650417382655892e-05,
+      "loss": 0.7258123159408569,
+      "step": 2646
+    },
+    {
+      "epoch": 1.1423948220064726,
+      "grad_norm": 1.0225578546524048,
+      "learning_rate": 9.64951486143119e-05,
+      "loss": 0.5709917545318604,
+      "step": 2648
+    },
+    {
+      "epoch": 1.1432578209277238,
+      "grad_norm": 1.0961005687713623,
+      "learning_rate": 9.64861121899694e-05,
+      "loss": 0.7353709936141968,
+      "step": 2650
+    },
+    {
+      "epoch": 1.1441208198489752,
+      "grad_norm": 0.9393200874328613,
+      "learning_rate": 9.64770645557105e-05,
+      "loss": 0.7083833813667297,
+      "step": 2652
+    },
+    {
+      "epoch": 1.1449838187702266,
+      "grad_norm": 1.089733362197876,
+      "learning_rate": 9.646800571371701e-05,
+      "loss": 0.6405303478240967,
+      "step": 2654
+    },
+    {
+      "epoch": 1.1458468176914778,
+      "grad_norm": 1.1581398248672485,
+      "learning_rate": 9.645893566617341e-05,
+      "loss": 0.617733359336853,
+      "step": 2656
+    },
+    {
+      "epoch": 1.1467098166127292,
+      "grad_norm": 1.1336220502853394,
+      "learning_rate": 9.644985441526693e-05,
+      "loss": 0.6948232650756836,
+      "step": 2658
+    },
+    {
+      "epoch": 1.1475728155339806,
+      "grad_norm": 1.028591513633728,
+      "learning_rate": 9.644076196318743e-05,
+      "loss": 0.6392919421195984,
+      "step": 2660
+    },
+    {
+      "epoch": 1.148435814455232,
+      "grad_norm": 1.2088173627853394,
+      "learning_rate": 9.643165831212756e-05,
+      "loss": 0.6384931206703186,
+      "step": 2662
+    },
+    {
+      "epoch": 1.1492988133764832,
+      "grad_norm": 0.9905838966369629,
+      "learning_rate": 9.642254346428259e-05,
+      "loss": 0.5909387469291687,
+      "step": 2664
+    },
+    {
+      "epoch": 1.1501618122977346,
+      "grad_norm": 1.1670424938201904,
+      "learning_rate": 9.641341742185054e-05,
+      "loss": 0.700904130935669,
+      "step": 2666
+    },
+    {
+      "epoch": 1.151024811218986,
+      "grad_norm": 1.164746880531311,
+      "learning_rate": 9.640428018703211e-05,
+      "loss": 0.6036953330039978,
+      "step": 2668
+    },
+    {
+      "epoch": 1.1518878101402374,
+      "grad_norm": 1.00895357131958,
+      "learning_rate": 9.639513176203072e-05,
+      "loss": 0.6158549785614014,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1527508090614886,
+      "grad_norm": 1.2236207723617554,
+      "learning_rate": 9.638597214905244e-05,
+      "loss": 0.7280431389808655,
+      "step": 2672
+    },
+    {
+      "epoch": 1.15361380798274,
+      "grad_norm": 1.0930732488632202,
+      "learning_rate": 9.637680135030609e-05,
+      "loss": 0.6434410810470581,
+      "step": 2674
+    },
+    {
+      "epoch": 1.1544768069039915,
+      "grad_norm": 1.366325855255127,
+      "learning_rate": 9.636761936800315e-05,
+      "loss": 0.6771191954612732,
+      "step": 2676
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 2.249835968017578,
+      "learning_rate": 9.635842620435785e-05,
+      "loss": 0.6382432579994202,
+      "step": 2678
+    },
+    {
+      "epoch": 1.156202804746494,
+      "grad_norm": 1.274483561515808,
+      "learning_rate": 9.634922186158704e-05,
+      "loss": 0.6733657121658325,
+      "step": 2680
+    },
+    {
+      "epoch": 1.1570658036677455,
+      "grad_norm": 1.3715782165527344,
+      "learning_rate": 9.634000634191033e-05,
+      "loss": 0.7601200938224792,
+      "step": 2682
+    },
+    {
+      "epoch": 1.1579288025889967,
+      "grad_norm": 1.1312460899353027,
+      "learning_rate": 9.633077964754998e-05,
+      "loss": 0.5959098935127258,
+      "step": 2684
+    },
+    {
+      "epoch": 1.158791801510248,
+      "grad_norm": 1.063233494758606,
+      "learning_rate": 9.632154178073098e-05,
+      "loss": 0.615096926689148,
+      "step": 2686
+    },
+    {
+      "epoch": 1.1596548004314995,
+      "grad_norm": 1.074887752532959,
+      "learning_rate": 9.6312292743681e-05,
+      "loss": 0.6553784608840942,
+      "step": 2688
+    },
+    {
+      "epoch": 1.160517799352751,
+      "grad_norm": 1.17954683303833,
+      "learning_rate": 9.630303253863042e-05,
+      "loss": 0.6963269114494324,
+      "step": 2690
+    },
+    {
+      "epoch": 1.161380798274002,
+      "grad_norm": 1.2549158334732056,
+      "learning_rate": 9.629376116781226e-05,
+      "loss": 0.6633999347686768,
+      "step": 2692
+    },
+    {
+      "epoch": 1.1622437971952535,
+      "grad_norm": 1.237196683883667,
+      "learning_rate": 9.628447863346229e-05,
+      "loss": 0.6727597117424011,
+      "step": 2694
+    },
+    {
+      "epoch": 1.163106796116505,
+      "grad_norm": 1.4047741889953613,
+      "learning_rate": 9.627518493781895e-05,
+      "loss": 0.7018169164657593,
+      "step": 2696
+    },
+    {
+      "epoch": 1.1639697950377563,
+      "grad_norm": 1.15412175655365,
+      "learning_rate": 9.626588008312339e-05,
+      "loss": 0.6566510796546936,
+      "step": 2698
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "grad_norm": 1.194110631942749,
+      "learning_rate": 9.625656407161938e-05,
+      "loss": 0.7017931342124939,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "eval_loss": 0.72404944896698,
+      "eval_runtime": 824.2877,
+      "eval_samples_per_second": 2.499,
+      "eval_steps_per_second": 2.499,
+      "step": 2700
+    },
+    {
+      "epoch": 1.165695792880259,
+      "grad_norm": 1.1512954235076904,
+      "learning_rate": 9.62472369055535e-05,
+      "loss": 0.7058532238006592,
+      "step": 2702
+    },
+    {
+      "epoch": 1.1665587918015103,
+      "grad_norm": 1.4743460416793823,
+      "learning_rate": 9.623789858717492e-05,
+      "loss": 0.7458376288414001,
+      "step": 2704
+    },
+    {
+      "epoch": 1.1674217907227615,
+      "grad_norm": 1.0526450872421265,
+      "learning_rate": 9.622854911873551e-05,
+      "loss": 0.6526424288749695,
+      "step": 2706
+    },
+    {
+      "epoch": 1.168284789644013,
+      "grad_norm": 1.0408949851989746,
+      "learning_rate": 9.62191885024899e-05,
+      "loss": 0.6055451035499573,
+      "step": 2708
+    },
+    {
+      "epoch": 1.1691477885652644,
+      "grad_norm": 1.214052677154541,
+      "learning_rate": 9.620981674069534e-05,
+      "loss": 0.666462779045105,
+      "step": 2710
+    },
+    {
+      "epoch": 1.1700107874865155,
+      "grad_norm": 1.2526963949203491,
+      "learning_rate": 9.62004338356118e-05,
+      "loss": 0.7208677530288696,
+      "step": 2712
+    },
+    {
+      "epoch": 1.170873786407767,
+      "grad_norm": 1.2097395658493042,
+      "learning_rate": 9.61910397895019e-05,
+      "loss": 0.6269269585609436,
+      "step": 2714
+    },
+    {
+      "epoch": 1.1717367853290184,
+      "grad_norm": 1.035733699798584,
+      "learning_rate": 9.618163460463096e-05,
+      "loss": 0.6063299179077148,
+      "step": 2716
+    },
+    {
+      "epoch": 1.1725997842502698,
+      "grad_norm": 1.2881336212158203,
+      "learning_rate": 9.617221828326706e-05,
+      "loss": 0.7084641456604004,
+      "step": 2718
+    },
+    {
+      "epoch": 1.173462783171521,
+      "grad_norm": 1.2808257341384888,
+      "learning_rate": 9.616279082768085e-05,
+      "loss": 0.6515718102455139,
+      "step": 2720
+    },
+    {
+      "epoch": 1.1743257820927724,
+      "grad_norm": 1.1174249649047852,
+      "learning_rate": 9.615335224014574e-05,
+      "loss": 0.6745902299880981,
+      "step": 2722
+    },
+    {
+      "epoch": 1.1751887810140238,
+      "grad_norm": 1.1222290992736816,
+      "learning_rate": 9.61439025229378e-05,
+      "loss": 0.6824861168861389,
+      "step": 2724
+    },
+    {
+      "epoch": 1.176051779935275,
+      "grad_norm": 1.156187891960144,
+      "learning_rate": 9.613444167833579e-05,
+      "loss": 0.6020817160606384,
+      "step": 2726
+    },
+    {
+      "epoch": 1.1769147788565264,
+      "grad_norm": 1.2502888441085815,
+      "learning_rate": 9.612496970862113e-05,
+      "loss": 0.6430572271347046,
+      "step": 2728
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 1.6157052516937256,
+      "learning_rate": 9.611548661607797e-05,
+      "loss": 0.7195103168487549,
+      "step": 2730
+    },
+    {
+      "epoch": 1.1786407766990292,
+      "grad_norm": 1.3196234703063965,
+      "learning_rate": 9.610599240299308e-05,
+      "loss": 0.710418164730072,
+      "step": 2732
+    },
+    {
+      "epoch": 1.1795037756202804,
+      "grad_norm": 1.2117719650268555,
+      "learning_rate": 9.6096487071656e-05,
+      "loss": 0.650095522403717,
+      "step": 2734
+    },
+    {
+      "epoch": 1.1803667745415318,
+      "grad_norm": 2.2723631858825684,
+      "learning_rate": 9.608697062435884e-05,
+      "loss": 0.7150285243988037,
+      "step": 2736
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 1.0523152351379395,
+      "learning_rate": 9.607744306339647e-05,
+      "loss": 0.6271997690200806,
+      "step": 2738
+    },
+    {
+      "epoch": 1.1820927723840344,
+      "grad_norm": 1.2567129135131836,
+      "learning_rate": 9.606790439106646e-05,
+      "loss": 0.659799337387085,
+      "step": 2740
+    },
+    {
+      "epoch": 1.1829557713052858,
+      "grad_norm": 1.1448849439620972,
+      "learning_rate": 9.605835460966896e-05,
+      "loss": 0.7097326517105103,
+      "step": 2742
+    },
+    {
+      "epoch": 1.1838187702265373,
+      "grad_norm": 1.3698351383209229,
+      "learning_rate": 9.604879372150688e-05,
+      "loss": 0.6692788004875183,
+      "step": 2744
+    },
+    {
+      "epoch": 1.1846817691477884,
+      "grad_norm": 1.1424204111099243,
+      "learning_rate": 9.603922172888577e-05,
+      "loss": 0.6453474760055542,
+      "step": 2746
+    },
+    {
+      "epoch": 1.1855447680690399,
+      "grad_norm": 1.2701301574707031,
+      "learning_rate": 9.60296386341139e-05,
+      "loss": 0.6197408437728882,
+      "step": 2748
+    },
+    {
+      "epoch": 1.1864077669902913,
+      "grad_norm": 1.1171884536743164,
+      "learning_rate": 9.602004443950217e-05,
+      "loss": 0.6525035500526428,
+      "step": 2750
+    },
+    {
+      "epoch": 1.1872707659115427,
+      "grad_norm": 1.132633090019226,
+      "learning_rate": 9.60104391473642e-05,
+      "loss": 0.6459797620773315,
+      "step": 2752
+    },
+    {
+      "epoch": 1.1881337648327939,
+      "grad_norm": 1.1362011432647705,
+      "learning_rate": 9.600082276001623e-05,
+      "loss": 0.6175895929336548,
+      "step": 2754
+    },
+    {
+      "epoch": 1.1889967637540453,
+      "grad_norm": 1.1216920614242554,
+      "learning_rate": 9.599119527977722e-05,
+      "loss": 0.6782001852989197,
+      "step": 2756
+    },
+    {
+      "epoch": 1.1898597626752967,
+      "grad_norm": 1.024833083152771,
+      "learning_rate": 9.59815567089688e-05,
+      "loss": 0.6387497782707214,
+      "step": 2758
+    },
+    {
+      "epoch": 1.190722761596548,
+      "grad_norm": 1.3539615869522095,
+      "learning_rate": 9.597190704991527e-05,
+      "loss": 0.6974280476570129,
+      "step": 2760
+    },
+    {
+      "epoch": 1.1915857605177993,
+      "grad_norm": 1.002167820930481,
+      "learning_rate": 9.59622463049436e-05,
+      "loss": 0.6039770245552063,
+      "step": 2762
+    },
+    {
+      "epoch": 1.1924487594390507,
+      "grad_norm": 1.150636076927185,
+      "learning_rate": 9.595257447638343e-05,
+      "loss": 0.6543651819229126,
+      "step": 2764
+    },
+    {
+      "epoch": 1.1933117583603021,
+      "grad_norm": 1.413001537322998,
+      "learning_rate": 9.594289156656709e-05,
+      "loss": 0.7388567328453064,
+      "step": 2766
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 1.2041958570480347,
+      "learning_rate": 9.593319757782955e-05,
+      "loss": 0.7128161191940308,
+      "step": 2768
+    },
+    {
+      "epoch": 1.1950377562028047,
+      "grad_norm": 1.2151374816894531,
+      "learning_rate": 9.592349251250847e-05,
+      "loss": 0.6593868136405945,
+      "step": 2770
+    },
+    {
+      "epoch": 1.1959007551240561,
+      "grad_norm": 1.6225353479385376,
+      "learning_rate": 9.59137763729442e-05,
+      "loss": 0.6128022074699402,
+      "step": 2772
+    },
+    {
+      "epoch": 1.1967637540453073,
+      "grad_norm": 1.1904171705245972,
+      "learning_rate": 9.590404916147976e-05,
+      "loss": 0.6537567377090454,
+      "step": 2774
+    },
+    {
+      "epoch": 1.1976267529665587,
+      "grad_norm": 1.159506916999817,
+      "learning_rate": 9.589431088046079e-05,
+      "loss": 0.6325924396514893,
+      "step": 2776
+    },
+    {
+      "epoch": 1.1984897518878102,
+      "grad_norm": 1.1314560174942017,
+      "learning_rate": 9.588456153223565e-05,
+      "loss": 0.6438963413238525,
+      "step": 2778
+    },
+    {
+      "epoch": 1.1993527508090616,
+      "grad_norm": 1.0182478427886963,
+      "learning_rate": 9.587480111915532e-05,
+      "loss": 0.6703475117683411,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2002157497303128,
+      "grad_norm": 1.5506517887115479,
+      "learning_rate": 9.586502964357352e-05,
+      "loss": 0.7694190144538879,
+      "step": 2782
+    },
+    {
+      "epoch": 1.2010787486515642,
+      "grad_norm": 1.0769826173782349,
+      "learning_rate": 9.585524710784659e-05,
+      "loss": 0.6117724180221558,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2019417475728156,
+      "grad_norm": 1.207175850868225,
+      "learning_rate": 9.584545351433352e-05,
+      "loss": 0.6375927329063416,
+      "step": 2786
+    },
+    {
+      "epoch": 1.202804746494067,
+      "grad_norm": 1.1291100978851318,
+      "learning_rate": 9.583564886539602e-05,
+      "loss": 0.6526537537574768,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2036677454153182,
+      "grad_norm": 1.3076926469802856,
+      "learning_rate": 9.582583316339843e-05,
+      "loss": 0.7080113887786865,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2045307443365696,
+      "grad_norm": 1.166823148727417,
+      "learning_rate": 9.581600641070773e-05,
+      "loss": 0.6415606141090393,
+      "step": 2792
+    },
+    {
+      "epoch": 1.205393743257821,
+      "grad_norm": 1.1248292922973633,
+      "learning_rate": 9.580616860969364e-05,
+      "loss": 0.6538968086242676,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2062567421790722,
+      "grad_norm": 1.406731128692627,
+      "learning_rate": 9.579631976272849e-05,
+      "loss": 0.694507896900177,
+      "step": 2796
+    },
+    {
+      "epoch": 1.2071197411003236,
+      "grad_norm": 1.185895562171936,
+      "learning_rate": 9.578645987218728e-05,
+      "loss": 0.7064871788024902,
+      "step": 2798
+    },
+    {
+      "epoch": 1.207982740021575,
+      "grad_norm": 1.0860708951950073,
+      "learning_rate": 9.577658894044768e-05,
+      "loss": 0.5790154933929443,
+      "step": 2800
+    },
+    {
+      "epoch": 1.207982740021575,
+      "eval_loss": 0.7220398187637329,
+      "eval_runtime": 831.5812,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2088457389428262,
+      "grad_norm": 1.2761744260787964,
+      "learning_rate": 9.576670696989001e-05,
+      "loss": 0.6645335555076599,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2097087378640776,
+      "grad_norm": 1.3502917289733887,
+      "learning_rate": 9.575681396289728e-05,
+      "loss": 0.6320346593856812,
+      "step": 2804
+    },
+    {
+      "epoch": 1.210571736785329,
+      "grad_norm": 1.2638593912124634,
+      "learning_rate": 9.574690992185514e-05,
+      "loss": 0.6297362446784973,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2114347357065804,
+      "grad_norm": 1.0787774324417114,
+      "learning_rate": 9.573699484915189e-05,
+      "loss": 0.7569923996925354,
+      "step": 2808
+    },
+    {
+      "epoch": 1.2122977346278316,
+      "grad_norm": 1.097254991531372,
+      "learning_rate": 9.572706874717851e-05,
+      "loss": 0.6712964773178101,
+      "step": 2810
+    },
+    {
+      "epoch": 1.213160733549083,
+      "grad_norm": 1.0995726585388184,
+      "learning_rate": 9.571713161832865e-05,
+      "loss": 0.6775450706481934,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2140237324703345,
+      "grad_norm": 1.7386797666549683,
+      "learning_rate": 9.570718346499859e-05,
+      "loss": 0.6691418290138245,
+      "step": 2814
+    },
+    {
+      "epoch": 1.2148867313915859,
+      "grad_norm": 1.2281049489974976,
+      "learning_rate": 9.569722428958728e-05,
+      "loss": 0.6406403183937073,
+      "step": 2816
+    },
+    {
+      "epoch": 1.215749730312837,
+      "grad_norm": 1.1326305866241455,
+      "learning_rate": 9.568725409449635e-05,
+      "loss": 0.7008584141731262,
+      "step": 2818
+    },
+    {
+      "epoch": 1.2166127292340885,
+      "grad_norm": 1.3039381504058838,
+      "learning_rate": 9.567727288213005e-05,
+      "loss": 0.7438415288925171,
+      "step": 2820
+    },
+    {
+      "epoch": 1.2174757281553399,
+      "grad_norm": 1.219136357307434,
+      "learning_rate": 9.566728065489531e-05,
+      "loss": 0.6937777996063232,
+      "step": 2822
+    },
+    {
+      "epoch": 1.218338727076591,
+      "grad_norm": 1.321285605430603,
+      "learning_rate": 9.56572774152017e-05,
+      "loss": 0.7586468458175659,
+      "step": 2824
+    },
+    {
+      "epoch": 1.2192017259978425,
+      "grad_norm": 1.3538484573364258,
+      "learning_rate": 9.564726316546147e-05,
+      "loss": 0.6356340050697327,
+      "step": 2826
+    },
+    {
+      "epoch": 1.220064724919094,
+      "grad_norm": 1.1015816926956177,
+      "learning_rate": 9.563723790808952e-05,
+      "loss": 0.6304616928100586,
+      "step": 2828
+    },
+    {
+      "epoch": 1.220927723840345,
+      "grad_norm": 0.9658390283584595,
+      "learning_rate": 9.562720164550337e-05,
+      "loss": 0.6459048986434937,
+      "step": 2830
+    },
+    {
+      "epoch": 1.2217907227615965,
+      "grad_norm": 1.3981443643569946,
+      "learning_rate": 9.561715438012323e-05,
+      "loss": 0.7034182548522949,
+      "step": 2832
+    },
+    {
+      "epoch": 1.222653721682848,
+      "grad_norm": 1.1737093925476074,
+      "learning_rate": 9.560709611437195e-05,
+      "loss": 0.6629860997200012,
+      "step": 2834
+    },
+    {
+      "epoch": 1.2235167206040993,
+      "grad_norm": 1.2071303129196167,
+      "learning_rate": 9.559702685067506e-05,
+      "loss": 0.7166537642478943,
+      "step": 2836
+    },
+    {
+      "epoch": 1.2243797195253505,
+      "grad_norm": 1.2520930767059326,
+      "learning_rate": 9.558694659146068e-05,
+      "loss": 0.6711867451667786,
+      "step": 2838
+    },
+    {
+      "epoch": 1.225242718446602,
+      "grad_norm": 1.2304543256759644,
+      "learning_rate": 9.557685533915964e-05,
+      "loss": 0.7135099768638611,
+      "step": 2840
+    },
+    {
+      "epoch": 1.2261057173678533,
+      "grad_norm": 1.1509613990783691,
+      "learning_rate": 9.556675309620539e-05,
+      "loss": 0.6279205083847046,
+      "step": 2842
+    },
+    {
+      "epoch": 1.2269687162891048,
+      "grad_norm": 1.2024176120758057,
+      "learning_rate": 9.555663986503406e-05,
+      "loss": 0.6848570704460144,
+      "step": 2844
+    },
+    {
+      "epoch": 1.227831715210356,
+      "grad_norm": 1.360257863998413,
+      "learning_rate": 9.55465156480844e-05,
+      "loss": 0.7109206914901733,
+      "step": 2846
+    },
+    {
+      "epoch": 1.2286947141316074,
+      "grad_norm": 1.1556848287582397,
+      "learning_rate": 9.55363804477978e-05,
+      "loss": 0.5983237028121948,
+      "step": 2848
+    },
+    {
+      "epoch": 1.2295577130528588,
+      "grad_norm": 1.131439447402954,
+      "learning_rate": 9.552623426661836e-05,
+      "loss": 0.6387401223182678,
+      "step": 2850
+    },
+    {
+      "epoch": 1.23042071197411,
+      "grad_norm": 1.1854766607284546,
+      "learning_rate": 9.551607710699273e-05,
+      "loss": 0.6584593057632446,
+      "step": 2852
+    },
+    {
+      "epoch": 1.2312837108953614,
+      "grad_norm": 1.1811782121658325,
+      "learning_rate": 9.55059089713703e-05,
+      "loss": 0.6284405589103699,
+      "step": 2854
+    },
+    {
+      "epoch": 1.2321467098166128,
+      "grad_norm": 1.1038708686828613,
+      "learning_rate": 9.549572986220307e-05,
+      "loss": 0.6183417439460754,
+      "step": 2856
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 1.3645228147506714,
+      "learning_rate": 9.548553978194567e-05,
+      "loss": 0.7266764640808105,
+      "step": 2858
+    },
+    {
+      "epoch": 1.2338727076591154,
+      "grad_norm": 1.1896798610687256,
+      "learning_rate": 9.547533873305542e-05,
+      "loss": 0.6580738425254822,
+      "step": 2860
+    },
+    {
+      "epoch": 1.2347357065803668,
+      "grad_norm": 1.1351168155670166,
+      "learning_rate": 9.546512671799224e-05,
+      "loss": 0.669810950756073,
+      "step": 2862
+    },
+    {
+      "epoch": 1.2355987055016182,
+      "grad_norm": 1.1937936544418335,
+      "learning_rate": 9.545490373921869e-05,
+      "loss": 0.730329155921936,
+      "step": 2864
+    },
+    {
+      "epoch": 1.2364617044228694,
+      "grad_norm": 1.094347357749939,
+      "learning_rate": 9.544466979920002e-05,
+      "loss": 0.6867446303367615,
+      "step": 2866
+    },
+    {
+      "epoch": 1.2373247033441208,
+      "grad_norm": 1.4509600400924683,
+      "learning_rate": 9.54344249004041e-05,
+      "loss": 0.735468327999115,
+      "step": 2868
+    },
+    {
+      "epoch": 1.2381877022653722,
+      "grad_norm": 1.1195857524871826,
+      "learning_rate": 9.542416904530142e-05,
+      "loss": 0.6316167116165161,
+      "step": 2870
+    },
+    {
+      "epoch": 1.2390507011866236,
+      "grad_norm": 1.0227746963500977,
+      "learning_rate": 9.541390223636517e-05,
+      "loss": 0.6637873649597168,
+      "step": 2872
+    },
+    {
+      "epoch": 1.2399137001078748,
+      "grad_norm": 1.1871414184570312,
+      "learning_rate": 9.540362447607111e-05,
+      "loss": 0.6579877734184265,
+      "step": 2874
+    },
+    {
+      "epoch": 1.2407766990291262,
+      "grad_norm": 1.3265694379806519,
+      "learning_rate": 9.539333576689768e-05,
+      "loss": 0.7563158273696899,
+      "step": 2876
+    },
+    {
+      "epoch": 1.2416396979503777,
+      "grad_norm": 1.0074095726013184,
+      "learning_rate": 9.538303611132598e-05,
+      "loss": 0.6373291015625,
+      "step": 2878
+    },
+    {
+      "epoch": 1.2425026968716288,
+      "grad_norm": 1.421204924583435,
+      "learning_rate": 9.537272551183971e-05,
+      "loss": 0.6989303827285767,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2433656957928803,
+      "grad_norm": 1.256486415863037,
+      "learning_rate": 9.536240397092522e-05,
+      "loss": 0.7590261697769165,
+      "step": 2882
+    },
+    {
+      "epoch": 1.2442286947141317,
+      "grad_norm": 1.1513274908065796,
+      "learning_rate": 9.53520714910715e-05,
+      "loss": 0.6652005314826965,
+      "step": 2884
+    },
+    {
+      "epoch": 1.2450916936353829,
+      "grad_norm": 1.2129496335983276,
+      "learning_rate": 9.53417280747702e-05,
+      "loss": 0.6510938405990601,
+      "step": 2886
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 1.1393296718597412,
+      "learning_rate": 9.533137372451557e-05,
+      "loss": 0.6405928730964661,
+      "step": 2888
+    },
+    {
+      "epoch": 1.2468176914778857,
+      "grad_norm": 1.1467162370681763,
+      "learning_rate": 9.532100844280451e-05,
+      "loss": 0.6819307208061218,
+      "step": 2890
+    },
+    {
+      "epoch": 1.247680690399137,
+      "grad_norm": 1.1516613960266113,
+      "learning_rate": 9.531063223213659e-05,
+      "loss": 0.6333560347557068,
+      "step": 2892
+    },
+    {
+      "epoch": 1.2485436893203883,
+      "grad_norm": 1.463189959526062,
+      "learning_rate": 9.530024509501396e-05,
+      "loss": 0.7259839177131653,
+      "step": 2894
+    },
+    {
+      "epoch": 1.2494066882416397,
+      "grad_norm": 1.0450547933578491,
+      "learning_rate": 9.528984703394144e-05,
+      "loss": 0.661693811416626,
+      "step": 2896
+    },
+    {
+      "epoch": 1.250269687162891,
+      "grad_norm": 1.0525774955749512,
+      "learning_rate": 9.527943805142648e-05,
+      "loss": 0.5889303684234619,
+      "step": 2898
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "grad_norm": 1.130048394203186,
+      "learning_rate": 9.526901814997914e-05,
+      "loss": 0.6380610466003418,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "eval_loss": 0.7182289958000183,
+      "eval_runtime": 840.8156,
+      "eval_samples_per_second": 2.45,
+      "eval_steps_per_second": 2.45,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2519956850053937,
+      "grad_norm": 1.0194777250289917,
+      "learning_rate": 9.525858733211217e-05,
+      "loss": 0.6397356986999512,
+      "step": 2902
+    },
+    {
+      "epoch": 1.2528586839266451,
+      "grad_norm": 1.3902885913848877,
+      "learning_rate": 9.524814560034087e-05,
+      "loss": 0.7931577563285828,
+      "step": 2904
+    },
+    {
+      "epoch": 1.2537216828478965,
+      "grad_norm": 1.258169412612915,
+      "learning_rate": 9.523769295718325e-05,
+      "loss": 0.6997087597846985,
+      "step": 2906
+    },
+    {
+      "epoch": 1.2545846817691477,
+      "grad_norm": 1.3403711318969727,
+      "learning_rate": 9.522722940515991e-05,
+      "loss": 0.6946825385093689,
+      "step": 2908
+    },
+    {
+      "epoch": 1.2554476806903991,
+      "grad_norm": 1.1724950075149536,
+      "learning_rate": 9.521675494679409e-05,
+      "loss": 0.6309347748756409,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2563106796116505,
+      "grad_norm": 1.210274338722229,
+      "learning_rate": 9.520626958461163e-05,
+      "loss": 0.6453542113304138,
+      "step": 2912
+    },
+    {
+      "epoch": 1.2571736785329017,
+      "grad_norm": 1.233112096786499,
+      "learning_rate": 9.519577332114106e-05,
+      "loss": 0.6747867465019226,
+      "step": 2914
+    },
+    {
+      "epoch": 1.2580366774541532,
+      "grad_norm": 1.1785951852798462,
+      "learning_rate": 9.51852661589135e-05,
+      "loss": 0.6729235053062439,
+      "step": 2916
+    },
+    {
+      "epoch": 1.2588996763754046,
+      "grad_norm": 1.0781456232070923,
+      "learning_rate": 9.51747481004627e-05,
+      "loss": 0.6581047177314758,
+      "step": 2918
+    },
+    {
+      "epoch": 1.2597626752966558,
+      "grad_norm": 1.1122184991836548,
+      "learning_rate": 9.516421914832506e-05,
+      "loss": 0.6576398015022278,
+      "step": 2920
+    },
+    {
+      "epoch": 1.2606256742179072,
+      "grad_norm": 1.3378703594207764,
+      "learning_rate": 9.515367930503956e-05,
+      "loss": 0.7026275396347046,
+      "step": 2922
+    },
+    {
+      "epoch": 1.2614886731391586,
+      "grad_norm": 1.127657175064087,
+      "learning_rate": 9.514312857314785e-05,
+      "loss": 0.6111703515052795,
+      "step": 2924
+    },
+    {
+      "epoch": 1.26235167206041,
+      "grad_norm": 1.2058607339859009,
+      "learning_rate": 9.513256695519419e-05,
+      "loss": 0.6698760986328125,
+      "step": 2926
+    },
+    {
+      "epoch": 1.2632146709816614,
+      "grad_norm": 1.2206004858016968,
+      "learning_rate": 9.512199445372546e-05,
+      "loss": 0.6793603897094727,
+      "step": 2928
+    },
+    {
+      "epoch": 1.2640776699029126,
+      "grad_norm": 1.167311668395996,
+      "learning_rate": 9.511141107129119e-05,
+      "loss": 0.709406316280365,
+      "step": 2930
+    },
+    {
+      "epoch": 1.264940668824164,
+      "grad_norm": 1.3347517251968384,
+      "learning_rate": 9.51008168104435e-05,
+      "loss": 0.685570240020752,
+      "step": 2932
+    },
+    {
+      "epoch": 1.2658036677454154,
+      "grad_norm": 1.186468482017517,
+      "learning_rate": 9.509021167373714e-05,
+      "loss": 0.6775280833244324,
+      "step": 2934
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 1.3934271335601807,
+      "learning_rate": 9.507959566372953e-05,
+      "loss": 0.6587507128715515,
+      "step": 2936
+    },
+    {
+      "epoch": 1.267529665587918,
+      "grad_norm": 1.3766281604766846,
+      "learning_rate": 9.506896878298061e-05,
+      "loss": 0.692707896232605,
+      "step": 2938
+    },
+    {
+      "epoch": 1.2683926645091694,
+      "grad_norm": 1.034883975982666,
+      "learning_rate": 9.505833103405307e-05,
+      "loss": 0.6481655836105347,
+      "step": 2940
+    },
+    {
+      "epoch": 1.2692556634304206,
+      "grad_norm": 1.113044261932373,
+      "learning_rate": 9.504768241951212e-05,
+      "loss": 0.5867606997489929,
+      "step": 2942
+    },
+    {
+      "epoch": 1.270118662351672,
+      "grad_norm": 1.0769455432891846,
+      "learning_rate": 9.503702294192563e-05,
+      "loss": 0.6147398948669434,
+      "step": 2944
+    },
+    {
+      "epoch": 1.2709816612729234,
+      "grad_norm": 1.100616216659546,
+      "learning_rate": 9.502635260386408e-05,
+      "loss": 0.650299608707428,
+      "step": 2946
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 1.1982520818710327,
+      "learning_rate": 9.501567140790057e-05,
+      "loss": 0.6896426677703857,
+      "step": 2948
+    },
+    {
+      "epoch": 1.272707659115426,
+      "grad_norm": 1.276911735534668,
+      "learning_rate": 9.500497935661085e-05,
+      "loss": 0.6780133247375488,
+      "step": 2950
+    },
+    {
+      "epoch": 1.2735706580366775,
+      "grad_norm": 1.131345272064209,
+      "learning_rate": 9.499427645257323e-05,
+      "loss": 0.6906163692474365,
+      "step": 2952
+    },
+    {
+      "epoch": 1.2744336569579289,
+      "grad_norm": 1.125251293182373,
+      "learning_rate": 9.498356269836868e-05,
+      "loss": 0.6418634057044983,
+      "step": 2954
+    },
+    {
+      "epoch": 1.2752966558791803,
+      "grad_norm": 1.2508443593978882,
+      "learning_rate": 9.497283809658078e-05,
+      "loss": 0.690779983997345,
+      "step": 2956
+    },
+    {
+      "epoch": 1.2761596548004315,
+      "grad_norm": 0.9558717608451843,
+      "learning_rate": 9.49621026497957e-05,
+      "loss": 0.645685613155365,
+      "step": 2958
+    },
+    {
+      "epoch": 1.2770226537216829,
+      "grad_norm": 1.1781591176986694,
+      "learning_rate": 9.495135636060226e-05,
+      "loss": 0.6313972473144531,
+      "step": 2960
+    },
+    {
+      "epoch": 1.2778856526429343,
+      "grad_norm": 0.9713736772537231,
+      "learning_rate": 9.494059923159189e-05,
+      "loss": 0.5853794813156128,
+      "step": 2962
+    },
+    {
+      "epoch": 1.2787486515641855,
+      "grad_norm": 1.3793153762817383,
+      "learning_rate": 9.492983126535859e-05,
+      "loss": 0.6966362595558167,
+      "step": 2964
+    },
+    {
+      "epoch": 1.279611650485437,
+      "grad_norm": 1.250563383102417,
+      "learning_rate": 9.491905246449903e-05,
+      "loss": 0.7281587719917297,
+      "step": 2966
+    },
+    {
+      "epoch": 1.2804746494066883,
+      "grad_norm": 1.3268930912017822,
+      "learning_rate": 9.490826283161247e-05,
+      "loss": 0.7322974801063538,
+      "step": 2968
+    },
+    {
+      "epoch": 1.2813376483279395,
+      "grad_norm": 1.3876935243606567,
+      "learning_rate": 9.489746236930077e-05,
+      "loss": 0.7875081896781921,
+      "step": 2970
+    },
+    {
+      "epoch": 1.282200647249191,
+      "grad_norm": 1.02049720287323,
+      "learning_rate": 9.488665108016842e-05,
+      "loss": 0.601652979850769,
+      "step": 2972
+    },
+    {
+      "epoch": 1.2830636461704423,
+      "grad_norm": 1.0868756771087646,
+      "learning_rate": 9.487582896682252e-05,
+      "loss": 0.6255780458450317,
+      "step": 2974
+    },
+    {
+      "epoch": 1.2839266450916935,
+      "grad_norm": 1.00597083568573,
+      "learning_rate": 9.486499603187275e-05,
+      "loss": 0.5967418551445007,
+      "step": 2976
+    },
+    {
+      "epoch": 1.284789644012945,
+      "grad_norm": 1.3068969249725342,
+      "learning_rate": 9.485415227793145e-05,
+      "loss": 0.719573438167572,
+      "step": 2978
+    },
+    {
+      "epoch": 1.2856526429341963,
+      "grad_norm": 1.2478386163711548,
+      "learning_rate": 9.484329770761353e-05,
+      "loss": 0.665522038936615,
+      "step": 2980
+    },
+    {
+      "epoch": 1.2865156418554478,
+      "grad_norm": 1.1220918893814087,
+      "learning_rate": 9.483243232353652e-05,
+      "loss": 0.6798994541168213,
+      "step": 2982
+    },
+    {
+      "epoch": 1.287378640776699,
+      "grad_norm": 1.1369609832763672,
+      "learning_rate": 9.482155612832057e-05,
+      "loss": 0.672002911567688,
+      "step": 2984
+    },
+    {
+      "epoch": 1.2882416396979504,
+      "grad_norm": 1.2077715396881104,
+      "learning_rate": 9.48106691245884e-05,
+      "loss": 0.6755662560462952,
+      "step": 2986
+    },
+    {
+      "epoch": 1.2891046386192018,
+      "grad_norm": 1.3299074172973633,
+      "learning_rate": 9.479977131496539e-05,
+      "loss": 0.7504231333732605,
+      "step": 2988
+    },
+    {
+      "epoch": 1.2899676375404532,
+      "grad_norm": 1.08982253074646,
+      "learning_rate": 9.478886270207948e-05,
+      "loss": 0.6837405562400818,
+      "step": 2990
+    },
+    {
+      "epoch": 1.2908306364617044,
+      "grad_norm": 1.1015464067459106,
+      "learning_rate": 9.477794328856124e-05,
+      "loss": 0.6552648544311523,
+      "step": 2992
+    },
+    {
+      "epoch": 1.2916936353829558,
+      "grad_norm": 1.1793229579925537,
+      "learning_rate": 9.476701307704384e-05,
+      "loss": 0.6160483360290527,
+      "step": 2994
+    },
+    {
+      "epoch": 1.2925566343042072,
+      "grad_norm": 1.4781638383865356,
+      "learning_rate": 9.475607207016306e-05,
+      "loss": 0.7406170964241028,
+      "step": 2996
+    },
+    {
+      "epoch": 1.2934196332254584,
+      "grad_norm": 1.2391449213027954,
+      "learning_rate": 9.474512027055724e-05,
+      "loss": 0.5950883626937866,
+      "step": 2998
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "grad_norm": 1.166535496711731,
+      "learning_rate": 9.473415768086738e-05,
+      "loss": 0.6587089896202087,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "eval_loss": 0.7160727977752686,
+      "eval_runtime": 830.8186,
+      "eval_samples_per_second": 2.479,
+      "eval_steps_per_second": 2.479,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2951456310679612,
+      "grad_norm": 0.9833650588989258,
+      "learning_rate": 9.472318430373706e-05,
+      "loss": 0.5802602767944336,
+      "step": 3002
+    },
+    {
+      "epoch": 1.2960086299892124,
+      "grad_norm": 1.047341227531433,
+      "learning_rate": 9.471220014181246e-05,
+      "loss": 0.6255064606666565,
+      "step": 3004
+    },
+    {
+      "epoch": 1.2968716289104638,
+      "grad_norm": 1.2761468887329102,
+      "learning_rate": 9.470120519774237e-05,
+      "loss": 0.7038446068763733,
+      "step": 3006
+    },
+    {
+      "epoch": 1.2977346278317152,
+      "grad_norm": 1.2568607330322266,
+      "learning_rate": 9.469019947417815e-05,
+      "loss": 0.7063520550727844,
+      "step": 3008
+    },
+    {
+      "epoch": 1.2985976267529666,
+      "grad_norm": 1.2337472438812256,
+      "learning_rate": 9.467918297377378e-05,
+      "loss": 0.6931142807006836,
+      "step": 3010
+    },
+    {
+      "epoch": 1.2994606256742178,
+      "grad_norm": 1.1557739973068237,
+      "learning_rate": 9.466815569918588e-05,
+      "loss": 0.6660345196723938,
+      "step": 3012
+    },
+    {
+      "epoch": 1.3003236245954692,
+      "grad_norm": 1.2341985702514648,
+      "learning_rate": 9.465711765307357e-05,
+      "loss": 0.6522690653800964,
+      "step": 3014
+    },
+    {
+      "epoch": 1.3011866235167207,
+      "grad_norm": 1.0495452880859375,
+      "learning_rate": 9.464606883809867e-05,
+      "loss": 0.6304616332054138,
+      "step": 3016
+    },
+    {
+      "epoch": 1.302049622437972,
+      "grad_norm": 1.1920377016067505,
+      "learning_rate": 9.463500925692554e-05,
+      "loss": 0.7271291613578796,
+      "step": 3018
+    },
+    {
+      "epoch": 1.3029126213592233,
+      "grad_norm": 1.0697529315948486,
+      "learning_rate": 9.462393891222112e-05,
+      "loss": 0.5725325345993042,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3037756202804747,
+      "grad_norm": 1.0559576749801636,
+      "learning_rate": 9.461285780665502e-05,
+      "loss": 0.5725870728492737,
+      "step": 3022
+    },
+    {
+      "epoch": 1.304638619201726,
+      "grad_norm": 1.1314871311187744,
+      "learning_rate": 9.460176594289937e-05,
+      "loss": 0.6498687863349915,
+      "step": 3024
+    },
+    {
+      "epoch": 1.3055016181229773,
+      "grad_norm": 1.0620028972625732,
+      "learning_rate": 9.459066332362892e-05,
+      "loss": 0.5473160147666931,
+      "step": 3026
+    },
+    {
+      "epoch": 1.3063646170442287,
+      "grad_norm": 1.245997667312622,
+      "learning_rate": 9.457954995152103e-05,
+      "loss": 0.6976684927940369,
+      "step": 3028
+    },
+    {
+      "epoch": 1.30722761596548,
+      "grad_norm": 1.0973089933395386,
+      "learning_rate": 9.456842582925565e-05,
+      "loss": 0.6469701528549194,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3080906148867313,
+      "grad_norm": 1.1558068990707397,
+      "learning_rate": 9.455729095951527e-05,
+      "loss": 0.6712765097618103,
+      "step": 3032
+    },
+    {
+      "epoch": 1.3089536138079827,
+      "grad_norm": 1.0367828607559204,
+      "learning_rate": 9.454614534498506e-05,
+      "loss": 0.6504483222961426,
+      "step": 3034
+    },
+    {
+      "epoch": 1.309816612729234,
+      "grad_norm": 1.3485349416732788,
+      "learning_rate": 9.453498898835272e-05,
+      "loss": 0.7000672221183777,
+      "step": 3036
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 1.2435184717178345,
+      "learning_rate": 9.452382189230853e-05,
+      "loss": 0.6946983337402344,
+      "step": 3038
+    },
+    {
+      "epoch": 1.3115426105717367,
+      "grad_norm": 1.0728851556777954,
+      "learning_rate": 9.451264405954541e-05,
+      "loss": 0.5928027033805847,
+      "step": 3040
+    },
+    {
+      "epoch": 1.3124056094929881,
+      "grad_norm": 1.2320033311843872,
+      "learning_rate": 9.450145549275884e-05,
+      "loss": 0.6589536070823669,
+      "step": 3042
+    },
+    {
+      "epoch": 1.3132686084142395,
+      "grad_norm": 1.143538236618042,
+      "learning_rate": 9.449025619464689e-05,
+      "loss": 0.6516439914703369,
+      "step": 3044
+    },
+    {
+      "epoch": 1.314131607335491,
+      "grad_norm": 1.3419140577316284,
+      "learning_rate": 9.447904616791023e-05,
+      "loss": 0.6889938116073608,
+      "step": 3046
+    },
+    {
+      "epoch": 1.3149946062567421,
+      "grad_norm": 1.4128848314285278,
+      "learning_rate": 9.44678254152521e-05,
+      "loss": 0.7203409075737,
+      "step": 3048
+    },
+    {
+      "epoch": 1.3158576051779936,
+      "grad_norm": 1.1409581899642944,
+      "learning_rate": 9.445659393937833e-05,
+      "loss": 0.656524658203125,
+      "step": 3050
+    },
+    {
+      "epoch": 1.316720604099245,
+      "grad_norm": 1.0904145240783691,
+      "learning_rate": 9.444535174299734e-05,
+      "loss": 0.6888052821159363,
+      "step": 3052
+    },
+    {
+      "epoch": 1.3175836030204962,
+      "grad_norm": 1.1533669233322144,
+      "learning_rate": 9.443409882882014e-05,
+      "loss": 0.6149572730064392,
+      "step": 3054
+    },
+    {
+      "epoch": 1.3184466019417476,
+      "grad_norm": 1.0775257349014282,
+      "learning_rate": 9.442283519956031e-05,
+      "loss": 0.6445736885070801,
+      "step": 3056
+    },
+    {
+      "epoch": 1.319309600862999,
+      "grad_norm": 1.1126797199249268,
+      "learning_rate": 9.441156085793403e-05,
+      "loss": 0.6723806858062744,
+      "step": 3058
+    },
+    {
+      "epoch": 1.3201725997842502,
+      "grad_norm": 1.1584032773971558,
+      "learning_rate": 9.440027580666008e-05,
+      "loss": 0.5901219248771667,
+      "step": 3060
+    },
+    {
+      "epoch": 1.3210355987055016,
+      "grad_norm": 1.0951672792434692,
+      "learning_rate": 9.438898004845975e-05,
+      "loss": 0.6816843748092651,
+      "step": 3062
+    },
+    {
+      "epoch": 1.321898597626753,
+      "grad_norm": 1.2165061235427856,
+      "learning_rate": 9.4377673586057e-05,
+      "loss": 0.6883957386016846,
+      "step": 3064
+    },
+    {
+      "epoch": 1.3227615965480042,
+      "grad_norm": 1.1537514925003052,
+      "learning_rate": 9.436635642217832e-05,
+      "loss": 0.6828808188438416,
+      "step": 3066
+    },
+    {
+      "epoch": 1.3236245954692556,
+      "grad_norm": 1.2186168432235718,
+      "learning_rate": 9.435502855955281e-05,
+      "loss": 0.6086717844009399,
+      "step": 3068
+    },
+    {
+      "epoch": 1.324487594390507,
+      "grad_norm": 1.0345975160598755,
+      "learning_rate": 9.43436900009121e-05,
+      "loss": 0.6000884771347046,
+      "step": 3070
+    },
+    {
+      "epoch": 1.3253505933117584,
+      "grad_norm": 1.1357595920562744,
+      "learning_rate": 9.433234074899044e-05,
+      "loss": 0.6812195181846619,
+      "step": 3072
+    },
+    {
+      "epoch": 1.3262135922330098,
+      "grad_norm": 1.1914640665054321,
+      "learning_rate": 9.432098080652467e-05,
+      "loss": 0.6666595935821533,
+      "step": 3074
+    },
+    {
+      "epoch": 1.327076591154261,
+      "grad_norm": 1.0881390571594238,
+      "learning_rate": 9.430961017625418e-05,
+      "loss": 0.6455721259117126,
+      "step": 3076
+    },
+    {
+      "epoch": 1.3279395900755124,
+      "grad_norm": 1.1859697103500366,
+      "learning_rate": 9.429822886092093e-05,
+      "loss": 0.5861079692840576,
+      "step": 3078
+    },
+    {
+      "epoch": 1.3288025889967638,
+      "grad_norm": 1.083703637123108,
+      "learning_rate": 9.42868368632695e-05,
+      "loss": 0.6173920035362244,
+      "step": 3080
+    },
+    {
+      "epoch": 1.329665587918015,
+      "grad_norm": 1.2820687294006348,
+      "learning_rate": 9.427543418604701e-05,
+      "loss": 0.817629873752594,
+      "step": 3082
+    },
+    {
+      "epoch": 1.3305285868392664,
+      "grad_norm": 1.1267426013946533,
+      "learning_rate": 9.426402083200315e-05,
+      "loss": 0.7454650402069092,
+      "step": 3084
+    },
+    {
+      "epoch": 1.3313915857605179,
+      "grad_norm": 1.0095305442810059,
+      "learning_rate": 9.425259680389022e-05,
+      "loss": 0.625328779220581,
+      "step": 3086
+    },
+    {
+      "epoch": 1.332254584681769,
+      "grad_norm": 1.183151364326477,
+      "learning_rate": 9.424116210446305e-05,
+      "loss": 0.5977230072021484,
+      "step": 3088
+    },
+    {
+      "epoch": 1.3331175836030205,
+      "grad_norm": 1.1312501430511475,
+      "learning_rate": 9.42297167364791e-05,
+      "loss": 0.6251013278961182,
+      "step": 3090
+    },
+    {
+      "epoch": 1.3339805825242719,
+      "grad_norm": 1.3226121664047241,
+      "learning_rate": 9.421826070269832e-05,
+      "loss": 0.7469011545181274,
+      "step": 3092
+    },
+    {
+      "epoch": 1.334843581445523,
+      "grad_norm": 1.1389979124069214,
+      "learning_rate": 9.420679400588334e-05,
+      "loss": 0.7012571096420288,
+      "step": 3094
+    },
+    {
+      "epoch": 1.3357065803667745,
+      "grad_norm": 1.2065027952194214,
+      "learning_rate": 9.419531664879925e-05,
+      "loss": 0.7112689018249512,
+      "step": 3096
+    },
+    {
+      "epoch": 1.3365695792880259,
+      "grad_norm": 1.2532200813293457,
+      "learning_rate": 9.418382863421379e-05,
+      "loss": 0.6219019293785095,
+      "step": 3098
+    },
+    {
+      "epoch": 1.3374325782092773,
+      "grad_norm": 1.1859256029129028,
+      "learning_rate": 9.417232996489724e-05,
+      "loss": 0.6905092597007751,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3374325782092773,
+      "eval_loss": 0.7127581834793091,
+      "eval_runtime": 821.1346,
+      "eval_samples_per_second": 2.509,
+      "eval_steps_per_second": 2.509,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3382955771305287,
+      "grad_norm": 1.172967791557312,
+      "learning_rate": 9.416082064362244e-05,
+      "loss": 0.6473057866096497,
+      "step": 3102
+    },
+    {
+      "epoch": 1.33915857605178,
+      "grad_norm": 1.2966243028640747,
+      "learning_rate": 9.414930067316482e-05,
+      "loss": 0.7377036213874817,
+      "step": 3104
+    },
+    {
+      "epoch": 1.3400215749730313,
+      "grad_norm": 1.0844253301620483,
+      "learning_rate": 9.413777005630237e-05,
+      "loss": 0.6842042207717896,
+      "step": 3106
+    },
+    {
+      "epoch": 1.3408845738942827,
+      "grad_norm": 0.930098831653595,
+      "learning_rate": 9.412622879581565e-05,
+      "loss": 0.662327229976654,
+      "step": 3108
+    },
+    {
+      "epoch": 1.341747572815534,
+      "grad_norm": 1.1382031440734863,
+      "learning_rate": 9.411467689448777e-05,
+      "loss": 0.6979284882545471,
+      "step": 3110
+    },
+    {
+      "epoch": 1.3426105717367853,
+      "grad_norm": 1.1555129289627075,
+      "learning_rate": 9.410311435510444e-05,
+      "loss": 0.7032427191734314,
+      "step": 3112
+    },
+    {
+      "epoch": 1.3434735706580367,
+      "grad_norm": 1.147678017616272,
+      "learning_rate": 9.409154118045389e-05,
+      "loss": 0.6282536387443542,
+      "step": 3114
+    },
+    {
+      "epoch": 1.344336569579288,
+      "grad_norm": 1.0925829410552979,
+      "learning_rate": 9.407995737332696e-05,
+      "loss": 0.6572934985160828,
+      "step": 3116
+    },
+    {
+      "epoch": 1.3451995685005393,
+      "grad_norm": 1.3632638454437256,
+      "learning_rate": 9.4068362936517e-05,
+      "loss": 0.7233583927154541,
+      "step": 3118
+    },
+    {
+      "epoch": 1.3460625674217908,
+      "grad_norm": 1.3994121551513672,
+      "learning_rate": 9.405675787282e-05,
+      "loss": 0.6465377807617188,
+      "step": 3120
+    },
+    {
+      "epoch": 1.346925566343042,
+      "grad_norm": 1.1862388849258423,
+      "learning_rate": 9.404514218503444e-05,
+      "loss": 0.6348391771316528,
+      "step": 3122
+    },
+    {
+      "epoch": 1.3477885652642934,
+      "grad_norm": 1.1724227666854858,
+      "learning_rate": 9.403351587596141e-05,
+      "loss": 0.6559093594551086,
+      "step": 3124
+    },
+    {
+      "epoch": 1.3486515641855448,
+      "grad_norm": 1.175567865371704,
+      "learning_rate": 9.402187894840452e-05,
+      "loss": 0.6693678498268127,
+      "step": 3126
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 1.1273375749588013,
+      "learning_rate": 9.401023140516999e-05,
+      "loss": 0.6766137480735779,
+      "step": 3128
+    },
+    {
+      "epoch": 1.3503775620280476,
+      "grad_norm": 1.2515393495559692,
+      "learning_rate": 9.399857324906652e-05,
+      "loss": 0.7133797407150269,
+      "step": 3130
+    },
+    {
+      "epoch": 1.3512405609492988,
+      "grad_norm": 1.1575038433074951,
+      "learning_rate": 9.398690448290547e-05,
+      "loss": 0.6572166681289673,
+      "step": 3132
+    },
+    {
+      "epoch": 1.3521035598705502,
+      "grad_norm": 1.2350742816925049,
+      "learning_rate": 9.397522510950072e-05,
+      "loss": 0.6389561891555786,
+      "step": 3134
+    },
+    {
+      "epoch": 1.3529665587918016,
+      "grad_norm": 0.9648056626319885,
+      "learning_rate": 9.396353513166867e-05,
+      "loss": 0.5810164213180542,
+      "step": 3136
+    },
+    {
+      "epoch": 1.3538295577130528,
+      "grad_norm": 1.147218108177185,
+      "learning_rate": 9.39518345522283e-05,
+      "loss": 0.7157397866249084,
+      "step": 3138
+    },
+    {
+      "epoch": 1.3546925566343042,
+      "grad_norm": 1.1948896646499634,
+      "learning_rate": 9.394012337400117e-05,
+      "loss": 0.6749072670936584,
+      "step": 3140
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 1.0945216417312622,
+      "learning_rate": 9.392840159981136e-05,
+      "loss": 0.63704514503479,
+      "step": 3142
+    },
+    {
+      "epoch": 1.3564185544768068,
+      "grad_norm": 1.0654480457305908,
+      "learning_rate": 9.391666923248555e-05,
+      "loss": 0.6804014444351196,
+      "step": 3144
+    },
+    {
+      "epoch": 1.3572815533980582,
+      "grad_norm": 1.3977962732315063,
+      "learning_rate": 9.390492627485292e-05,
+      "loss": 0.7055573463439941,
+      "step": 3146
+    },
+    {
+      "epoch": 1.3581445523193096,
+      "grad_norm": 1.1607128381729126,
+      "learning_rate": 9.389317272974526e-05,
+      "loss": 0.7383730411529541,
+      "step": 3148
+    },
+    {
+      "epoch": 1.3590075512405608,
+      "grad_norm": 1.192213773727417,
+      "learning_rate": 9.388140859999686e-05,
+      "loss": 0.7339736223220825,
+      "step": 3150
+    },
+    {
+      "epoch": 1.3598705501618122,
+      "grad_norm": 1.1597269773483276,
+      "learning_rate": 9.386963388844459e-05,
+      "loss": 0.6564369797706604,
+      "step": 3152
+    },
+    {
+      "epoch": 1.3607335490830637,
+      "grad_norm": 1.2374637126922607,
+      "learning_rate": 9.385784859792788e-05,
+      "loss": 0.7393131256103516,
+      "step": 3154
+    },
+    {
+      "epoch": 1.361596548004315,
+      "grad_norm": 1.1975921392440796,
+      "learning_rate": 9.384605273128868e-05,
+      "loss": 0.6794763207435608,
+      "step": 3156
+    },
+    {
+      "epoch": 1.3624595469255665,
+      "grad_norm": 1.158282995223999,
+      "learning_rate": 9.383424629137154e-05,
+      "loss": 0.5906237363815308,
+      "step": 3158
+    },
+    {
+      "epoch": 1.3633225458468177,
+      "grad_norm": 1.0605288743972778,
+      "learning_rate": 9.38224292810235e-05,
+      "loss": 0.6813581585884094,
+      "step": 3160
+    },
+    {
+      "epoch": 1.364185544768069,
+      "grad_norm": 0.9797501564025879,
+      "learning_rate": 9.38106017030942e-05,
+      "loss": 0.639259934425354,
+      "step": 3162
+    },
+    {
+      "epoch": 1.3650485436893205,
+      "grad_norm": 1.1023533344268799,
+      "learning_rate": 9.379876356043578e-05,
+      "loss": 0.667378306388855,
+      "step": 3164
+    },
+    {
+      "epoch": 1.3659115426105717,
+      "grad_norm": 1.1105194091796875,
+      "learning_rate": 9.378691485590299e-05,
+      "loss": 0.671601414680481,
+      "step": 3166
+    },
+    {
+      "epoch": 1.366774541531823,
+      "grad_norm": 0.9441736936569214,
+      "learning_rate": 9.377505559235308e-05,
+      "loss": 0.623394787311554,
+      "step": 3168
+    },
+    {
+      "epoch": 1.3676375404530745,
+      "grad_norm": 1.308560848236084,
+      "learning_rate": 9.376318577264585e-05,
+      "loss": 0.6558915972709656,
+      "step": 3170
+    },
+    {
+      "epoch": 1.3685005393743257,
+      "grad_norm": 1.065693974494934,
+      "learning_rate": 9.375130539964363e-05,
+      "loss": 0.6431660652160645,
+      "step": 3172
+    },
+    {
+      "epoch": 1.369363538295577,
+      "grad_norm": 1.027817726135254,
+      "learning_rate": 9.373941447621137e-05,
+      "loss": 0.6541599631309509,
+      "step": 3174
+    },
+    {
+      "epoch": 1.3702265372168285,
+      "grad_norm": 1.1651864051818848,
+      "learning_rate": 9.372751300521649e-05,
+      "loss": 0.6710350513458252,
+      "step": 3176
+    },
+    {
+      "epoch": 1.3710895361380797,
+      "grad_norm": 1.1066170930862427,
+      "learning_rate": 9.371560098952896e-05,
+      "loss": 0.6525241732597351,
+      "step": 3178
+    },
+    {
+      "epoch": 1.3719525350593311,
+      "grad_norm": 1.0942546129226685,
+      "learning_rate": 9.370367843202133e-05,
+      "loss": 0.6362038850784302,
+      "step": 3180
+    },
+    {
+      "epoch": 1.3728155339805825,
+      "grad_norm": 1.175432562828064,
+      "learning_rate": 9.369174533556868e-05,
+      "loss": 0.6017009615898132,
+      "step": 3182
+    },
+    {
+      "epoch": 1.373678532901834,
+      "grad_norm": 1.3605095148086548,
+      "learning_rate": 9.367980170304857e-05,
+      "loss": 0.6853877305984497,
+      "step": 3184
+    },
+    {
+      "epoch": 1.3745415318230851,
+      "grad_norm": 1.224565029144287,
+      "learning_rate": 9.366784753734121e-05,
+      "loss": 0.6434065699577332,
+      "step": 3186
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 1.1151230335235596,
+      "learning_rate": 9.365588284132927e-05,
+      "loss": 0.596630334854126,
+      "step": 3188
+    },
+    {
+      "epoch": 1.376267529665588,
+      "grad_norm": 1.2674700021743774,
+      "learning_rate": 9.364390761789799e-05,
+      "loss": 0.6716568470001221,
+      "step": 3190
+    },
+    {
+      "epoch": 1.3771305285868394,
+      "grad_norm": 1.3221330642700195,
+      "learning_rate": 9.363192186993512e-05,
+      "loss": 0.6230685710906982,
+      "step": 3192
+    },
+    {
+      "epoch": 1.3779935275080906,
+      "grad_norm": 1.2978439331054688,
+      "learning_rate": 9.3619925600331e-05,
+      "loss": 0.6495190858840942,
+      "step": 3194
+    },
+    {
+      "epoch": 1.378856526429342,
+      "grad_norm": 1.1542012691497803,
+      "learning_rate": 9.360791881197846e-05,
+      "loss": 0.6534912586212158,
+      "step": 3196
+    },
+    {
+      "epoch": 1.3797195253505934,
+      "grad_norm": 1.1294280290603638,
+      "learning_rate": 9.359590150777288e-05,
+      "loss": 0.5706549286842346,
+      "step": 3198
+    },
+    {
+      "epoch": 1.3805825242718446,
+      "grad_norm": 1.0622307062149048,
+      "learning_rate": 9.358387369061218e-05,
+      "loss": 0.6193811893463135,
+      "step": 3200
+    },
+    {
+      "epoch": 1.3805825242718446,
+      "eval_loss": 0.7103806734085083,
+      "eval_runtime": 831.4047,
+      "eval_samples_per_second": 2.478,
+      "eval_steps_per_second": 2.478,
+      "step": 3200
+    },
+    {
+      "epoch": 1.381445523193096,
+      "grad_norm": 1.1722023487091064,
+      "learning_rate": 9.357183536339683e-05,
+      "loss": 0.6021195650100708,
+      "step": 3202
+    },
+    {
+      "epoch": 1.3823085221143474,
+      "grad_norm": 1.2733373641967773,
+      "learning_rate": 9.355978652902978e-05,
+      "loss": 0.7195428013801575,
+      "step": 3204
+    },
+    {
+      "epoch": 1.3831715210355986,
+      "grad_norm": 1.019609808921814,
+      "learning_rate": 9.354772719041658e-05,
+      "loss": 0.6314851641654968,
+      "step": 3206
+    },
+    {
+      "epoch": 1.38403451995685,
+      "grad_norm": 1.36198890209198,
+      "learning_rate": 9.35356573504653e-05,
+      "loss": 0.722943902015686,
+      "step": 3208
+    },
+    {
+      "epoch": 1.3848975188781014,
+      "grad_norm": 1.1291067600250244,
+      "learning_rate": 9.35235770120865e-05,
+      "loss": 0.7058439254760742,
+      "step": 3210
+    },
+    {
+      "epoch": 1.3857605177993528,
+      "grad_norm": 1.2160366773605347,
+      "learning_rate": 9.351148617819332e-05,
+      "loss": 0.6186655759811401,
+      "step": 3212
+    },
+    {
+      "epoch": 1.386623516720604,
+      "grad_norm": 1.1663079261779785,
+      "learning_rate": 9.349938485170139e-05,
+      "loss": 0.5912084579467773,
+      "step": 3214
+    },
+    {
+      "epoch": 1.3874865156418554,
+      "grad_norm": 1.2145130634307861,
+      "learning_rate": 9.34872730355289e-05,
+      "loss": 0.6494051218032837,
+      "step": 3216
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 1.018477439880371,
+      "learning_rate": 9.347515073259657e-05,
+      "loss": 0.6258159279823303,
+      "step": 3218
+    },
+    {
+      "epoch": 1.3892125134843583,
+      "grad_norm": 1.2019606828689575,
+      "learning_rate": 9.346301794582761e-05,
+      "loss": 0.7076614499092102,
+      "step": 3220
+    },
+    {
+      "epoch": 1.3900755124056094,
+      "grad_norm": 1.072648048400879,
+      "learning_rate": 9.345087467814782e-05,
+      "loss": 0.5958849191665649,
+      "step": 3222
+    },
+    {
+      "epoch": 1.3909385113268609,
+      "grad_norm": 1.087908387184143,
+      "learning_rate": 9.343872093248546e-05,
+      "loss": 0.5659268498420715,
+      "step": 3224
+    },
+    {
+      "epoch": 1.3918015102481123,
+      "grad_norm": 1.2628878355026245,
+      "learning_rate": 9.342655671177138e-05,
+      "loss": 0.6369889974594116,
+      "step": 3226
+    },
+    {
+      "epoch": 1.3926645091693635,
+      "grad_norm": 1.1580097675323486,
+      "learning_rate": 9.341438201893891e-05,
+      "loss": 0.64048171043396,
+      "step": 3228
+    },
+    {
+      "epoch": 1.3935275080906149,
+      "grad_norm": 1.2230379581451416,
+      "learning_rate": 9.340219685692395e-05,
+      "loss": 0.7039018273353577,
+      "step": 3230
+    },
+    {
+      "epoch": 1.3943905070118663,
+      "grad_norm": 1.1932404041290283,
+      "learning_rate": 9.339000122866486e-05,
+      "loss": 0.6461225152015686,
+      "step": 3232
+    },
+    {
+      "epoch": 1.3952535059331175,
+      "grad_norm": 1.4337340593338013,
+      "learning_rate": 9.337779513710257e-05,
+      "loss": 0.6538844108581543,
+      "step": 3234
+    },
+    {
+      "epoch": 1.396116504854369,
+      "grad_norm": 1.2300320863723755,
+      "learning_rate": 9.336557858518055e-05,
+      "loss": 0.6788939237594604,
+      "step": 3236
+    },
+    {
+      "epoch": 1.3969795037756203,
+      "grad_norm": 1.1031018495559692,
+      "learning_rate": 9.335335157584473e-05,
+      "loss": 0.6180470585823059,
+      "step": 3238
+    },
+    {
+      "epoch": 1.3978425026968715,
+      "grad_norm": 1.1876481771469116,
+      "learning_rate": 9.334111411204363e-05,
+      "loss": 0.7097930908203125,
+      "step": 3240
+    },
+    {
+      "epoch": 1.398705501618123,
+      "grad_norm": 1.1510672569274902,
+      "learning_rate": 9.332886619672823e-05,
+      "loss": 0.6521070003509521,
+      "step": 3242
+    },
+    {
+      "epoch": 1.3995685005393743,
+      "grad_norm": 1.2715339660644531,
+      "learning_rate": 9.331660783285208e-05,
+      "loss": 0.6738741397857666,
+      "step": 3244
+    },
+    {
+      "epoch": 1.4004314994606257,
+      "grad_norm": 1.1528351306915283,
+      "learning_rate": 9.330433902337124e-05,
+      "loss": 0.6261277794837952,
+      "step": 3246
+    },
+    {
+      "epoch": 1.4012944983818771,
+      "grad_norm": 1.1715319156646729,
+      "learning_rate": 9.329205977124424e-05,
+      "loss": 0.7071060538291931,
+      "step": 3248
+    },
+    {
+      "epoch": 1.4021574973031283,
+      "grad_norm": 1.1527135372161865,
+      "learning_rate": 9.32797700794322e-05,
+      "loss": 0.6339676380157471,
+      "step": 3250
+    },
+    {
+      "epoch": 1.4030204962243797,
+      "grad_norm": 1.1546231508255005,
+      "learning_rate": 9.32674699508987e-05,
+      "loss": 0.6572067737579346,
+      "step": 3252
+    },
+    {
+      "epoch": 1.4038834951456312,
+      "grad_norm": 1.1085712909698486,
+      "learning_rate": 9.325515938860988e-05,
+      "loss": 0.6182453036308289,
+      "step": 3254
+    },
+    {
+      "epoch": 1.4047464940668823,
+      "grad_norm": 1.2346831560134888,
+      "learning_rate": 9.324283839553437e-05,
+      "loss": 0.6558552980422974,
+      "step": 3256
+    },
+    {
+      "epoch": 1.4056094929881338,
+      "grad_norm": 1.1873912811279297,
+      "learning_rate": 9.323050697464331e-05,
+      "loss": 0.6406651735305786,
+      "step": 3258
+    },
+    {
+      "epoch": 1.4064724919093852,
+      "grad_norm": 1.5326859951019287,
+      "learning_rate": 9.32181651289104e-05,
+      "loss": 0.6869475841522217,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4073354908306364,
+      "grad_norm": 1.2141598463058472,
+      "learning_rate": 9.320581286131178e-05,
+      "loss": 0.6805770397186279,
+      "step": 3262
+    },
+    {
+      "epoch": 1.4081984897518878,
+      "grad_norm": 1.232690453529358,
+      "learning_rate": 9.319345017482615e-05,
+      "loss": 0.6193398833274841,
+      "step": 3264
+    },
+    {
+      "epoch": 1.4090614886731392,
+      "grad_norm": 1.0616317987442017,
+      "learning_rate": 9.318107707243475e-05,
+      "loss": 0.5868768095970154,
+      "step": 3266
+    },
+    {
+      "epoch": 1.4099244875943904,
+      "grad_norm": 1.091909408569336,
+      "learning_rate": 9.316869355712127e-05,
+      "loss": 0.6360217928886414,
+      "step": 3268
+    },
+    {
+      "epoch": 1.4107874865156418,
+      "grad_norm": 1.0247875452041626,
+      "learning_rate": 9.315629963187194e-05,
+      "loss": 0.6082828044891357,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4116504854368932,
+      "grad_norm": 1.2529716491699219,
+      "learning_rate": 9.314389529967551e-05,
+      "loss": 0.6340673565864563,
+      "step": 3272
+    },
+    {
+      "epoch": 1.4125134843581446,
+      "grad_norm": 0.9816965460777283,
+      "learning_rate": 9.31314805635232e-05,
+      "loss": 0.6333131790161133,
+      "step": 3274
+    },
+    {
+      "epoch": 1.413376483279396,
+      "grad_norm": 1.2140034437179565,
+      "learning_rate": 9.311905542640882e-05,
+      "loss": 0.7256702780723572,
+      "step": 3276
+    },
+    {
+      "epoch": 1.4142394822006472,
+      "grad_norm": 1.2171046733856201,
+      "learning_rate": 9.310661989132859e-05,
+      "loss": 0.7062730193138123,
+      "step": 3278
+    },
+    {
+      "epoch": 1.4151024811218986,
+      "grad_norm": 1.0224506855010986,
+      "learning_rate": 9.309417396128131e-05,
+      "loss": 0.6192088723182678,
+      "step": 3280
+    },
+    {
+      "epoch": 1.41596548004315,
+      "grad_norm": 1.1904140710830688,
+      "learning_rate": 9.308171763926823e-05,
+      "loss": 0.6522074341773987,
+      "step": 3282
+    },
+    {
+      "epoch": 1.4168284789644012,
+      "grad_norm": 1.1734132766723633,
+      "learning_rate": 9.306925092829317e-05,
+      "loss": 0.6472417712211609,
+      "step": 3284
+    },
+    {
+      "epoch": 1.4176914778856526,
+      "grad_norm": 1.2349748611450195,
+      "learning_rate": 9.30567738313624e-05,
+      "loss": 0.6990411877632141,
+      "step": 3286
+    },
+    {
+      "epoch": 1.418554476806904,
+      "grad_norm": 1.1732749938964844,
+      "learning_rate": 9.304428635148472e-05,
+      "loss": 0.6181937456130981,
+      "step": 3288
+    },
+    {
+      "epoch": 1.4194174757281552,
+      "grad_norm": 1.3659436702728271,
+      "learning_rate": 9.303178849167142e-05,
+      "loss": 0.7471824288368225,
+      "step": 3290
+    },
+    {
+      "epoch": 1.4202804746494067,
+      "grad_norm": 1.0865631103515625,
+      "learning_rate": 9.301928025493634e-05,
+      "loss": 0.5541732907295227,
+      "step": 3292
+    },
+    {
+      "epoch": 1.421143473570658,
+      "grad_norm": 1.2114793062210083,
+      "learning_rate": 9.300676164429574e-05,
+      "loss": 0.6934989094734192,
+      "step": 3294
+    },
+    {
+      "epoch": 1.4220064724919093,
+      "grad_norm": 1.1783509254455566,
+      "learning_rate": 9.299423266276846e-05,
+      "loss": 0.6887713670730591,
+      "step": 3296
+    },
+    {
+      "epoch": 1.4228694714131607,
+      "grad_norm": 1.3001595735549927,
+      "learning_rate": 9.298169331337577e-05,
+      "loss": 0.6956127882003784,
+      "step": 3298
+    },
+    {
+      "epoch": 1.423732470334412,
+      "grad_norm": 1.1657462120056152,
+      "learning_rate": 9.296914359914151e-05,
+      "loss": 0.690113365650177,
+      "step": 3300
+    },
+    {
+      "epoch": 1.423732470334412,
+      "eval_loss": 0.7086145281791687,
+      "eval_runtime": 838.887,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 3300
+    },
+    {
+      "epoch": 1.4245954692556635,
+      "grad_norm": 1.1632269620895386,
+      "learning_rate": 9.295658352309198e-05,
+      "loss": 0.6801355481147766,
+      "step": 3302
+    },
+    {
+      "epoch": 1.425458468176915,
+      "grad_norm": 1.3355532884597778,
+      "learning_rate": 9.2944013088256e-05,
+      "loss": 0.6801818013191223,
+      "step": 3304
+    },
+    {
+      "epoch": 1.426321467098166,
+      "grad_norm": 1.0911835432052612,
+      "learning_rate": 9.293143229766484e-05,
+      "loss": 0.633465051651001,
+      "step": 3306
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 1.1873823404312134,
+      "learning_rate": 9.29188411543523e-05,
+      "loss": 0.6677361130714417,
+      "step": 3308
+    },
+    {
+      "epoch": 1.428047464940669,
+      "grad_norm": 1.2115873098373413,
+      "learning_rate": 9.290623966135472e-05,
+      "loss": 0.6741982102394104,
+      "step": 3310
+    },
+    {
+      "epoch": 1.42891046386192,
+      "grad_norm": 1.2389469146728516,
+      "learning_rate": 9.289362782171084e-05,
+      "loss": 0.7169438600540161,
+      "step": 3312
+    },
+    {
+      "epoch": 1.4297734627831715,
+      "grad_norm": 1.2088500261306763,
+      "learning_rate": 9.288100563846201e-05,
+      "loss": 0.7112365961074829,
+      "step": 3314
+    },
+    {
+      "epoch": 1.430636461704423,
+      "grad_norm": 1.1279292106628418,
+      "learning_rate": 9.286837311465196e-05,
+      "loss": 0.6910304427146912,
+      "step": 3316
+    },
+    {
+      "epoch": 1.4314994606256741,
+      "grad_norm": 1.1812585592269897,
+      "learning_rate": 9.285573025332698e-05,
+      "loss": 0.7288362979888916,
+      "step": 3318
+    },
+    {
+      "epoch": 1.4323624595469255,
+      "grad_norm": 1.5522732734680176,
+      "learning_rate": 9.284307705753587e-05,
+      "loss": 0.6918224692344666,
+      "step": 3320
+    },
+    {
+      "epoch": 1.433225458468177,
+      "grad_norm": 1.2825061082839966,
+      "learning_rate": 9.283041353032984e-05,
+      "loss": 0.6756533980369568,
+      "step": 3322
+    },
+    {
+      "epoch": 1.4340884573894281,
+      "grad_norm": 1.0962797403335571,
+      "learning_rate": 9.281773967476269e-05,
+      "loss": 0.6005351543426514,
+      "step": 3324
+    },
+    {
+      "epoch": 1.4349514563106796,
+      "grad_norm": 1.249868631362915,
+      "learning_rate": 9.280505549389063e-05,
+      "loss": 0.6576942205429077,
+      "step": 3326
+    },
+    {
+      "epoch": 1.435814455231931,
+      "grad_norm": 1.2587590217590332,
+      "learning_rate": 9.279236099077239e-05,
+      "loss": 0.6686707735061646,
+      "step": 3328
+    },
+    {
+      "epoch": 1.4366774541531824,
+      "grad_norm": 1.0961607694625854,
+      "learning_rate": 9.277965616846922e-05,
+      "loss": 0.5858182311058044,
+      "step": 3330
+    },
+    {
+      "epoch": 1.4375404530744338,
+      "grad_norm": 1.199666142463684,
+      "learning_rate": 9.276694103004481e-05,
+      "loss": 0.6661836504936218,
+      "step": 3332
+    },
+    {
+      "epoch": 1.438403451995685,
+      "grad_norm": 1.2444990873336792,
+      "learning_rate": 9.275421557856536e-05,
+      "loss": 0.6922421455383301,
+      "step": 3334
+    },
+    {
+      "epoch": 1.4392664509169364,
+      "grad_norm": 1.3198295831680298,
+      "learning_rate": 9.274147981709957e-05,
+      "loss": 0.750586986541748,
+      "step": 3336
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 1.0782018899917603,
+      "learning_rate": 9.27287337487186e-05,
+      "loss": 0.6817078590393066,
+      "step": 3338
+    },
+    {
+      "epoch": 1.440992448759439,
+      "grad_norm": 1.1463382244110107,
+      "learning_rate": 9.271597737649611e-05,
+      "loss": 0.6163688898086548,
+      "step": 3340
+    },
+    {
+      "epoch": 1.4418554476806904,
+      "grad_norm": 1.0618164539337158,
+      "learning_rate": 9.270321070350825e-05,
+      "loss": 0.6624670028686523,
+      "step": 3342
+    },
+    {
+      "epoch": 1.4427184466019418,
+      "grad_norm": 1.1389427185058594,
+      "learning_rate": 9.269043373283362e-05,
+      "loss": 0.6802749037742615,
+      "step": 3344
+    },
+    {
+      "epoch": 1.443581445523193,
+      "grad_norm": 1.082173466682434,
+      "learning_rate": 9.267764646755335e-05,
+      "loss": 0.6428707838058472,
+      "step": 3346
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 1.1741442680358887,
+      "learning_rate": 9.266484891075103e-05,
+      "loss": 0.6853904128074646,
+      "step": 3348
+    },
+    {
+      "epoch": 1.4453074433656958,
+      "grad_norm": 1.1328803300857544,
+      "learning_rate": 9.265204106551273e-05,
+      "loss": 0.6889524459838867,
+      "step": 3350
+    },
+    {
+      "epoch": 1.446170442286947,
+      "grad_norm": 1.0607874393463135,
+      "learning_rate": 9.2639222934927e-05,
+      "loss": 0.6037911176681519,
+      "step": 3352
+    },
+    {
+      "epoch": 1.4470334412081984,
+      "grad_norm": 1.1584129333496094,
+      "learning_rate": 9.262639452208487e-05,
+      "loss": 0.6671991944313049,
+      "step": 3354
+    },
+    {
+      "epoch": 1.4478964401294498,
+      "grad_norm": 1.1881403923034668,
+      "learning_rate": 9.261355583007987e-05,
+      "loss": 0.7165482044219971,
+      "step": 3356
+    },
+    {
+      "epoch": 1.4487594390507013,
+      "grad_norm": 1.2093054056167603,
+      "learning_rate": 9.260070686200797e-05,
+      "loss": 0.7108280062675476,
+      "step": 3358
+    },
+    {
+      "epoch": 1.4496224379719527,
+      "grad_norm": 1.1105186939239502,
+      "learning_rate": 9.258784762096768e-05,
+      "loss": 0.718152642250061,
+      "step": 3360
+    },
+    {
+      "epoch": 1.4504854368932039,
+      "grad_norm": 1.028255820274353,
+      "learning_rate": 9.25749781100599e-05,
+      "loss": 0.6449500322341919,
+      "step": 3362
+    },
+    {
+      "epoch": 1.4513484358144553,
+      "grad_norm": 1.235719919204712,
+      "learning_rate": 9.25620983323881e-05,
+      "loss": 0.6895458102226257,
+      "step": 3364
+    },
+    {
+      "epoch": 1.4522114347357067,
+      "grad_norm": 1.3693572282791138,
+      "learning_rate": 9.254920829105816e-05,
+      "loss": 0.7243028879165649,
+      "step": 3366
+    },
+    {
+      "epoch": 1.4530744336569579,
+      "grad_norm": 1.1795028448104858,
+      "learning_rate": 9.253630798917843e-05,
+      "loss": 0.6177746057510376,
+      "step": 3368
+    },
+    {
+      "epoch": 1.4539374325782093,
+      "grad_norm": 1.0759676694869995,
+      "learning_rate": 9.25233974298598e-05,
+      "loss": 0.6603507399559021,
+      "step": 3370
+    },
+    {
+      "epoch": 1.4548004314994607,
+      "grad_norm": 0.9495601058006287,
+      "learning_rate": 9.251047661621556e-05,
+      "loss": 0.5989331603050232,
+      "step": 3372
+    },
+    {
+      "epoch": 1.455663430420712,
+      "grad_norm": 1.1334787607192993,
+      "learning_rate": 9.249754555136154e-05,
+      "loss": 0.5743664503097534,
+      "step": 3374
+    },
+    {
+      "epoch": 1.4565264293419633,
+      "grad_norm": 1.114450216293335,
+      "learning_rate": 9.248460423841598e-05,
+      "loss": 0.5905263423919678,
+      "step": 3376
+    },
+    {
+      "epoch": 1.4573894282632147,
+      "grad_norm": 1.2400883436203003,
+      "learning_rate": 9.247165268049963e-05,
+      "loss": 0.6751091480255127,
+      "step": 3378
+    },
+    {
+      "epoch": 1.458252427184466,
+      "grad_norm": 1.2845438718795776,
+      "learning_rate": 9.245869088073569e-05,
+      "loss": 0.6872944235801697,
+      "step": 3380
+    },
+    {
+      "epoch": 1.4591154261057173,
+      "grad_norm": 1.054388165473938,
+      "learning_rate": 9.244571884224987e-05,
+      "loss": 0.6467806100845337,
+      "step": 3382
+    },
+    {
+      "epoch": 1.4599784250269687,
+      "grad_norm": 1.0959769487380981,
+      "learning_rate": 9.243273656817029e-05,
+      "loss": 0.6486450433731079,
+      "step": 3384
+    },
+    {
+      "epoch": 1.4608414239482201,
+      "grad_norm": 1.6487104892730713,
+      "learning_rate": 9.241974406162756e-05,
+      "loss": 0.7024005651473999,
+      "step": 3386
+    },
+    {
+      "epoch": 1.4617044228694713,
+      "grad_norm": 1.3018934726715088,
+      "learning_rate": 9.240674132575481e-05,
+      "loss": 0.7081685662269592,
+      "step": 3388
+    },
+    {
+      "epoch": 1.4625674217907227,
+      "grad_norm": 0.9295057058334351,
+      "learning_rate": 9.239372836368754e-05,
+      "loss": 0.5842618346214294,
+      "step": 3390
+    },
+    {
+      "epoch": 1.4634304207119742,
+      "grad_norm": 1.0556540489196777,
+      "learning_rate": 9.238070517856378e-05,
+      "loss": 0.6355351209640503,
+      "step": 3392
+    },
+    {
+      "epoch": 1.4642934196332256,
+      "grad_norm": 1.2247962951660156,
+      "learning_rate": 9.236767177352403e-05,
+      "loss": 0.6340579390525818,
+      "step": 3394
+    },
+    {
+      "epoch": 1.4651564185544768,
+      "grad_norm": 1.1625045537948608,
+      "learning_rate": 9.235462815171123e-05,
+      "loss": 0.6405525803565979,
+      "step": 3396
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 1.009028673171997,
+      "learning_rate": 9.234157431627077e-05,
+      "loss": 0.5884681940078735,
+      "step": 3398
+    },
+    {
+      "epoch": 1.4668824163969796,
+      "grad_norm": 1.174851894378662,
+      "learning_rate": 9.232851027035056e-05,
+      "loss": 0.6532622575759888,
+      "step": 3400
+    },
+    {
+      "epoch": 1.4668824163969796,
+      "eval_loss": 0.7067069411277771,
+      "eval_runtime": 829.9411,
+      "eval_samples_per_second": 2.482,
+      "eval_steps_per_second": 2.482,
+      "step": 3400
+    },
+    {
+      "epoch": 1.4677454153182308,
+      "grad_norm": 0.9843166470527649,
+      "learning_rate": 9.231543601710091e-05,
+      "loss": 0.6505036354064941,
+      "step": 3402
+    },
+    {
+      "epoch": 1.4686084142394822,
+      "grad_norm": 1.1260782480239868,
+      "learning_rate": 9.230235155967461e-05,
+      "loss": 0.5809159874916077,
+      "step": 3404
+    },
+    {
+      "epoch": 1.4694714131607336,
+      "grad_norm": 1.3986146450042725,
+      "learning_rate": 9.228925690122696e-05,
+      "loss": 0.6431854367256165,
+      "step": 3406
+    },
+    {
+      "epoch": 1.4703344120819848,
+      "grad_norm": 1.145108938217163,
+      "learning_rate": 9.227615204491562e-05,
+      "loss": 0.6126174330711365,
+      "step": 3408
+    },
+    {
+      "epoch": 1.4711974110032362,
+      "grad_norm": 1.0973509550094604,
+      "learning_rate": 9.22630369939008e-05,
+      "loss": 0.6271735429763794,
+      "step": 3410
+    },
+    {
+      "epoch": 1.4720604099244876,
+      "grad_norm": 1.0548614263534546,
+      "learning_rate": 9.224991175134515e-05,
+      "loss": 0.5894950032234192,
+      "step": 3412
+    },
+    {
+      "epoch": 1.472923408845739,
+      "grad_norm": 1.3288220167160034,
+      "learning_rate": 9.223677632041374e-05,
+      "loss": 0.6621272563934326,
+      "step": 3414
+    },
+    {
+      "epoch": 1.4737864077669902,
+      "grad_norm": 1.0312891006469727,
+      "learning_rate": 9.222363070427411e-05,
+      "loss": 0.6911466717720032,
+      "step": 3416
+    },
+    {
+      "epoch": 1.4746494066882416,
+      "grad_norm": 1.1039631366729736,
+      "learning_rate": 9.221047490609629e-05,
+      "loss": 0.5945214033126831,
+      "step": 3418
+    },
+    {
+      "epoch": 1.475512405609493,
+      "grad_norm": 1.1555835008621216,
+      "learning_rate": 9.219730892905273e-05,
+      "loss": 0.6231427192687988,
+      "step": 3420
+    },
+    {
+      "epoch": 1.4763754045307445,
+      "grad_norm": 1.0504940748214722,
+      "learning_rate": 9.218413277631835e-05,
+      "loss": 0.6413036584854126,
+      "step": 3422
+    },
+    {
+      "epoch": 1.4772384034519956,
+      "grad_norm": 1.0855275392532349,
+      "learning_rate": 9.217094645107053e-05,
+      "loss": 0.663884699344635,
+      "step": 3424
+    },
+    {
+      "epoch": 1.478101402373247,
+      "grad_norm": 1.068493366241455,
+      "learning_rate": 9.215774995648907e-05,
+      "loss": 0.6161345839500427,
+      "step": 3426
+    },
+    {
+      "epoch": 1.4789644012944985,
+      "grad_norm": 1.0962451696395874,
+      "learning_rate": 9.214454329575626e-05,
+      "loss": 0.6693077087402344,
+      "step": 3428
+    },
+    {
+      "epoch": 1.4798274002157497,
+      "grad_norm": 1.070582389831543,
+      "learning_rate": 9.213132647205682e-05,
+      "loss": 0.6362645030021667,
+      "step": 3430
+    },
+    {
+      "epoch": 1.480690399137001,
+      "grad_norm": 1.201202154159546,
+      "learning_rate": 9.211809948857793e-05,
+      "loss": 0.7378971576690674,
+      "step": 3432
+    },
+    {
+      "epoch": 1.4815533980582525,
+      "grad_norm": 1.1988805532455444,
+      "learning_rate": 9.210486234850921e-05,
+      "loss": 0.686621367931366,
+      "step": 3434
+    },
+    {
+      "epoch": 1.4824163969795037,
+      "grad_norm": 1.2886041402816772,
+      "learning_rate": 9.209161505504274e-05,
+      "loss": 0.6873248219490051,
+      "step": 3436
+    },
+    {
+      "epoch": 1.483279395900755,
+      "grad_norm": 1.0028132200241089,
+      "learning_rate": 9.207835761137307e-05,
+      "loss": 0.624991774559021,
+      "step": 3438
+    },
+    {
+      "epoch": 1.4841423948220065,
+      "grad_norm": 1.139201045036316,
+      "learning_rate": 9.206509002069715e-05,
+      "loss": 0.6064573526382446,
+      "step": 3440
+    },
+    {
+      "epoch": 1.4850053937432577,
+      "grad_norm": 1.0371732711791992,
+      "learning_rate": 9.205181228621438e-05,
+      "loss": 0.6139655709266663,
+      "step": 3442
+    },
+    {
+      "epoch": 1.485868392664509,
+      "grad_norm": 1.1208233833312988,
+      "learning_rate": 9.203852441112665e-05,
+      "loss": 0.6176325082778931,
+      "step": 3444
+    },
+    {
+      "epoch": 1.4867313915857605,
+      "grad_norm": 1.2438154220581055,
+      "learning_rate": 9.20252263986383e-05,
+      "loss": 0.6920754313468933,
+      "step": 3446
+    },
+    {
+      "epoch": 1.487594390507012,
+      "grad_norm": 1.2117916345596313,
+      "learning_rate": 9.201191825195604e-05,
+      "loss": 0.6859762072563171,
+      "step": 3448
+    },
+    {
+      "epoch": 1.4884573894282633,
+      "grad_norm": 1.2274118661880493,
+      "learning_rate": 9.199859997428908e-05,
+      "loss": 0.7358171939849854,
+      "step": 3450
+    },
+    {
+      "epoch": 1.4893203883495145,
+      "grad_norm": 1.2810527086257935,
+      "learning_rate": 9.198527156884907e-05,
+      "loss": 0.6878830194473267,
+      "step": 3452
+    },
+    {
+      "epoch": 1.490183387270766,
+      "grad_norm": 1.2344684600830078,
+      "learning_rate": 9.197193303885008e-05,
+      "loss": 0.6995986104011536,
+      "step": 3454
+    },
+    {
+      "epoch": 1.4910463861920173,
+      "grad_norm": 1.233686923980713,
+      "learning_rate": 9.195858438750866e-05,
+      "loss": 0.6423236131668091,
+      "step": 3456
+    },
+    {
+      "epoch": 1.4919093851132685,
+      "grad_norm": 1.3485876321792603,
+      "learning_rate": 9.194522561804377e-05,
+      "loss": 0.7513291835784912,
+      "step": 3458
+    },
+    {
+      "epoch": 1.49277238403452,
+      "grad_norm": 0.9913640022277832,
+      "learning_rate": 9.19318567336768e-05,
+      "loss": 0.6667183637619019,
+      "step": 3460
+    },
+    {
+      "epoch": 1.4936353829557714,
+      "grad_norm": 1.146099328994751,
+      "learning_rate": 9.191847773763161e-05,
+      "loss": 0.643630862236023,
+      "step": 3462
+    },
+    {
+      "epoch": 1.4944983818770226,
+      "grad_norm": 1.0300167798995972,
+      "learning_rate": 9.190508863313448e-05,
+      "loss": 0.655176043510437,
+      "step": 3464
+    },
+    {
+      "epoch": 1.495361380798274,
+      "grad_norm": 1.285008430480957,
+      "learning_rate": 9.189168942341416e-05,
+      "loss": 0.6792252659797668,
+      "step": 3466
+    },
+    {
+      "epoch": 1.4962243797195254,
+      "grad_norm": 1.1455490589141846,
+      "learning_rate": 9.187828011170175e-05,
+      "loss": 0.7081018686294556,
+      "step": 3468
+    },
+    {
+      "epoch": 1.4970873786407766,
+      "grad_norm": 1.2560410499572754,
+      "learning_rate": 9.186486070123091e-05,
+      "loss": 0.6613248586654663,
+      "step": 3470
+    },
+    {
+      "epoch": 1.497950377562028,
+      "grad_norm": 1.2391352653503418,
+      "learning_rate": 9.185143119523761e-05,
+      "loss": 0.6299360990524292,
+      "step": 3472
+    },
+    {
+      "epoch": 1.4988133764832794,
+      "grad_norm": 1.134585976600647,
+      "learning_rate": 9.183799159696035e-05,
+      "loss": 0.5910941362380981,
+      "step": 3474
+    },
+    {
+      "epoch": 1.4996763754045308,
+      "grad_norm": 1.1744983196258545,
+      "learning_rate": 9.182454190964005e-05,
+      "loss": 0.7077166438102722,
+      "step": 3476
+    },
+    {
+      "epoch": 1.5005393743257822,
+      "grad_norm": 1.1142816543579102,
+      "learning_rate": 9.181108213651999e-05,
+      "loss": 0.6168769598007202,
+      "step": 3478
+    },
+    {
+      "epoch": 1.5014023732470334,
+      "grad_norm": 1.1463894844055176,
+      "learning_rate": 9.179761228084597e-05,
+      "loss": 0.6587874293327332,
+      "step": 3480
+    },
+    {
+      "epoch": 1.5022653721682848,
+      "grad_norm": 1.3945616483688354,
+      "learning_rate": 9.178413234586616e-05,
+      "loss": 0.7696266770362854,
+      "step": 3482
+    },
+    {
+      "epoch": 1.5031283710895362,
+      "grad_norm": 1.3266242742538452,
+      "learning_rate": 9.177064233483121e-05,
+      "loss": 0.7142559885978699,
+      "step": 3484
+    },
+    {
+      "epoch": 1.5039913700107874,
+      "grad_norm": 1.1857004165649414,
+      "learning_rate": 9.175714225099417e-05,
+      "loss": 0.7085118293762207,
+      "step": 3486
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 1.1070430278778076,
+      "learning_rate": 9.174363209761049e-05,
+      "loss": 0.6643039584159851,
+      "step": 3488
+    },
+    {
+      "epoch": 1.5057173678532902,
+      "grad_norm": 1.0877243280410767,
+      "learning_rate": 9.173011187793814e-05,
+      "loss": 0.6282537579536438,
+      "step": 3490
+    },
+    {
+      "epoch": 1.5065803667745414,
+      "grad_norm": 1.2648247480392456,
+      "learning_rate": 9.171658159523741e-05,
+      "loss": 0.6811800599098206,
+      "step": 3492
+    },
+    {
+      "epoch": 1.5074433656957928,
+      "grad_norm": 1.1535227298736572,
+      "learning_rate": 9.170304125277111e-05,
+      "loss": 0.6442922353744507,
+      "step": 3494
+    },
+    {
+      "epoch": 1.5083063646170443,
+      "grad_norm": 1.1175323724746704,
+      "learning_rate": 9.16894908538044e-05,
+      "loss": 0.717490553855896,
+      "step": 3496
+    },
+    {
+      "epoch": 1.5091693635382954,
+      "grad_norm": 1.1923470497131348,
+      "learning_rate": 9.167593040160491e-05,
+      "loss": 0.6537191867828369,
+      "step": 3498
+    },
+    {
+      "epoch": 1.510032362459547,
+      "grad_norm": 1.3302042484283447,
+      "learning_rate": 9.166235989944268e-05,
+      "loss": 0.7314199805259705,
+      "step": 3500
+    },
+    {
+      "epoch": 1.510032362459547,
+      "eval_loss": 0.7047909498214722,
+      "eval_runtime": 821.8811,
+      "eval_samples_per_second": 2.506,
+      "eval_steps_per_second": 2.506,
+      "step": 3500
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.6469600804753183e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-3500/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-3500/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-3500/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/README.md b/sft_devstral_24B/checkpoints/checkpoint-4000/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-4000/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-4000/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ca75408787ff2ae72ab9835adad75b336bbad7a0
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01e0fe0810eaa9f99f82257d536602a4bd3b801ea597104efe093189374232b4
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-4000/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..bb294f1902e27f02bae4edb3af888dd2c6d9c134
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40c9ff9af5de27f6f1bb88703d6f620486a8e83c41ec592ed5f2631d1e961804
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-4000/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..3e27a5edf53c942f94ac540767dd719aa0ce38ab
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:991dc52763274f771544bb1435e987bfa7d76b2e6e5aa1054e2710d99f90777f
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-4000/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..41f0b8653464cddb7d98f9399407492422eb6e39
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d87db0a0bb48dbb8be0e1b140af067cf2e0b18d888a7a7295f1516be0b0b72ba
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-4000/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3b11370a6e406c24dde2c258c963c1902ae57ee
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/trainer_state.json
@@ -0,0 +1,14363 @@
+{
+  "best_global_step": 4000,
+  "best_metric": 0.6941319704055786,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-4000",
+  "epoch": 1.7257820927723841,
+  "eval_steps": 100,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4323624595469256,
+      "grad_norm": 1.5475730895996094,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.7443642616271973,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4332254584681769,
+      "grad_norm": 1.4621447324752808,
+      "learning_rate": 9.011680143755616e-05,
+      "loss": 0.8210510015487671,
+      "step": 1004
+    },
+    {
+      "epoch": 0.43408845738942825,
+      "grad_norm": 1.3651922941207886,
+      "learning_rate": 9.029649595687331e-05,
+      "loss": 0.8130766153335571,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4349514563106796,
+      "grad_norm": 1.3587357997894287,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.79145747423172,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 1.464652180671692,
+      "learning_rate": 9.065588499550764e-05,
+      "loss": 0.8286946415901184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4366774541531823,
+      "grad_norm": 1.3912625312805176,
+      "learning_rate": 9.083557951482479e-05,
+      "loss": 0.717319667339325,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4375404530744337,
+      "grad_norm": 1.2045491933822632,
+      "learning_rate": 9.101527403414196e-05,
+      "loss": 0.7294418215751648,
+      "step": 1014
+    },
+    {
+      "epoch": 0.438403451995685,
+      "grad_norm": 1.3449069261550903,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.7366700768470764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43926645091693634,
+      "grad_norm": 2.901384115219116,
+      "learning_rate": 9.137466307277629e-05,
+      "loss": 0.8551293611526489,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 1.6648688316345215,
+      "learning_rate": 9.155435759209344e-05,
+      "loss": 0.8944112658500671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44099244875943905,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 9.17340521114106e-05,
+      "loss": 0.7644214630126953,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4418554476806904,
+      "grad_norm": 1.4643231630325317,
+      "learning_rate": 9.191374663072777e-05,
+      "loss": 0.8644799590110779,
+      "step": 1024
+    },
+    {
+      "epoch": 0.44271844660194176,
+      "grad_norm": 1.4690274000167847,
+      "learning_rate": 9.209344115004492e-05,
+      "loss": 0.8281022906303406,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4435814455231931,
+      "grad_norm": 1.3321986198425293,
+      "learning_rate": 9.227313566936209e-05,
+      "loss": 0.7432141900062561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.8143638372421265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4453074433656958,
+      "grad_norm": 1.747295618057251,
+      "learning_rate": 9.263252470799642e-05,
+      "loss": 0.8102239370346069,
+      "step": 1032
+    },
+    {
+      "epoch": 0.44617044228694713,
+      "grad_norm": 1.3273073434829712,
+      "learning_rate": 9.281221922731357e-05,
+      "loss": 0.7768088579177856,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4470334412081985,
+      "grad_norm": 1.5019903182983398,
+      "learning_rate": 9.299191374663073e-05,
+      "loss": 0.7886938452720642,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44789644012944985,
+      "grad_norm": 1.6711527109146118,
+      "learning_rate": 9.31716082659479e-05,
+      "loss": 0.8152258396148682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 1.8504793643951416,
+      "learning_rate": 9.335130278526505e-05,
+      "loss": 0.8339349627494812,
+      "step": 1040
+    },
+    {
+      "epoch": 0.44962243797195256,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 9.353099730458221e-05,
+      "loss": 0.7798911929130554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.45048543689320386,
+      "grad_norm": 1.3361700773239136,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.7676728963851929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4513484358144552,
+      "grad_norm": 1.338934063911438,
+      "learning_rate": 9.389038634321654e-05,
+      "loss": 0.7729206085205078,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4522114347357066,
+      "grad_norm": 1.4347442388534546,
+      "learning_rate": 9.40700808625337e-05,
+      "loss": 0.7923640608787537,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 1.4451628923416138,
+      "learning_rate": 9.424977538185086e-05,
+      "loss": 0.8068996071815491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4539374325782093,
+      "grad_norm": 1.1904500722885132,
+      "learning_rate": 9.442946990116803e-05,
+      "loss": 0.7532910704612732,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45480043149946064,
+      "grad_norm": 1.4103976488113403,
+      "learning_rate": 9.460916442048518e-05,
+      "loss": 0.8236854672431946,
+      "step": 1054
+    },
+    {
+      "epoch": 0.455663430420712,
+      "grad_norm": 1.3762754201889038,
+      "learning_rate": 9.478885893980234e-05,
+      "loss": 0.7532716393470764,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4565264293419633,
+      "grad_norm": 1.461421012878418,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.8506031632423401,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 1.7207789421081543,
+      "learning_rate": 9.514824797843667e-05,
+      "loss": 0.7824664115905762,
+      "step": 1060
+    },
+    {
+      "epoch": 0.458252427184466,
+      "grad_norm": 1.5946974754333496,
+      "learning_rate": 9.532794249775382e-05,
+      "loss": 0.8674250245094299,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4591154261057174,
+      "grad_norm": 1.3552565574645996,
+      "learning_rate": 9.550763701707099e-05,
+      "loss": 0.7851845622062683,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45997842502696873,
+      "grad_norm": 1.414617657661438,
+      "learning_rate": 9.568733153638815e-05,
+      "loss": 0.8890700936317444,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4608414239482201,
+      "grad_norm": 1.3322749137878418,
+      "learning_rate": 9.58670260557053e-05,
+      "loss": 0.8226593732833862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 1.41420316696167,
+      "learning_rate": 9.604672057502247e-05,
+      "loss": 0.7663303017616272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.46256742179072274,
+      "grad_norm": 1.324967384338379,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.7122516632080078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4634304207119741,
+      "grad_norm": 1.2046154737472534,
+      "learning_rate": 9.640610961365679e-05,
+      "loss": 0.7177618145942688,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46429341963322546,
+      "grad_norm": 1.5221766233444214,
+      "learning_rate": 9.658580413297395e-05,
+      "loss": 0.7618930339813232,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4651564185544768,
+      "grad_norm": 1.3517179489135742,
+      "learning_rate": 9.67654986522911e-05,
+      "loss": 0.7483780384063721,
+      "step": 1078
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.747865080833435,
+      "learning_rate": 9.694519317160827e-05,
+      "loss": 0.8507670164108276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4668824163969795,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 9.712488769092543e-05,
+      "loss": 0.7893540859222412,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46774541531823083,
+      "grad_norm": 1.6290462017059326,
+      "learning_rate": 9.730458221024258e-05,
+      "loss": 0.7891272306442261,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4686084142394822,
+      "grad_norm": 1.6605802774429321,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.841018795967102,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46947141316073354,
+      "grad_norm": 1.388964295387268,
+      "learning_rate": 9.766397124887691e-05,
+      "loss": 0.8099003434181213,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 1.3901978731155396,
+      "learning_rate": 9.784366576819407e-05,
+      "loss": 0.7177658677101135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.47119741100323626,
+      "grad_norm": 1.3341423273086548,
+      "learning_rate": 9.802336028751123e-05,
+      "loss": 0.8285676836967468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4720604099244876,
+      "grad_norm": 1.5451914072036743,
+      "learning_rate": 9.82030548068284e-05,
+      "loss": 0.7582666873931885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47292340884573897,
+      "grad_norm": 1.5395640134811401,
+      "learning_rate": 9.838274932614556e-05,
+      "loss": 0.7293475270271301,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47378640776699027,
+      "grad_norm": 1.5005141496658325,
+      "learning_rate": 9.856244384546271e-05,
+      "loss": 0.8858493566513062,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 1.3562248945236206,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.7176282405853271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "eval_loss": 0.8152187466621399,
+      "eval_runtime": 835.0394,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.475512405609493,
+      "grad_norm": 1.5226716995239258,
+      "learning_rate": 9.892183288409704e-05,
+      "loss": 0.8048524856567383,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47637540453074434,
+      "grad_norm": 1.3544842004776,
+      "learning_rate": 9.91015274034142e-05,
+      "loss": 0.7954329252243042,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4772384034519957,
+      "grad_norm": 1.5169771909713745,
+      "learning_rate": 9.928122192273136e-05,
+      "loss": 0.804015040397644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47810140237324705,
+      "grad_norm": 1.4011081457138062,
+      "learning_rate": 9.946091644204852e-05,
+      "loss": 0.7652344703674316,
+      "step": 1108
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 1.4163414239883423,
+      "learning_rate": 9.964061096136569e-05,
+      "loss": 0.8501335382461548,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4798274002157497,
+      "grad_norm": 1.7406005859375,
+      "learning_rate": 9.982030548068284e-05,
+      "loss": 0.9024044871330261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.48069039913700107,
+      "grad_norm": 1.1490557193756104,
+      "learning_rate": 0.0001,
+      "loss": 0.7308287024497986,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4815533980582524,
+      "grad_norm": 1.2512775659561157,
+      "learning_rate": 9.99999939713627e-05,
+      "loss": 0.7273974418640137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4824163969795038,
+      "grad_norm": 1.3944242000579834,
+      "learning_rate": 9.999997588545227e-05,
+      "loss": 0.8071790337562561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 1.2209112644195557,
+      "learning_rate": 9.999994574227303e-05,
+      "loss": 0.6586678624153137,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4841423948220065,
+      "grad_norm": 1.7328095436096191,
+      "learning_rate": 9.999990354183228e-05,
+      "loss": 0.8529233336448669,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4850053937432578,
+      "grad_norm": 1.3309327363967896,
+      "learning_rate": 9.999984928414021e-05,
+      "loss": 0.7529556155204773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.48586839266450915,
+      "grad_norm": 1.22034752368927,
+      "learning_rate": 9.999978296920987e-05,
+      "loss": 0.6780388355255127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4867313915857605,
+      "grad_norm": 1.975162386894226,
+      "learning_rate": 9.999970459705727e-05,
+      "loss": 0.7941693067550659,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 1.4939875602722168,
+      "learning_rate": 9.999961416770132e-05,
+      "loss": 0.7895156145095825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4884573894282632,
+      "grad_norm": 1.4237034320831299,
+      "learning_rate": 9.999951168116379e-05,
+      "loss": 0.7662149667739868,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4893203883495146,
+      "grad_norm": 1.24979567527771,
+      "learning_rate": 9.999939713746942e-05,
+      "loss": 0.8079231381416321,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49018338727076594,
+      "grad_norm": 1.2252353429794312,
+      "learning_rate": 9.999927053664585e-05,
+      "loss": 0.7616434693336487,
+      "step": 1136
+    },
+    {
+      "epoch": 0.49104638619201724,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.999913187872357e-05,
+      "loss": 0.8576793074607849,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 1.2108005285263062,
+      "learning_rate": 9.999898116373605e-05,
+      "loss": 0.8060292601585388,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49277238403451995,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 9.99988183917196e-05,
+      "loss": 0.7532875537872314,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4936353829557713,
+      "grad_norm": 1.261606216430664,
+      "learning_rate": 9.99986435627135e-05,
+      "loss": 0.7241997718811035,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49449838187702266,
+      "grad_norm": 1.2678134441375732,
+      "learning_rate": 9.99984566767599e-05,
+      "loss": 0.78656005859375,
+      "step": 1146
+    },
+    {
+      "epoch": 0.495361380798274,
+      "grad_norm": 1.4185810089111328,
+      "learning_rate": 9.999825773390387e-05,
+      "loss": 0.7423107624053955,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 1.3218237161636353,
+      "learning_rate": 9.999804673419338e-05,
+      "loss": 0.8178372383117676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4970873786407767,
+      "grad_norm": 1.9698240756988525,
+      "learning_rate": 9.99978236776793e-05,
+      "loss": 0.8016347885131836,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49795037756202803,
+      "grad_norm": 1.2228590250015259,
+      "learning_rate": 9.999758856441544e-05,
+      "loss": 0.782421886920929,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4988133764832794,
+      "grad_norm": 1.2753938436508179,
+      "learning_rate": 9.999734139445849e-05,
+      "loss": 0.7235615849494934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.49967637540453075,
+      "grad_norm": 1.353155493736267,
+      "learning_rate": 9.999708216786805e-05,
+      "loss": 0.7686187028884888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 1.4242770671844482,
+      "learning_rate": 9.999681088470663e-05,
+      "loss": 0.7842718362808228,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5014023732470334,
+      "grad_norm": 1.501827359199524,
+      "learning_rate": 9.999652754503966e-05,
+      "loss": 0.8324156999588013,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5022653721682848,
+      "grad_norm": 1.5403931140899658,
+      "learning_rate": 9.999623214893545e-05,
+      "loss": 0.7797855138778687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5031283710895361,
+      "grad_norm": 1.4075721502304077,
+      "learning_rate": 9.999592469646525e-05,
+      "loss": 0.7667348384857178,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5039913700107875,
+      "grad_norm": 1.4411870241165161,
+      "learning_rate": 9.99956051877032e-05,
+      "loss": 0.8153020739555359,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.4197871685028076,
+      "learning_rate": 9.999527362272632e-05,
+      "loss": 0.8548861742019653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5057173678532901,
+      "grad_norm": 1.5030468702316284,
+      "learning_rate": 9.999493000161458e-05,
+      "loss": 0.7978097796440125,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5065803667745415,
+      "grad_norm": 1.483534574508667,
+      "learning_rate": 9.999457432445088e-05,
+      "loss": 0.7990087866783142,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5074433656957928,
+      "grad_norm": 1.5424920320510864,
+      "learning_rate": 9.999420659132093e-05,
+      "loss": 0.8072807788848877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5083063646170443,
+      "grad_norm": 1.129503846168518,
+      "learning_rate": 9.999382680231344e-05,
+      "loss": 0.7216097712516785,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 1.3208131790161133,
+      "learning_rate": 9.999343495752001e-05,
+      "loss": 0.8637414574623108,
+      "step": 1180
+    },
+    {
+      "epoch": 0.510032362459547,
+      "grad_norm": 1.5239129066467285,
+      "learning_rate": 9.999303105703508e-05,
+      "loss": 0.8561362028121948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5108953613807983,
+      "grad_norm": 1.3193504810333252,
+      "learning_rate": 9.99926151009561e-05,
+      "loss": 0.7514780759811401,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5117583603020496,
+      "grad_norm": 1.312649130821228,
+      "learning_rate": 9.999218708938336e-05,
+      "loss": 0.8634516596794128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.512621359223301,
+      "grad_norm": 1.2677429914474487,
+      "learning_rate": 9.999174702242006e-05,
+      "loss": 0.7493832111358643,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 1.3901816606521606,
+      "learning_rate": 9.999129490017234e-05,
+      "loss": 0.8981389999389648,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5143473570658037,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 9.999083072274921e-05,
+      "loss": 0.6771019697189331,
+      "step": 1192
+    },
+    {
+      "epoch": 0.515210355987055,
+      "grad_norm": 1.1299076080322266,
+      "learning_rate": 9.999035449026261e-05,
+      "loss": 0.789888858795166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5160733549083064,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 9.998986620282738e-05,
+      "loss": 0.6656325459480286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5169363538295577,
+      "grad_norm": 1.3438260555267334,
+      "learning_rate": 9.998936586056127e-05,
+      "loss": 0.795595645904541,
+      "step": 1198
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 1.45097017288208,
+      "learning_rate": 9.998885346358495e-05,
+      "loss": 0.7769836187362671,
+      "step": 1200
+    },
+    {
+      "epoch": 0.517799352750809,
+      "eval_loss": 0.8070361018180847,
+      "eval_runtime": 823.9296,
+      "eval_samples_per_second": 2.5,
+      "eval_steps_per_second": 2.5,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5186623516720604,
+      "grad_norm": 1.3791370391845703,
+      "learning_rate": 9.998832901202195e-05,
+      "loss": 0.7527318000793457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5195253505933117,
+      "grad_norm": 1.2574659585952759,
+      "learning_rate": 9.998779250599876e-05,
+      "loss": 0.8021673560142517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5203883495145631,
+      "grad_norm": 1.3305809497833252,
+      "learning_rate": 9.998724394564476e-05,
+      "loss": 0.776639461517334,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5212513484358144,
+      "grad_norm": 1.2330507040023804,
+      "learning_rate": 9.998668333109223e-05,
+      "loss": 0.7381181120872498,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 1.0482324361801147,
+      "learning_rate": 9.998611066247635e-05,
+      "loss": 0.6595721244812012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5229773462783172,
+      "grad_norm": 1.5456492900848389,
+      "learning_rate": 9.998552593993522e-05,
+      "loss": 0.8388534784317017,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5238403451995685,
+      "grad_norm": 1.3730167150497437,
+      "learning_rate": 9.998492916360986e-05,
+      "loss": 0.7593658566474915,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5247033441208199,
+      "grad_norm": 1.2668508291244507,
+      "learning_rate": 9.998432033364416e-05,
+      "loss": 0.7796505093574524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5255663430420712,
+      "grad_norm": 1.1175068616867065,
+      "learning_rate": 9.998369945018493e-05,
+      "loss": 0.7493594288825989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 1.121848464012146,
+      "learning_rate": 9.99830665133819e-05,
+      "loss": 0.7791322469711304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5272923408845739,
+      "grad_norm": 1.49788236618042,
+      "learning_rate": 9.998242152338772e-05,
+      "loss": 0.8299469947814941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5281553398058253,
+      "grad_norm": 1.227125644683838,
+      "learning_rate": 9.998176448035791e-05,
+      "loss": 0.7911439538002014,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5290183387270766,
+      "grad_norm": 1.176081895828247,
+      "learning_rate": 9.998109538445091e-05,
+      "loss": 0.6487553119659424,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5298813376483279,
+      "grad_norm": 1.169835090637207,
+      "learning_rate": 9.99804142358281e-05,
+      "loss": 0.7331781983375549,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 1.4085001945495605,
+      "learning_rate": 9.997972103465367e-05,
+      "loss": 0.8198121190071106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5316073354908306,
+      "grad_norm": 1.8634135723114014,
+      "learning_rate": 9.997901578109484e-05,
+      "loss": 0.7099791169166565,
+      "step": 1232
+    },
+    {
+      "epoch": 0.532470334412082,
+      "grad_norm": 1.4719511270523071,
+      "learning_rate": 9.997829847532166e-05,
+      "loss": 0.7416951656341553,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.2430769205093384,
+      "learning_rate": 9.99775691175071e-05,
+      "loss": 0.7583408355712891,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5341963322545846,
+      "grad_norm": 1.2502130270004272,
+      "learning_rate": 9.997682770782706e-05,
+      "loss": 0.7639391422271729,
+      "step": 1238
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 1.3577396869659424,
+      "learning_rate": 9.997607424646032e-05,
+      "loss": 0.8304449319839478,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5359223300970873,
+      "grad_norm": 1.0986218452453613,
+      "learning_rate": 9.997530873358854e-05,
+      "loss": 0.7331952452659607,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5367853290183388,
+      "grad_norm": 1.4607514142990112,
+      "learning_rate": 9.997453116939637e-05,
+      "loss": 0.8173373341560364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.53764832793959,
+      "grad_norm": 1.2897552251815796,
+      "learning_rate": 9.997374155407128e-05,
+      "loss": 0.7677372694015503,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5385113268608415,
+      "grad_norm": 1.4581117630004883,
+      "learning_rate": 9.997293988780372e-05,
+      "loss": 0.8524296283721924,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 1.2029366493225098,
+      "learning_rate": 9.997212617078698e-05,
+      "loss": 0.7920866012573242,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5402373247033441,
+      "grad_norm": 1.109483242034912,
+      "learning_rate": 9.997130040321727e-05,
+      "loss": 0.7889736294746399,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5411003236245955,
+      "grad_norm": 1.1930333375930786,
+      "learning_rate": 9.997046258529377e-05,
+      "loss": 0.6593763828277588,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5419633225458468,
+      "grad_norm": 1.3668162822723389,
+      "learning_rate": 9.996961271721845e-05,
+      "loss": 0.7768212556838989,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5428263214670982,
+      "grad_norm": 1.2901886701583862,
+      "learning_rate": 9.996875079919632e-05,
+      "loss": 0.7392467260360718,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.366720199584961,
+      "learning_rate": 9.996787683143518e-05,
+      "loss": 0.6961479187011719,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5445523193096009,
+      "grad_norm": 1.2925949096679688,
+      "learning_rate": 9.99669908141458e-05,
+      "loss": 0.801517128944397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5454153182308522,
+      "grad_norm": 1.1967244148254395,
+      "learning_rate": 9.996609274754182e-05,
+      "loss": 0.7287030816078186,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5462783171521035,
+      "grad_norm": 1.419996738433838,
+      "learning_rate": 9.996518263183985e-05,
+      "loss": 0.7791255116462708,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5471413160733549,
+      "grad_norm": 1.3544845581054688,
+      "learning_rate": 9.996426046725933e-05,
+      "loss": 0.6665297150611877,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 1.6272345781326294,
+      "learning_rate": 9.996332625402262e-05,
+      "loss": 0.7902728915214539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5488673139158576,
+      "grad_norm": 1.4264246225357056,
+      "learning_rate": 9.996237999235502e-05,
+      "loss": 0.8483795523643494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5497303128371089,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 9.996142168248473e-05,
+      "loss": 0.7432113289833069,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5505933117583603,
+      "grad_norm": 0.9958710074424744,
+      "learning_rate": 9.996045132464283e-05,
+      "loss": 0.6721545457839966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5514563106796116,
+      "grad_norm": 1.371719241142273,
+      "learning_rate": 9.995946891906331e-05,
+      "loss": 0.7770468592643738,
+      "step": 1278
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 1.1906969547271729,
+      "learning_rate": 9.995847446598307e-05,
+      "loss": 0.816495418548584,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5531823085221144,
+      "grad_norm": 1.262184977531433,
+      "learning_rate": 9.995746796564195e-05,
+      "loss": 0.7122345566749573,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5540453074433657,
+      "grad_norm": 1.454692006111145,
+      "learning_rate": 9.995644941828262e-05,
+      "loss": 0.8476479649543762,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5549083063646171,
+      "grad_norm": 1.44277024269104,
+      "learning_rate": 9.995541882415072e-05,
+      "loss": 0.7595248818397522,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5557713052858684,
+      "grad_norm": 1.151465654373169,
+      "learning_rate": 9.995437618349478e-05,
+      "loss": 0.7413201928138733,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 1.3160182237625122,
+      "learning_rate": 9.995332149656621e-05,
+      "loss": 0.7242997884750366,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5574973031283711,
+      "grad_norm": 1.2748576402664185,
+      "learning_rate": 9.995225476361936e-05,
+      "loss": 0.7779543995857239,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5583603020496224,
+      "grad_norm": 1.3639990091323853,
+      "learning_rate": 9.995117598491145e-05,
+      "loss": 0.8075780868530273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5592233009708738,
+      "grad_norm": 1.0247999429702759,
+      "learning_rate": 9.995008516070263e-05,
+      "loss": 0.6940524578094482,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5600862998921251,
+      "grad_norm": 1.3383198976516724,
+      "learning_rate": 9.994898229125597e-05,
+      "loss": 0.799100399017334,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 1.37660551071167,
+      "learning_rate": 9.994786737683738e-05,
+      "loss": 0.8038942217826843,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "eval_loss": 0.7962777018547058,
+      "eval_runtime": 827.8798,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5618122977346278,
+      "grad_norm": 1.5424575805664062,
+      "learning_rate": 9.994674041771574e-05,
+      "loss": 0.8553245067596436,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5626752966558792,
+      "grad_norm": 1.1047813892364502,
+      "learning_rate": 9.994560141416282e-05,
+      "loss": 0.6650274991989136,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5635382955771305,
+      "grad_norm": 1.4532393217086792,
+      "learning_rate": 9.994445036645328e-05,
+      "loss": 0.794471025466919,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5644012944983818,
+      "grad_norm": 1.2588927745819092,
+      "learning_rate": 9.994328727486468e-05,
+      "loss": 0.7241487503051758,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 1.4981781244277954,
+      "learning_rate": 9.994211213967748e-05,
+      "loss": 0.7781044244766235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5661272923408845,
+      "grad_norm": 1.426277995109558,
+      "learning_rate": 9.99409249611751e-05,
+      "loss": 0.7772427201271057,
+      "step": 1312
+    },
+    {
+      "epoch": 0.566990291262136,
+      "grad_norm": 1.1987197399139404,
+      "learning_rate": 9.993972573964379e-05,
+      "loss": 0.7443774342536926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5678532901833873,
+      "grad_norm": 1.1490470170974731,
+      "learning_rate": 9.993851447537275e-05,
+      "loss": 0.7203393578529358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5687162891046387,
+      "grad_norm": 1.3586851358413696,
+      "learning_rate": 9.993729116865406e-05,
+      "loss": 0.8038793206214905,
+      "step": 1318
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 1.4160653352737427,
+      "learning_rate": 9.993605581978274e-05,
+      "loss": 0.827308714389801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5704422869471413,
+      "grad_norm": 1.2725251913070679,
+      "learning_rate": 9.993480842905667e-05,
+      "loss": 0.8183196783065796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5713052858683927,
+      "grad_norm": 1.3789756298065186,
+      "learning_rate": 9.993354899677665e-05,
+      "loss": 0.8268020153045654,
+      "step": 1324
+    },
+    {
+      "epoch": 0.572168284789644,
+      "grad_norm": 1.2641910314559937,
+      "learning_rate": 9.993227752324639e-05,
+      "loss": 0.7434715628623962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5730312837108954,
+      "grad_norm": 1.2429786920547485,
+      "learning_rate": 9.99309940087725e-05,
+      "loss": 0.7636955976486206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 1.3281534910202026,
+      "learning_rate": 9.992969845366449e-05,
+      "loss": 0.794617772102356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.574757281553398,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 9.992839085823477e-05,
+      "loss": 0.7658806443214417,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5756202804746494,
+      "grad_norm": 1.4143259525299072,
+      "learning_rate": 9.99270712227987e-05,
+      "loss": 0.8483052849769592,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5764832793959007,
+      "grad_norm": 1.2387818098068237,
+      "learning_rate": 9.992573954767449e-05,
+      "loss": 0.8610434532165527,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5773462783171521,
+      "grad_norm": 1.3307154178619385,
+      "learning_rate": 9.992439583318322e-05,
+      "loss": 0.8021210432052612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 1.293459415435791,
+      "learning_rate": 9.992304007964896e-05,
+      "loss": 0.7536460757255554,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5790722761596548,
+      "grad_norm": 1.1508948802947998,
+      "learning_rate": 9.992167228739865e-05,
+      "loss": 0.7519120573997498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5799352750809061,
+      "grad_norm": 1.5906150341033936,
+      "learning_rate": 9.992029245676211e-05,
+      "loss": 0.784415066242218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5807982740021574,
+      "grad_norm": 1.1421746015548706,
+      "learning_rate": 9.99189005880721e-05,
+      "loss": 0.8217549324035645,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5816612729234089,
+      "grad_norm": 1.0767277479171753,
+      "learning_rate": 9.991749668166422e-05,
+      "loss": 0.7446820735931396,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 1.147594928741455,
+      "learning_rate": 9.991608073787706e-05,
+      "loss": 0.8426011800765991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5833872707659116,
+      "grad_norm": 1.6618242263793945,
+      "learning_rate": 9.991465275705206e-05,
+      "loss": 0.8185623288154602,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5842502696871629,
+      "grad_norm": 1.414926290512085,
+      "learning_rate": 9.991321273953355e-05,
+      "loss": 0.7521756887435913,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5851132686084143,
+      "grad_norm": 1.298245906829834,
+      "learning_rate": 9.991176068566882e-05,
+      "loss": 0.7517471313476562,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5859762675296656,
+      "grad_norm": 1.3243110179901123,
+      "learning_rate": 9.9910296595808e-05,
+      "loss": 0.749305248260498,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 1.1953799724578857,
+      "learning_rate": 9.990882047030413e-05,
+      "loss": 0.6883649230003357,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5877022653721683,
+      "grad_norm": 1.3934388160705566,
+      "learning_rate": 9.990733230951322e-05,
+      "loss": 0.7859632968902588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5885652642934196,
+      "grad_norm": 1.2212340831756592,
+      "learning_rate": 9.990583211379409e-05,
+      "loss": 0.7313099503517151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.589428263214671,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 9.990431988350853e-05,
+      "loss": 0.8007960319519043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5902912621359223,
+      "grad_norm": 1.1826236248016357,
+      "learning_rate": 9.990279561902118e-05,
+      "loss": 0.773026168346405,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 1.1181561946868896,
+      "learning_rate": 9.990125932069964e-05,
+      "loss": 0.7469213604927063,
+      "step": 1370
+    },
+    {
+      "epoch": 0.592017259978425,
+      "grad_norm": 1.3735432624816895,
+      "learning_rate": 9.989971098891439e-05,
+      "loss": 0.7788950204849243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5928802588996763,
+      "grad_norm": 1.1906198263168335,
+      "learning_rate": 9.989815062403877e-05,
+      "loss": 0.8117855191230774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5937432578209277,
+      "grad_norm": 1.326940894126892,
+      "learning_rate": 9.989657822644907e-05,
+      "loss": 0.7536540627479553,
+      "step": 1376
+    },
+    {
+      "epoch": 0.594606256742179,
+      "grad_norm": 1.4508434534072876,
+      "learning_rate": 9.989499379652447e-05,
+      "loss": 0.860320508480072,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 1.114499568939209,
+      "learning_rate": 9.989339733464704e-05,
+      "loss": 0.749508798122406,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5963322545846818,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 9.989178884120176e-05,
+      "loss": 0.7294815182685852,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5971952535059332,
+      "grad_norm": 1.291913628578186,
+      "learning_rate": 9.989016831657652e-05,
+      "loss": 0.7797688245773315,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5980582524271845,
+      "grad_norm": 1.1245452165603638,
+      "learning_rate": 9.988853576116208e-05,
+      "loss": 0.7824407815933228,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5989212513484358,
+      "grad_norm": 1.2384979724884033,
+      "learning_rate": 9.988689117535215e-05,
+      "loss": 0.714794933795929,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 1.1945903301239014,
+      "learning_rate": 9.988523455954332e-05,
+      "loss": 0.7072597742080688,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6006472491909385,
+      "grad_norm": 1.3414266109466553,
+      "learning_rate": 9.988356591413504e-05,
+      "loss": 0.7075814008712769,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6015102481121899,
+      "grad_norm": 1.4575660228729248,
+      "learning_rate": 9.988188523952973e-05,
+      "loss": 0.8237126469612122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6023732470334412,
+      "grad_norm": 1.3836476802825928,
+      "learning_rate": 9.988019253613265e-05,
+      "loss": 0.7991626858711243,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6032362459546926,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.987848780435201e-05,
+      "loss": 0.6936330199241638,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 1.171294093132019,
+      "learning_rate": 9.98767710445989e-05,
+      "loss": 0.7834921479225159,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "eval_loss": 0.7866972088813782,
+      "eval_runtime": 836.0759,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6049622437971952,
+      "grad_norm": 1.5934456586837769,
+      "learning_rate": 9.987504225728727e-05,
+      "loss": 0.765119194984436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6058252427184466,
+      "grad_norm": 1.231372356414795,
+      "learning_rate": 9.987330144283407e-05,
+      "loss": 0.752061665058136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6066882416396979,
+      "grad_norm": 1.315215826034546,
+      "learning_rate": 9.987154860165903e-05,
+      "loss": 0.7475255727767944,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6075512405609493,
+      "grad_norm": 1.2640653848648071,
+      "learning_rate": 9.986978373418488e-05,
+      "loss": 0.777019202709198,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 1.085493803024292,
+      "learning_rate": 9.98680068408372e-05,
+      "loss": 0.6682190299034119,
+      "step": 1410
+    },
+    {
+      "epoch": 0.609277238403452,
+      "grad_norm": 1.4247251749038696,
+      "learning_rate": 9.986621792204447e-05,
+      "loss": 0.7765674591064453,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6101402373247033,
+      "grad_norm": 1.2993831634521484,
+      "learning_rate": 9.986441697823808e-05,
+      "loss": 0.7429938316345215,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6110032362459547,
+      "grad_norm": 2.614980936050415,
+      "learning_rate": 9.986260400985233e-05,
+      "loss": 0.7838386297225952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6118662351672061,
+      "grad_norm": 1.4305615425109863,
+      "learning_rate": 9.98607790173244e-05,
+      "loss": 0.7679475545883179,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 1.1412850618362427,
+      "learning_rate": 9.98589420010944e-05,
+      "loss": 0.6889669895172119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6135922330097088,
+      "grad_norm": 1.372313141822815,
+      "learning_rate": 9.98570929616053e-05,
+      "loss": 0.7659094333648682,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6144552319309601,
+      "grad_norm": 1.2629581689834595,
+      "learning_rate": 9.985523189930297e-05,
+      "loss": 0.7404800653457642,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6153182308522115,
+      "grad_norm": 1.5652254819869995,
+      "learning_rate": 9.985335881463623e-05,
+      "loss": 0.8082308769226074,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6161812297734628,
+      "grad_norm": 1.2268856763839722,
+      "learning_rate": 9.985147370805675e-05,
+      "loss": 0.7773997187614441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 1.2190957069396973,
+      "learning_rate": 9.984957658001912e-05,
+      "loss": 0.7412725687026978,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6179072276159655,
+      "grad_norm": 1.481990098953247,
+      "learning_rate": 9.984766743098081e-05,
+      "loss": 0.7257611751556396,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6187702265372168,
+      "grad_norm": 1.0297027826309204,
+      "learning_rate": 9.984574626140223e-05,
+      "loss": 0.7151464819908142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6196332254584682,
+      "grad_norm": 1.2404571771621704,
+      "learning_rate": 9.984381307174663e-05,
+      "loss": 0.7855883240699768,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6204962243797195,
+      "grad_norm": 1.2952802181243896,
+      "learning_rate": 9.984186786248023e-05,
+      "loss": 0.6360295414924622,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.5405724048614502,
+      "learning_rate": 9.983991063407206e-05,
+      "loss": 0.7280036807060242,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.303699254989624,
+      "learning_rate": 9.983794138699411e-05,
+      "loss": 0.7553068995475769,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6230852211434735,
+      "grad_norm": 1.1452997922897339,
+      "learning_rate": 9.983596012172127e-05,
+      "loss": 0.7329047918319702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.623948220064725,
+      "grad_norm": 1.2648253440856934,
+      "learning_rate": 9.983396683873133e-05,
+      "loss": 0.753168523311615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6248112189859762,
+      "grad_norm": 1.2095013856887817,
+      "learning_rate": 9.983196153850493e-05,
+      "loss": 0.7378705739974976,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 1.2855647802352905,
+      "learning_rate": 9.982994422152564e-05,
+      "loss": 0.7910448908805847,
+      "step": 1450
+    },
+    {
+      "epoch": 0.626537216828479,
+      "grad_norm": 1.3895999193191528,
+      "learning_rate": 9.982791488827994e-05,
+      "loss": 0.785302996635437,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6274002157497303,
+      "grad_norm": 1.2398446798324585,
+      "learning_rate": 9.982587353925719e-05,
+      "loss": 0.7542029023170471,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6282632146709817,
+      "grad_norm": 1.2903637886047363,
+      "learning_rate": 9.982382017494965e-05,
+      "loss": 0.8081737756729126,
+      "step": 1456
+    },
+    {
+      "epoch": 0.629126213592233,
+      "grad_norm": 1.223611831665039,
+      "learning_rate": 9.98217547958525e-05,
+      "loss": 0.7886892557144165,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 1.3036459684371948,
+      "learning_rate": 9.981967740246377e-05,
+      "loss": 0.844115138053894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6308522114347357,
+      "grad_norm": 1.6421090364456177,
+      "learning_rate": 9.98175879952844e-05,
+      "loss": 0.7784068584442139,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6317152103559871,
+      "grad_norm": 1.244330644607544,
+      "learning_rate": 9.981548657481828e-05,
+      "loss": 0.746979832649231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6325782092772384,
+      "grad_norm": 1.0323492288589478,
+      "learning_rate": 9.981337314157215e-05,
+      "loss": 0.6680153608322144,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6334412081984897,
+      "grad_norm": 1.2238314151763916,
+      "learning_rate": 9.981124769605563e-05,
+      "loss": 0.7103193402290344,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 1.2647801637649536,
+      "learning_rate": 9.98091102387813e-05,
+      "loss": 0.7426751255989075,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6351672060409924,
+      "grad_norm": 1.1923259496688843,
+      "learning_rate": 9.980696077026456e-05,
+      "loss": 0.7664520144462585,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6360302049622438,
+      "grad_norm": 1.0503894090652466,
+      "learning_rate": 9.980479929102377e-05,
+      "loss": 0.7348257899284363,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6368932038834951,
+      "grad_norm": 1.2223190069198608,
+      "learning_rate": 9.980262580158014e-05,
+      "loss": 0.687703013420105,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6377562028047465,
+      "grad_norm": 1.1469025611877441,
+      "learning_rate": 9.980044030245782e-05,
+      "loss": 0.7903051972389221,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 1.2581169605255127,
+      "learning_rate": 9.979824279418381e-05,
+      "loss": 0.7920540571212769,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6394822006472491,
+      "grad_norm": 1.2616536617279053,
+      "learning_rate": 9.979603327728806e-05,
+      "loss": 0.7030006647109985,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6403451995685006,
+      "grad_norm": 1.3970606327056885,
+      "learning_rate": 9.979381175230335e-05,
+      "loss": 0.7489339709281921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6412081984897519,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 9.979157821976542e-05,
+      "loss": 0.6733010411262512,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6420711974110033,
+      "grad_norm": 1.4987449645996094,
+      "learning_rate": 9.978933268021286e-05,
+      "loss": 0.8221960663795471,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 1.2843966484069824,
+      "learning_rate": 9.978707513418719e-05,
+      "loss": 0.8829001188278198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.643797195253506,
+      "grad_norm": 1.175879955291748,
+      "learning_rate": 9.978480558223278e-05,
+      "loss": 0.7605764269828796,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6446601941747573,
+      "grad_norm": 1.1991835832595825,
+      "learning_rate": 9.978252402489692e-05,
+      "loss": 0.7618141174316406,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6455231930960086,
+      "grad_norm": 1.167635202407837,
+      "learning_rate": 9.978023046272983e-05,
+      "loss": 0.7071146368980408,
+      "step": 1496
+    },
+    {
+      "epoch": 0.64638619201726,
+      "grad_norm": 1.263371229171753,
+      "learning_rate": 9.977792489628456e-05,
+      "loss": 0.7559816837310791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 1.2108519077301025,
+      "learning_rate": 9.977560732611713e-05,
+      "loss": 0.7253944873809814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "eval_loss": 0.778505802154541,
+      "eval_runtime": 836.2317,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6481121898597627,
+      "grad_norm": 1.4794813394546509,
+      "learning_rate": 9.977327775278636e-05,
+      "loss": 0.8212620615959167,
+      "step": 1502
+    },
+    {
+      "epoch": 0.648975188781014,
+      "grad_norm": 1.2120516300201416,
+      "learning_rate": 9.977093617685404e-05,
+      "loss": 0.7220476865768433,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6498381877022654,
+      "grad_norm": 1.3129044771194458,
+      "learning_rate": 9.976858259888484e-05,
+      "loss": 0.7791266441345215,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6507011866235167,
+      "grad_norm": 1.146490454673767,
+      "learning_rate": 9.976621701944628e-05,
+      "loss": 0.7052001357078552,
+      "step": 1508
+    },
+    {
+      "epoch": 0.651564185544768,
+      "grad_norm": 1.1903722286224365,
+      "learning_rate": 9.976383943910885e-05,
+      "loss": 0.7135450839996338,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6524271844660194,
+      "grad_norm": 1.4106701612472534,
+      "learning_rate": 9.976144985844589e-05,
+      "loss": 0.8114889860153198,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6532901833872707,
+      "grad_norm": 1.245661973953247,
+      "learning_rate": 9.97590482780336e-05,
+      "loss": 0.7501562237739563,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6541531823085222,
+      "grad_norm": 1.2918543815612793,
+      "learning_rate": 9.975663469845114e-05,
+      "loss": 0.7697334289550781,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6550161812297735,
+      "grad_norm": 1.2699558734893799,
+      "learning_rate": 9.975420912028053e-05,
+      "loss": 0.7457168102264404,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6558791801510249,
+      "grad_norm": 1.3968263864517212,
+      "learning_rate": 9.975177154410668e-05,
+      "loss": 0.8368204236030579,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6567421790722762,
+      "grad_norm": 1.2161848545074463,
+      "learning_rate": 9.974932197051741e-05,
+      "loss": 0.7355788350105286,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6576051779935275,
+      "grad_norm": 1.129886507987976,
+      "learning_rate": 9.974686040010341e-05,
+      "loss": 0.8023449182510376,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6584681769147789,
+      "grad_norm": 1.1114557981491089,
+      "learning_rate": 9.974438683345829e-05,
+      "loss": 0.7965474724769592,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6593311758360302,
+      "grad_norm": 1.5741950273513794,
+      "learning_rate": 9.974190127117853e-05,
+      "loss": 0.7176838517189026,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 1.1439597606658936,
+      "learning_rate": 9.973940371386352e-05,
+      "loss": 0.7114462852478027,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6610571736785329,
+      "grad_norm": 1.273659348487854,
+      "learning_rate": 9.973689416211553e-05,
+      "loss": 0.754712700843811,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6619201725997842,
+      "grad_norm": 3.9874372482299805,
+      "learning_rate": 9.973437261653973e-05,
+      "loss": 0.6720641255378723,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6627831715210356,
+      "grad_norm": 1.2506502866744995,
+      "learning_rate": 9.973183907774417e-05,
+      "loss": 0.793043851852417,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6636461704422869,
+      "grad_norm": 1.43257737159729,
+      "learning_rate": 9.972929354633982e-05,
+      "loss": 0.7806380987167358,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6645091693635383,
+      "grad_norm": 1.4794938564300537,
+      "learning_rate": 9.97267360229405e-05,
+      "loss": 0.7176088094711304,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6653721682847896,
+      "grad_norm": 1.2398169040679932,
+      "learning_rate": 9.972416650816297e-05,
+      "loss": 0.7145742774009705,
+      "step": 1542
+    },
+    {
+      "epoch": 0.666235167206041,
+      "grad_norm": 1.1309953927993774,
+      "learning_rate": 9.972158500262683e-05,
+      "loss": 0.8025617003440857,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6670981661272923,
+      "grad_norm": 1.2479451894760132,
+      "learning_rate": 9.971899150695463e-05,
+      "loss": 0.7808589339256287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6679611650485436,
+      "grad_norm": 1.2538350820541382,
+      "learning_rate": 9.971638602177175e-05,
+      "loss": 0.7096936106681824,
+      "step": 1548
+    },
+    {
+      "epoch": 0.668824163969795,
+      "grad_norm": 1.1431525945663452,
+      "learning_rate": 9.97137685477065e-05,
+      "loss": 0.6529367566108704,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6696871628910464,
+      "grad_norm": 1.5966559648513794,
+      "learning_rate": 9.971113908539008e-05,
+      "loss": 0.7780346870422363,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6705501618122978,
+      "grad_norm": 1.186475157737732,
+      "learning_rate": 9.970849763545658e-05,
+      "loss": 0.6889166831970215,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6714131607335491,
+      "grad_norm": 1.1922476291656494,
+      "learning_rate": 9.970584419854295e-05,
+      "loss": 0.7154867649078369,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6722761596548005,
+      "grad_norm": 1.30132257938385,
+      "learning_rate": 9.970317877528906e-05,
+      "loss": 0.7419366240501404,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6731391585760518,
+      "grad_norm": 1.1539888381958008,
+      "learning_rate": 9.970050136633769e-05,
+      "loss": 0.7856602072715759,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6740021574973031,
+      "grad_norm": 1.3056161403656006,
+      "learning_rate": 9.969781197233445e-05,
+      "loss": 0.6834115386009216,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6748651564185545,
+      "grad_norm": 1.2899249792099,
+      "learning_rate": 9.96951105939279e-05,
+      "loss": 0.8490727543830872,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6757281553398058,
+      "grad_norm": 1.2182374000549316,
+      "learning_rate": 9.969239723176946e-05,
+      "loss": 0.7462526559829712,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6765911542610572,
+      "grad_norm": 1.1231273412704468,
+      "learning_rate": 9.968967188651344e-05,
+      "loss": 0.7145209908485413,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6774541531823085,
+      "grad_norm": 1.1776760816574097,
+      "learning_rate": 9.968693455881703e-05,
+      "loss": 0.7911707162857056,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6783171521035599,
+      "grad_norm": 1.3542319536209106,
+      "learning_rate": 9.968418524934035e-05,
+      "loss": 0.6946635246276855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6791801510248112,
+      "grad_norm": 1.3073060512542725,
+      "learning_rate": 9.96814239587464e-05,
+      "loss": 0.7597488164901733,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6800431499460625,
+      "grad_norm": 1.3205567598342896,
+      "learning_rate": 9.967865068770099e-05,
+      "loss": 0.6763246655464172,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6809061488673139,
+      "grad_norm": 1.1734344959259033,
+      "learning_rate": 9.967586543687295e-05,
+      "loss": 0.7560655474662781,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6817691477885652,
+      "grad_norm": 1.445770025253296,
+      "learning_rate": 9.967306820693388e-05,
+      "loss": 0.8050205707550049,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6826321467098166,
+      "grad_norm": 1.1185686588287354,
+      "learning_rate": 9.967025899855836e-05,
+      "loss": 0.6628211140632629,
+      "step": 1582
+    },
+    {
+      "epoch": 0.683495145631068,
+      "grad_norm": 1.3831934928894043,
+      "learning_rate": 9.966743781242379e-05,
+      "loss": 0.7684114575386047,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6843581445523194,
+      "grad_norm": 1.2417019605636597,
+      "learning_rate": 9.966460464921047e-05,
+      "loss": 0.7084140777587891,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6852211434735707,
+      "grad_norm": 1.2138086557388306,
+      "learning_rate": 9.966175950960163e-05,
+      "loss": 0.6817399263381958,
+      "step": 1588
+    },
+    {
+      "epoch": 0.686084142394822,
+      "grad_norm": 1.399778962135315,
+      "learning_rate": 9.965890239428339e-05,
+      "loss": 0.8422354459762573,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6869471413160734,
+      "grad_norm": 1.4348714351654053,
+      "learning_rate": 9.965603330394468e-05,
+      "loss": 0.7432681322097778,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6878101402373247,
+      "grad_norm": 1.192425012588501,
+      "learning_rate": 9.965315223927741e-05,
+      "loss": 0.7571184635162354,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6886731391585761,
+      "grad_norm": 1.2834038734436035,
+      "learning_rate": 9.96502592009763e-05,
+      "loss": 0.802443265914917,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6895361380798274,
+      "grad_norm": 1.1650444269180298,
+      "learning_rate": 9.964735418973901e-05,
+      "loss": 0.8113405704498291,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "grad_norm": 1.1714438199996948,
+      "learning_rate": 9.964443720626606e-05,
+      "loss": 0.7507118582725525,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "eval_loss": 0.7706249952316284,
+      "eval_runtime": 821.4894,
+      "eval_samples_per_second": 2.508,
+      "eval_steps_per_second": 2.508,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6912621359223301,
+      "grad_norm": 1.175459861755371,
+      "learning_rate": 9.964150825126088e-05,
+      "loss": 0.7422796487808228,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6921251348435814,
+      "grad_norm": 1.6418036222457886,
+      "learning_rate": 9.963856732542978e-05,
+      "loss": 0.7125853300094604,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6929881337648328,
+      "grad_norm": 1.3256275653839111,
+      "learning_rate": 9.963561442948193e-05,
+      "loss": 0.757859468460083,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6938511326860841,
+      "grad_norm": 1.19981050491333,
+      "learning_rate": 9.963264956412944e-05,
+      "loss": 0.6849292516708374,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6947141316073355,
+      "grad_norm": 1.0392169952392578,
+      "learning_rate": 9.962967273008723e-05,
+      "loss": 0.7636215090751648,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6955771305285868,
+      "grad_norm": 1.1106646060943604,
+      "learning_rate": 9.962668392807318e-05,
+      "loss": 0.7208617329597473,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6964401294498382,
+      "grad_norm": 1.1569621562957764,
+      "learning_rate": 9.962368315880801e-05,
+      "loss": 0.671805739402771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6973031283710895,
+      "grad_norm": 1.2157769203186035,
+      "learning_rate": 9.962067042301537e-05,
+      "loss": 0.7648762464523315,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6981661272923408,
+      "grad_norm": 1.4804017543792725,
+      "learning_rate": 9.961764572142173e-05,
+      "loss": 0.8490509390830994,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 1.0344431400299072,
+      "learning_rate": 9.961460905475651e-05,
+      "loss": 0.5950891971588135,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6998921251348436,
+      "grad_norm": 1.1109060049057007,
+      "learning_rate": 9.9611560423752e-05,
+      "loss": 0.7288329005241394,
+      "step": 1622
+    },
+    {
+      "epoch": 0.700755124056095,
+      "grad_norm": 1.096996545791626,
+      "learning_rate": 9.960849982914331e-05,
+      "loss": 0.7074542045593262,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7016181229773463,
+      "grad_norm": 1.3109341859817505,
+      "learning_rate": 9.960542727166853e-05,
+      "loss": 0.7348865270614624,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7024811218985976,
+      "grad_norm": 1.3402032852172852,
+      "learning_rate": 9.96023427520686e-05,
+      "loss": 0.7658320665359497,
+      "step": 1628
+    },
+    {
+      "epoch": 0.703344120819849,
+      "grad_norm": 1.3712561130523682,
+      "learning_rate": 9.959924627108732e-05,
+      "loss": 0.809951901435852,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7042071197411003,
+      "grad_norm": 1.2278292179107666,
+      "learning_rate": 9.959613782947139e-05,
+      "loss": 0.6834598779678345,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7050701186623517,
+      "grad_norm": 1.0510832071304321,
+      "learning_rate": 9.95930174279704e-05,
+      "loss": 0.7416030764579773,
+      "step": 1634
+    },
+    {
+      "epoch": 0.705933117583603,
+      "grad_norm": 1.130991816520691,
+      "learning_rate": 9.958988506733684e-05,
+      "loss": 0.7304092645645142,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7067961165048544,
+      "grad_norm": 1.3328672647476196,
+      "learning_rate": 9.958674074832603e-05,
+      "loss": 0.8230003118515015,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7076591154261057,
+      "grad_norm": 1.3842110633850098,
+      "learning_rate": 9.958358447169625e-05,
+      "loss": 0.8413107991218567,
+      "step": 1640
+    },
+    {
+      "epoch": 0.708522114347357,
+      "grad_norm": 1.3143812417984009,
+      "learning_rate": 9.958041623820858e-05,
+      "loss": 0.77432781457901,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7093851132686084,
+      "grad_norm": 1.4727009534835815,
+      "learning_rate": 9.957723604862704e-05,
+      "loss": 0.7563844323158264,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7102481121898597,
+      "grad_norm": 1.1892327070236206,
+      "learning_rate": 9.957404390371854e-05,
+      "loss": 0.7085853219032288,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.124180555343628,
+      "learning_rate": 9.957083980425282e-05,
+      "loss": 0.6837730407714844,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 1.2224774360656738,
+      "learning_rate": 9.956762375100255e-05,
+      "loss": 0.7255740165710449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7128371089536139,
+      "grad_norm": 1.360425591468811,
+      "learning_rate": 9.956439574474326e-05,
+      "loss": 0.7849710583686829,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7137001078748652,
+      "grad_norm": 1.2818151712417603,
+      "learning_rate": 9.956115578625338e-05,
+      "loss": 0.7669665813446045,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7145631067961165,
+      "grad_norm": 1.2415164709091187,
+      "learning_rate": 9.95579038763142e-05,
+      "loss": 0.7052890658378601,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7154261057173679,
+      "grad_norm": 1.3983269929885864,
+      "learning_rate": 9.955464001570992e-05,
+      "loss": 0.8300023674964905,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7162891046386192,
+      "grad_norm": 1.1591373682022095,
+      "learning_rate": 9.955136420522759e-05,
+      "loss": 0.7742894887924194,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7171521035598706,
+      "grad_norm": 1.2503128051757812,
+      "learning_rate": 9.954807644565715e-05,
+      "loss": 0.8458141684532166,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7180151024811219,
+      "grad_norm": 1.2954912185668945,
+      "learning_rate": 9.954477673779146e-05,
+      "loss": 0.7044742107391357,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7188781014023733,
+      "grad_norm": 1.1669788360595703,
+      "learning_rate": 9.954146508242619e-05,
+      "loss": 0.7917183041572571,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7197411003236246,
+      "grad_norm": 1.1466871500015259,
+      "learning_rate": 9.953814148035996e-05,
+      "loss": 0.7174562811851501,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7206040992448759,
+      "grad_norm": 1.367224097251892,
+      "learning_rate": 9.953480593239421e-05,
+      "loss": 0.8925880193710327,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7214670981661273,
+      "grad_norm": 1.049497365951538,
+      "learning_rate": 9.953145843933334e-05,
+      "loss": 0.70949387550354,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7223300970873786,
+      "grad_norm": 0.9995084404945374,
+      "learning_rate": 9.952809900198455e-05,
+      "loss": 0.6731935143470764,
+      "step": 1674
+    },
+    {
+      "epoch": 0.72319309600863,
+      "grad_norm": 1.100899577140808,
+      "learning_rate": 9.952472762115794e-05,
+      "loss": 0.7316275835037231,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7240560949298813,
+      "grad_norm": 1.2594135999679565,
+      "learning_rate": 9.952134429766655e-05,
+      "loss": 0.7379520535469055,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7249190938511327,
+      "grad_norm": 1.3124911785125732,
+      "learning_rate": 9.951794903232621e-05,
+      "loss": 0.7072247266769409,
+      "step": 1680
+    },
+    {
+      "epoch": 0.725782092772384,
+      "grad_norm": 1.212144374847412,
+      "learning_rate": 9.95145418259557e-05,
+      "loss": 0.745680034160614,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7266450916936353,
+      "grad_norm": 1.3913586139678955,
+      "learning_rate": 9.951112267937663e-05,
+      "loss": 0.796852707862854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7275080906148867,
+      "grad_norm": 1.3227289915084839,
+      "learning_rate": 9.950769159341353e-05,
+      "loss": 0.700892984867096,
+      "step": 1686
+    },
+    {
+      "epoch": 0.728371089536138,
+      "grad_norm": 1.4656381607055664,
+      "learning_rate": 9.950424856889378e-05,
+      "loss": 0.7973095178604126,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7292340884573895,
+      "grad_norm": 1.1698973178863525,
+      "learning_rate": 9.950079360664766e-05,
+      "loss": 0.7811378836631775,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7300970873786408,
+      "grad_norm": 1.201695203781128,
+      "learning_rate": 9.949732670750831e-05,
+      "loss": 0.753012478351593,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7309600862998922,
+      "grad_norm": 1.3465325832366943,
+      "learning_rate": 9.949384787231175e-05,
+      "loss": 0.7620642781257629,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7318230852211435,
+      "grad_norm": 1.2191137075424194,
+      "learning_rate": 9.94903571018969e-05,
+      "loss": 0.7479873299598694,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7326860841423948,
+      "grad_norm": 1.1412187814712524,
+      "learning_rate": 9.948685439710554e-05,
+      "loss": 0.7327720522880554,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "grad_norm": 1.310327410697937,
+      "learning_rate": 9.948333975878233e-05,
+      "loss": 0.761475682258606,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "eval_loss": 0.7649155259132385,
+      "eval_runtime": 828.3542,
+      "eval_samples_per_second": 2.487,
+      "eval_steps_per_second": 2.487,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7344120819848975,
+      "grad_norm": 1.3031405210494995,
+      "learning_rate": 9.947981318777482e-05,
+      "loss": 0.7564467191696167,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7352750809061489,
+      "grad_norm": 1.2672377824783325,
+      "learning_rate": 9.94762746849334e-05,
+      "loss": 0.7689576745033264,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7361380798274002,
+      "grad_norm": 1.1159595251083374,
+      "learning_rate": 9.947272425111138e-05,
+      "loss": 0.7923810482025146,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7370010787486516,
+      "grad_norm": 1.1190921068191528,
+      "learning_rate": 9.946916188716495e-05,
+      "loss": 0.6540790796279907,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 1.373832106590271,
+      "learning_rate": 9.946558759395312e-05,
+      "loss": 0.7267473936080933,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7387270765911542,
+      "grad_norm": 1.2346374988555908,
+      "learning_rate": 9.946200137233785e-05,
+      "loss": 0.7693060636520386,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7395900755124056,
+      "grad_norm": 1.1629668474197388,
+      "learning_rate": 9.945840322318391e-05,
+      "loss": 0.7751407027244568,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7404530744336569,
+      "grad_norm": 1.2268708944320679,
+      "learning_rate": 9.945479314735899e-05,
+      "loss": 0.7798545360565186,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7413160733549083,
+      "grad_norm": 0.9968738555908203,
+      "learning_rate": 9.945117114573366e-05,
+      "loss": 0.6128867268562317,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7421790722761596,
+      "grad_norm": 1.2135992050170898,
+      "learning_rate": 9.944753721918133e-05,
+      "loss": 0.7536954283714294,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7430420711974111,
+      "grad_norm": 1.1361218690872192,
+      "learning_rate": 9.94438913685783e-05,
+      "loss": 0.7034479379653931,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7439050701186624,
+      "grad_norm": 1.0728967189788818,
+      "learning_rate": 9.944023359480378e-05,
+      "loss": 0.7285382747650146,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7447680690399137,
+      "grad_norm": 1.3613945245742798,
+      "learning_rate": 9.94365638987398e-05,
+      "loss": 0.7865195274353027,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7456310679611651,
+      "grad_norm": 1.0985898971557617,
+      "learning_rate": 9.94328822812713e-05,
+      "loss": 0.7511224746704102,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7464940668824164,
+      "grad_norm": 1.1506208181381226,
+      "learning_rate": 9.942918874328606e-05,
+      "loss": 0.732453465461731,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7473570658036678,
+      "grad_norm": 1.0995100736618042,
+      "learning_rate": 9.94254832856748e-05,
+      "loss": 0.702409029006958,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7482200647249191,
+      "grad_norm": 1.0075604915618896,
+      "learning_rate": 9.942176590933106e-05,
+      "loss": 0.7515703439712524,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7490830636461704,
+      "grad_norm": 1.3360812664031982,
+      "learning_rate": 9.941803661515126e-05,
+      "loss": 0.6910237073898315,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7499460625674218,
+      "grad_norm": 1.0172319412231445,
+      "learning_rate": 9.941429540403471e-05,
+      "loss": 0.7185606956481934,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7508090614886731,
+      "grad_norm": 1.2721658945083618,
+      "learning_rate": 9.941054227688358e-05,
+      "loss": 0.7136536836624146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7516720604099245,
+      "grad_norm": 1.0775413513183594,
+      "learning_rate": 9.940677723460292e-05,
+      "loss": 0.7595619559288025,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7525350593311758,
+      "grad_norm": 1.3436479568481445,
+      "learning_rate": 9.940300027810067e-05,
+      "loss": 0.7602618336677551,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7533980582524272,
+      "grad_norm": 1.2611958980560303,
+      "learning_rate": 9.93992114082876e-05,
+      "loss": 0.7954604029655457,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7542610571736785,
+      "grad_norm": 1.2794841527938843,
+      "learning_rate": 9.93954106260774e-05,
+      "loss": 0.7661190629005432,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7551240560949298,
+      "grad_norm": 1.3042595386505127,
+      "learning_rate": 9.939159793238661e-05,
+      "loss": 0.6855312585830688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7559870550161812,
+      "grad_norm": 1.2799923419952393,
+      "learning_rate": 9.938777332813463e-05,
+      "loss": 0.7970905900001526,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7568500539374325,
+      "grad_norm": 1.169205904006958,
+      "learning_rate": 9.938393681424375e-05,
+      "loss": 0.7072843313217163,
+      "step": 1754
+    },
+    {
+      "epoch": 0.757713052858684,
+      "grad_norm": 1.3833354711532593,
+      "learning_rate": 9.938008839163913e-05,
+      "loss": 0.7134629487991333,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7585760517799353,
+      "grad_norm": 1.223452091217041,
+      "learning_rate": 9.93762280612488e-05,
+      "loss": 0.7873107194900513,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7594390507011867,
+      "grad_norm": 1.2740365266799927,
+      "learning_rate": 9.937235582400367e-05,
+      "loss": 0.7680654525756836,
+      "step": 1760
+    },
+    {
+      "epoch": 0.760302049622438,
+      "grad_norm": 1.2564697265625,
+      "learning_rate": 9.93684716808375e-05,
+      "loss": 0.734126627445221,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7611650485436893,
+      "grad_norm": 1.3858847618103027,
+      "learning_rate": 9.936457563268694e-05,
+      "loss": 0.7972320914268494,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7620280474649407,
+      "grad_norm": 1.2705003023147583,
+      "learning_rate": 9.93606676804915e-05,
+      "loss": 0.7904276847839355,
+      "step": 1766
+    },
+    {
+      "epoch": 0.762891046386192,
+      "grad_norm": 1.210766077041626,
+      "learning_rate": 9.935674782519358e-05,
+      "loss": 0.7542425990104675,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7637540453074434,
+      "grad_norm": 1.4367541074752808,
+      "learning_rate": 9.93528160677384e-05,
+      "loss": 0.7975004315376282,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7646170442286947,
+      "grad_norm": 1.0994079113006592,
+      "learning_rate": 9.934887240907414e-05,
+      "loss": 0.7537807822227478,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7654800431499461,
+      "grad_norm": 1.111660122871399,
+      "learning_rate": 9.934491685015174e-05,
+      "loss": 0.7333855628967285,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7663430420711974,
+      "grad_norm": 1.2879043817520142,
+      "learning_rate": 9.934094939192508e-05,
+      "loss": 0.7472840547561646,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7672060409924487,
+      "grad_norm": 1.3618478775024414,
+      "learning_rate": 9.933697003535094e-05,
+      "loss": 0.7644904255867004,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7680690399137001,
+      "grad_norm": 1.1235666275024414,
+      "learning_rate": 9.933297878138887e-05,
+      "loss": 0.7872052192687988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7689320388349514,
+      "grad_norm": 1.3741775751113892,
+      "learning_rate": 9.932897563100136e-05,
+      "loss": 0.7343124747276306,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7697950377562028,
+      "grad_norm": 1.2115039825439453,
+      "learning_rate": 9.932496058515375e-05,
+      "loss": 0.7486474514007568,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7706580366774541,
+      "grad_norm": 1.183457851409912,
+      "learning_rate": 9.932093364481426e-05,
+      "loss": 0.7333025932312012,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7715210355987056,
+      "grad_norm": 1.2841869592666626,
+      "learning_rate": 9.931689481095394e-05,
+      "loss": 0.7241485118865967,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7723840345199569,
+      "grad_norm": 1.2312977313995361,
+      "learning_rate": 9.931284408454678e-05,
+      "loss": 0.7510396242141724,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7732470334412082,
+      "grad_norm": 1.144895315170288,
+      "learning_rate": 9.930878146656956e-05,
+      "loss": 0.7108346819877625,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7741100323624596,
+      "grad_norm": 1.18603515625,
+      "learning_rate": 9.9304706958002e-05,
+      "loss": 0.7687607407569885,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7749730312837109,
+      "grad_norm": 1.2827935218811035,
+      "learning_rate": 9.93006205598266e-05,
+      "loss": 0.7826682329177856,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7758360302049623,
+      "grad_norm": 1.164852261543274,
+      "learning_rate": 9.929652227302881e-05,
+      "loss": 0.6970428824424744,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 1.0439956188201904,
+      "learning_rate": 9.92924120985969e-05,
+      "loss": 0.7238826751708984,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "eval_loss": 0.7598537802696228,
+      "eval_runtime": 838.9645,
+      "eval_samples_per_second": 2.455,
+      "eval_steps_per_second": 2.455,
+      "step": 1800
+    },
+    {
+      "epoch": 0.777562028047465,
+      "grad_norm": 1.2464938163757324,
+      "learning_rate": 9.928829003752202e-05,
+      "loss": 0.7723641395568848,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7784250269687163,
+      "grad_norm": 1.2445050477981567,
+      "learning_rate": 9.92841560907982e-05,
+      "loss": 0.7154937982559204,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7792880258899676,
+      "grad_norm": 1.2235661745071411,
+      "learning_rate": 9.928001025942231e-05,
+      "loss": 0.8325800895690918,
+      "step": 1806
+    },
+    {
+      "epoch": 0.780151024811219,
+      "grad_norm": 1.268066644668579,
+      "learning_rate": 9.927585254439409e-05,
+      "loss": 0.7093122601509094,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7810140237324703,
+      "grad_norm": 1.1731390953063965,
+      "learning_rate": 9.927168294671618e-05,
+      "loss": 0.7601775527000427,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7818770226537217,
+      "grad_norm": 1.0685608386993408,
+      "learning_rate": 9.926750146739403e-05,
+      "loss": 0.6889236569404602,
+      "step": 1812
+    },
+    {
+      "epoch": 0.782740021574973,
+      "grad_norm": 1.1098639965057373,
+      "learning_rate": 9.926330810743602e-05,
+      "loss": 0.7397196888923645,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7836030204962244,
+      "grad_norm": 1.3606436252593994,
+      "learning_rate": 9.925910286785332e-05,
+      "loss": 0.7180029153823853,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7844660194174757,
+      "grad_norm": 1.143418550491333,
+      "learning_rate": 9.925488574966004e-05,
+      "loss": 0.7039036750793457,
+      "step": 1818
+    },
+    {
+      "epoch": 0.785329018338727,
+      "grad_norm": 1.180968999862671,
+      "learning_rate": 9.92506567538731e-05,
+      "loss": 0.741872251033783,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7861920172599784,
+      "grad_norm": 1.1953548192977905,
+      "learning_rate": 9.924641588151231e-05,
+      "loss": 0.7580198049545288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7870550161812297,
+      "grad_norm": 1.3049067258834839,
+      "learning_rate": 9.924216313360033e-05,
+      "loss": 0.7479216456413269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7879180151024812,
+      "grad_norm": 1.3962669372558594,
+      "learning_rate": 9.92378985111627e-05,
+      "loss": 0.7382357716560364,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7887810140237325,
+      "grad_norm": 1.1519217491149902,
+      "learning_rate": 9.92336220152278e-05,
+      "loss": 0.6356467604637146,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7896440129449838,
+      "grad_norm": 1.2544089555740356,
+      "learning_rate": 9.922933364682689e-05,
+      "loss": 0.7674193978309631,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7905070118662352,
+      "grad_norm": 1.1358295679092407,
+      "learning_rate": 9.922503340699412e-05,
+      "loss": 0.7083898782730103,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7913700107874865,
+      "grad_norm": 1.1744481325149536,
+      "learning_rate": 9.922072129676644e-05,
+      "loss": 0.6273066401481628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7922330097087379,
+      "grad_norm": 1.20509672164917,
+      "learning_rate": 9.92163973171837e-05,
+      "loss": 0.808181643486023,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7930960086299892,
+      "grad_norm": 1.2035597562789917,
+      "learning_rate": 9.921206146928863e-05,
+      "loss": 0.6631553769111633,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7939590075512406,
+      "grad_norm": 1.058695912361145,
+      "learning_rate": 9.920771375412678e-05,
+      "loss": 0.72016841173172,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7948220064724919,
+      "grad_norm": 1.035791039466858,
+      "learning_rate": 9.92033541727466e-05,
+      "loss": 0.7240060567855835,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7956850053937432,
+      "grad_norm": 1.1845753192901611,
+      "learning_rate": 9.919898272619935e-05,
+      "loss": 0.743224561214447,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7965480043149946,
+      "grad_norm": 1.2942636013031006,
+      "learning_rate": 9.919459941553922e-05,
+      "loss": 0.7064419388771057,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7974110032362459,
+      "grad_norm": 1.2736715078353882,
+      "learning_rate": 9.91902042418232e-05,
+      "loss": 0.805932343006134,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7982740021574973,
+      "grad_norm": 1.2803614139556885,
+      "learning_rate": 9.918579720611119e-05,
+      "loss": 0.8050802946090698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7991370010787486,
+      "grad_norm": 1.2225072383880615,
+      "learning_rate": 9.91813783094659e-05,
+      "loss": 0.7013180255889893,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3375571966171265,
+      "learning_rate": 9.917694755295294e-05,
+      "loss": 0.7415283918380737,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8008629989212513,
+      "grad_norm": 1.2209177017211914,
+      "learning_rate": 9.917250493764079e-05,
+      "loss": 0.7264692783355713,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8017259978425026,
+      "grad_norm": 1.1158864498138428,
+      "learning_rate": 9.916805046460074e-05,
+      "loss": 0.7083719372749329,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8025889967637541,
+      "grad_norm": 1.171000361442566,
+      "learning_rate": 9.916358413490696e-05,
+      "loss": 0.820396900177002,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8034519956850054,
+      "grad_norm": 0.9942941069602966,
+      "learning_rate": 9.91591059496365e-05,
+      "loss": 0.6754014492034912,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8043149946062568,
+      "grad_norm": 1.239795207977295,
+      "learning_rate": 9.915461590986926e-05,
+      "loss": 0.813154399394989,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8051779935275081,
+      "grad_norm": 1.076019525527954,
+      "learning_rate": 9.915011401668796e-05,
+      "loss": 0.745580792427063,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8060409924487595,
+      "grad_norm": 1.2134130001068115,
+      "learning_rate": 9.914560027117826e-05,
+      "loss": 0.7564949989318848,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8069039913700108,
+      "grad_norm": 1.1774994134902954,
+      "learning_rate": 9.914107467442859e-05,
+      "loss": 0.665183424949646,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8077669902912621,
+      "grad_norm": 1.0606465339660645,
+      "learning_rate": 9.91365372275303e-05,
+      "loss": 0.7597744464874268,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8086299892125135,
+      "grad_norm": 1.397161841392517,
+      "learning_rate": 9.913198793157757e-05,
+      "loss": 0.7453591227531433,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8094929881337648,
+      "grad_norm": 1.250003695487976,
+      "learning_rate": 9.912742678766742e-05,
+      "loss": 0.715792179107666,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8103559870550162,
+      "grad_norm": 1.2343473434448242,
+      "learning_rate": 9.912285379689979e-05,
+      "loss": 0.7094268798828125,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8112189859762675,
+      "grad_norm": 1.453764796257019,
+      "learning_rate": 9.911826896037739e-05,
+      "loss": 0.6652816534042358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8120819848975189,
+      "grad_norm": 1.030903935432434,
+      "learning_rate": 9.911367227920587e-05,
+      "loss": 0.6764529347419739,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8129449838187702,
+      "grad_norm": 1.2725969552993774,
+      "learning_rate": 9.91090637544937e-05,
+      "loss": 0.7043389678001404,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8138079827400215,
+      "grad_norm": 1.1610769033432007,
+      "learning_rate": 9.910444338735216e-05,
+      "loss": 0.6960558295249939,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8146709816612729,
+      "grad_norm": 1.2049180269241333,
+      "learning_rate": 9.909981117889548e-05,
+      "loss": 0.8029319643974304,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 1.1671873331069946,
+      "learning_rate": 9.909516713024068e-05,
+      "loss": 0.683114767074585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8163969795037757,
+      "grad_norm": 1.247768759727478,
+      "learning_rate": 9.909051124250764e-05,
+      "loss": 0.7193703055381775,
+      "step": 1892
+    },
+    {
+      "epoch": 0.817259978425027,
+      "grad_norm": 1.1376357078552246,
+      "learning_rate": 9.908584351681911e-05,
+      "loss": 0.7957084774971008,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8181229773462784,
+      "grad_norm": 1.1918323040008545,
+      "learning_rate": 9.90811639543007e-05,
+      "loss": 0.7168112993240356,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8189859762675297,
+      "grad_norm": 1.1249650716781616,
+      "learning_rate": 9.907647255608087e-05,
+      "loss": 0.7519245147705078,
+      "step": 1898
+    },
+    {
+      "epoch": 0.819848975188781,
+      "grad_norm": 1.0688576698303223,
+      "learning_rate": 9.907176932329092e-05,
+      "loss": 0.7496176362037659,
+      "step": 1900
+    },
+    {
+      "epoch": 0.819848975188781,
+      "eval_loss": 0.7533769607543945,
+      "eval_runtime": 830.2894,
+      "eval_samples_per_second": 2.481,
+      "eval_steps_per_second": 2.481,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8207119741100324,
+      "grad_norm": 0.9264270663261414,
+      "learning_rate": 9.906705425706499e-05,
+      "loss": 0.639653205871582,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8215749730312837,
+      "grad_norm": 1.2231628894805908,
+      "learning_rate": 9.906232735854016e-05,
+      "loss": 0.7555551528930664,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8224379719525351,
+      "grad_norm": 1.3369156122207642,
+      "learning_rate": 9.905758862885624e-05,
+      "loss": 0.7113592028617859,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8233009708737864,
+      "grad_norm": 1.217238187789917,
+      "learning_rate": 9.905283806915597e-05,
+      "loss": 0.6823481321334839,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8241639697950378,
+      "grad_norm": 1.2462831735610962,
+      "learning_rate": 9.904807568058495e-05,
+      "loss": 0.7347100377082825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8250269687162891,
+      "grad_norm": 1.2507221698760986,
+      "learning_rate": 9.904330146429157e-05,
+      "loss": 0.7981513738632202,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8258899676375404,
+      "grad_norm": 1.2246814966201782,
+      "learning_rate": 9.903851542142715e-05,
+      "loss": 0.759779155254364,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8267529665587918,
+      "grad_norm": 1.1746087074279785,
+      "learning_rate": 9.903371755314579e-05,
+      "loss": 0.741083562374115,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8276159654800431,
+      "grad_norm": 1.1913539171218872,
+      "learning_rate": 9.902890786060448e-05,
+      "loss": 0.7491101622581482,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8284789644012945,
+      "grad_norm": 1.2342501878738403,
+      "learning_rate": 9.902408634496307e-05,
+      "loss": 0.7655978202819824,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8293419633225458,
+      "grad_norm": 1.075892686843872,
+      "learning_rate": 9.901925300738424e-05,
+      "loss": 0.7199848890304565,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8302049622437971,
+      "grad_norm": 1.1517611742019653,
+      "learning_rate": 9.901440784903353e-05,
+      "loss": 0.8325693011283875,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8310679611650486,
+      "grad_norm": 1.2084628343582153,
+      "learning_rate": 9.900955087107933e-05,
+      "loss": 0.7674157023429871,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8319309600862999,
+      "grad_norm": 1.334679365158081,
+      "learning_rate": 9.900468207469287e-05,
+      "loss": 0.7415181398391724,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8327939590075513,
+      "grad_norm": 1.0505887269973755,
+      "learning_rate": 9.899980146104824e-05,
+      "loss": 0.672336220741272,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8336569579288026,
+      "grad_norm": 1.415797233581543,
+      "learning_rate": 9.899490903132238e-05,
+      "loss": 0.8077969551086426,
+      "step": 1932
+    },
+    {
+      "epoch": 0.834519956850054,
+      "grad_norm": 1.1793439388275146,
+      "learning_rate": 9.899000478669507e-05,
+      "loss": 0.8053045272827148,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8353829557713053,
+      "grad_norm": 1.196115493774414,
+      "learning_rate": 9.898508872834896e-05,
+      "loss": 0.7585914731025696,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8362459546925566,
+      "grad_norm": 1.2181763648986816,
+      "learning_rate": 9.898016085746953e-05,
+      "loss": 0.691072940826416,
+      "step": 1938
+    },
+    {
+      "epoch": 0.837108953613808,
+      "grad_norm": 1.2929240465164185,
+      "learning_rate": 9.897522117524512e-05,
+      "loss": 0.7723121643066406,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8379719525350593,
+      "grad_norm": 1.2859764099121094,
+      "learning_rate": 9.89702696828669e-05,
+      "loss": 0.8236626386642456,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8388349514563107,
+      "grad_norm": 1.2043784856796265,
+      "learning_rate": 9.89653063815289e-05,
+      "loss": 0.7315511703491211,
+      "step": 1944
+    },
+    {
+      "epoch": 0.839697950377562,
+      "grad_norm": 1.2063398361206055,
+      "learning_rate": 9.896033127242802e-05,
+      "loss": 0.6707082986831665,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8405609492988134,
+      "grad_norm": 1.1906715631484985,
+      "learning_rate": 9.895534435676393e-05,
+      "loss": 0.6920633912086487,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 1.122567892074585,
+      "learning_rate": 9.895034563573926e-05,
+      "loss": 0.7134768962860107,
+      "step": 1950
+    },
+    {
+      "epoch": 0.842286947141316,
+      "grad_norm": 1.0681787729263306,
+      "learning_rate": 9.894533511055943e-05,
+      "loss": 0.6333967447280884,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8431499460625674,
+      "grad_norm": 1.5216220617294312,
+      "learning_rate": 9.894031278243266e-05,
+      "loss": 0.7744026780128479,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8440129449838187,
+      "grad_norm": 1.0934535264968872,
+      "learning_rate": 9.89352786525701e-05,
+      "loss": 0.7678086757659912,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8448759439050701,
+      "grad_norm": 1.223285436630249,
+      "learning_rate": 9.89302327221857e-05,
+      "loss": 0.7344630360603333,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8457389428263214,
+      "grad_norm": 1.1123322248458862,
+      "learning_rate": 9.892517499249623e-05,
+      "loss": 0.76033616065979,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8466019417475729,
+      "grad_norm": 1.0472193956375122,
+      "learning_rate": 9.89201054647214e-05,
+      "loss": 0.7176017761230469,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8474649406688242,
+      "grad_norm": 1.8538368940353394,
+      "learning_rate": 9.891502414008365e-05,
+      "loss": 0.7220107316970825,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8483279395900755,
+      "grad_norm": 1.2280668020248413,
+      "learning_rate": 9.890993101980834e-05,
+      "loss": 0.7399556636810303,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8491909385113269,
+      "grad_norm": 1.1482552289962769,
+      "learning_rate": 9.890482610512364e-05,
+      "loss": 0.7700466513633728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8500539374325782,
+      "grad_norm": 1.1877986192703247,
+      "learning_rate": 9.889970939726061e-05,
+      "loss": 0.6666419506072998,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8509169363538296,
+      "grad_norm": 1.407792329788208,
+      "learning_rate": 9.88945808974531e-05,
+      "loss": 0.7883278131484985,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8517799352750809,
+      "grad_norm": 1.1426323652267456,
+      "learning_rate": 9.888944060693781e-05,
+      "loss": 0.7441611289978027,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8526429341963323,
+      "grad_norm": 1.2660622596740723,
+      "learning_rate": 9.888428852695432e-05,
+      "loss": 0.6464171409606934,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8535059331175836,
+      "grad_norm": 1.1639022827148438,
+      "learning_rate": 9.887912465874502e-05,
+      "loss": 0.7242478728294373,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 1.2546422481536865,
+      "learning_rate": 9.887394900355516e-05,
+      "loss": 0.7431875467300415,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8552319309600863,
+      "grad_norm": 1.1057130098342896,
+      "learning_rate": 9.886876156263282e-05,
+      "loss": 0.7662370204925537,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8560949298813376,
+      "grad_norm": 1.1017125844955444,
+      "learning_rate": 9.886356233722894e-05,
+      "loss": 0.7120938897132874,
+      "step": 1984
+    },
+    {
+      "epoch": 0.856957928802589,
+      "grad_norm": 1.3548704385757446,
+      "learning_rate": 9.885835132859727e-05,
+      "loss": 0.789219856262207,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8578209277238403,
+      "grad_norm": 1.1920398473739624,
+      "learning_rate": 9.885312853799443e-05,
+      "loss": 0.7599530816078186,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8586839266450917,
+      "grad_norm": 1.170946478843689,
+      "learning_rate": 9.884789396667988e-05,
+      "loss": 0.8315809369087219,
+      "step": 1990
+    },
+    {
+      "epoch": 0.859546925566343,
+      "grad_norm": 1.2530180215835571,
+      "learning_rate": 9.88426476159159e-05,
+      "loss": 0.713994026184082,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8604099244875943,
+      "grad_norm": 1.1056885719299316,
+      "learning_rate": 9.883738948696765e-05,
+      "loss": 0.6656507849693298,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8612729234088458,
+      "grad_norm": 1.169382929801941,
+      "learning_rate": 9.883211958110308e-05,
+      "loss": 0.7164257764816284,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8621359223300971,
+      "grad_norm": 1.1970325708389282,
+      "learning_rate": 9.882683789959299e-05,
+      "loss": 0.6885558366775513,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "grad_norm": 1.1700761318206787,
+      "learning_rate": 9.882154444371107e-05,
+      "loss": 0.7265812754631042,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "eval_loss": 0.7487027645111084,
+      "eval_runtime": 825.141,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8638619201725998,
+      "grad_norm": 1.3120243549346924,
+      "learning_rate": 9.88162392147338e-05,
+      "loss": 0.7371920347213745,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8647249190938512,
+      "grad_norm": 1.098697304725647,
+      "learning_rate": 9.881092221394049e-05,
+      "loss": 0.7453275918960571,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8655879180151025,
+      "grad_norm": 1.1961677074432373,
+      "learning_rate": 9.880559344261334e-05,
+      "loss": 0.7562432289123535,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8664509169363538,
+      "grad_norm": 1.1655291318893433,
+      "learning_rate": 9.880025290203735e-05,
+      "loss": 0.7654905319213867,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8673139158576052,
+      "grad_norm": 1.0979351997375488,
+      "learning_rate": 9.879490059350036e-05,
+      "loss": 0.6807888746261597,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8681769147788565,
+      "grad_norm": 1.0378475189208984,
+      "learning_rate": 9.878953651829306e-05,
+      "loss": 0.6887364387512207,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8690399137001079,
+      "grad_norm": 2.0228428840637207,
+      "learning_rate": 9.878416067770898e-05,
+      "loss": 0.718578577041626,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8699029126213592,
+      "grad_norm": 1.1610580682754517,
+      "learning_rate": 9.877877307304446e-05,
+      "loss": 0.7424600720405579,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8707659115426106,
+      "grad_norm": 1.1415833234786987,
+      "learning_rate": 9.877337370559873e-05,
+      "loss": 0.7450630068778992,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8716289104638619,
+      "grad_norm": 1.1008672714233398,
+      "learning_rate": 9.876796257667379e-05,
+      "loss": 0.7398725748062134,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8724919093851132,
+      "grad_norm": 1.1662763357162476,
+      "learning_rate": 9.876253968757452e-05,
+      "loss": 0.7122271060943604,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8733549083063646,
+      "grad_norm": 1.1835166215896606,
+      "learning_rate": 9.875710503960862e-05,
+      "loss": 0.7026485800743103,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8742179072276159,
+      "grad_norm": 1.1602214574813843,
+      "learning_rate": 9.875165863408666e-05,
+      "loss": 0.7211011052131653,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8750809061488674,
+      "grad_norm": 1.222589373588562,
+      "learning_rate": 9.874620047232198e-05,
+      "loss": 0.7565322518348694,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8759439050701187,
+      "grad_norm": 1.203661561012268,
+      "learning_rate": 9.87407305556308e-05,
+      "loss": 0.7173075675964355,
+      "step": 2030
+    },
+    {
+      "epoch": 0.87680690399137,
+      "grad_norm": 1.3316307067871094,
+      "learning_rate": 9.87352488853322e-05,
+      "loss": 0.820809006690979,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8776699029126214,
+      "grad_norm": 1.3071057796478271,
+      "learning_rate": 9.872975546274799e-05,
+      "loss": 0.8689352869987488,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8785329018338727,
+      "grad_norm": 1.3557701110839844,
+      "learning_rate": 9.872425028920294e-05,
+      "loss": 0.7090175151824951,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8793959007551241,
+      "grad_norm": 1.2234264612197876,
+      "learning_rate": 9.871873336602459e-05,
+      "loss": 0.6883202195167542,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8802588996763754,
+      "grad_norm": 1.3064714670181274,
+      "learning_rate": 9.87132046945433e-05,
+      "loss": 0.7134175896644592,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8811218985976268,
+      "grad_norm": 1.2502511739730835,
+      "learning_rate": 9.870766427609232e-05,
+      "loss": 0.7127227783203125,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8819848975188781,
+      "grad_norm": 1.1354259252548218,
+      "learning_rate": 9.870211211200766e-05,
+      "loss": 0.719437301158905,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8828478964401294,
+      "grad_norm": 1.1706442832946777,
+      "learning_rate": 9.869654820362821e-05,
+      "loss": 0.6933091282844543,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8837108953613808,
+      "grad_norm": 1.0227864980697632,
+      "learning_rate": 9.86909725522957e-05,
+      "loss": 0.6825079321861267,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8845738942826321,
+      "grad_norm": 1.103418231010437,
+      "learning_rate": 9.868538515935465e-05,
+      "loss": 0.7306133508682251,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8854368932038835,
+      "grad_norm": 1.0938395261764526,
+      "learning_rate": 9.867978602615245e-05,
+      "loss": 0.6956495046615601,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8862998921251348,
+      "grad_norm": 1.115692377090454,
+      "learning_rate": 9.86741751540393e-05,
+      "loss": 0.7740456461906433,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8871628910463862,
+      "grad_norm": 1.0264382362365723,
+      "learning_rate": 9.866855254436824e-05,
+      "loss": 0.6996002197265625,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8880258899676375,
+      "grad_norm": 1.2523163557052612,
+      "learning_rate": 9.866291819849511e-05,
+      "loss": 0.760357677936554,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.0595446825027466,
+      "learning_rate": 9.865727211777864e-05,
+      "loss": 0.6389813423156738,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8897518878101403,
+      "grad_norm": 1.198530912399292,
+      "learning_rate": 9.865161430358036e-05,
+      "loss": 0.7859060168266296,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8906148867313916,
+      "grad_norm": 1.0271193981170654,
+      "learning_rate": 9.864594475726461e-05,
+      "loss": 0.7120063900947571,
+      "step": 2064
+    },
+    {
+      "epoch": 0.891477885652643,
+      "grad_norm": 1.1619046926498413,
+      "learning_rate": 9.864026348019857e-05,
+      "loss": 0.6751878261566162,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8923408845738943,
+      "grad_norm": 1.2812464237213135,
+      "learning_rate": 9.863457047375227e-05,
+      "loss": 0.610360860824585,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 1.112628698348999,
+      "learning_rate": 9.862886573929855e-05,
+      "loss": 0.6540104746818542,
+      "step": 2070
+    },
+    {
+      "epoch": 0.894066882416397,
+      "grad_norm": 1.6257048845291138,
+      "learning_rate": 9.862314927821306e-05,
+      "loss": 0.726277232170105,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8949298813376483,
+      "grad_norm": 1.227334976196289,
+      "learning_rate": 9.861742109187432e-05,
+      "loss": 0.7798765301704407,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8957928802588997,
+      "grad_norm": 1.0728148221969604,
+      "learning_rate": 9.861168118166366e-05,
+      "loss": 0.637127161026001,
+      "step": 2076
+    },
+    {
+      "epoch": 0.896655879180151,
+      "grad_norm": 1.220282793045044,
+      "learning_rate": 9.860592954896523e-05,
+      "loss": 0.7332831025123596,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8975188781014024,
+      "grad_norm": 1.2287596464157104,
+      "learning_rate": 9.8600166195166e-05,
+      "loss": 0.7337175607681274,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8983818770226537,
+      "grad_norm": 1.0701712369918823,
+      "learning_rate": 9.859439112165579e-05,
+      "loss": 0.7422571182250977,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8992448759439051,
+      "grad_norm": 1.1583118438720703,
+      "learning_rate": 9.858860432982721e-05,
+      "loss": 0.7010003328323364,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9001078748651564,
+      "grad_norm": 0.9644797444343567,
+      "learning_rate": 9.858280582107575e-05,
+      "loss": 0.6211312413215637,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9009708737864077,
+      "grad_norm": 1.2448554039001465,
+      "learning_rate": 9.857699559679968e-05,
+      "loss": 0.7299860715866089,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9018338727076591,
+      "grad_norm": 1.2735580205917358,
+      "learning_rate": 9.85711736584001e-05,
+      "loss": 0.7461655735969543,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9026968716289104,
+      "grad_norm": 1.36841881275177,
+      "learning_rate": 9.856534000728093e-05,
+      "loss": 0.7218145132064819,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9035598705501618,
+      "grad_norm": 1.1471478939056396,
+      "learning_rate": 9.855949464484898e-05,
+      "loss": 0.7045709490776062,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9044228694714131,
+      "grad_norm": 1.058254361152649,
+      "learning_rate": 9.85536375725138e-05,
+      "loss": 0.7189744710922241,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9052858683926646,
+      "grad_norm": 1.4070709943771362,
+      "learning_rate": 9.854776879168779e-05,
+      "loss": 0.740775465965271,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 1.2182649374008179,
+      "learning_rate": 9.854188830378622e-05,
+      "loss": 0.8191262483596802,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "eval_loss": 0.7449111342430115,
+      "eval_runtime": 832.5135,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9070118662351672,
+      "grad_norm": 1.139978289604187,
+      "learning_rate": 9.853599611022708e-05,
+      "loss": 0.7295873761177063,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9078748651564186,
+      "grad_norm": 1.1802685260772705,
+      "learning_rate": 9.853009221243127e-05,
+      "loss": 0.7141753435134888,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9087378640776699,
+      "grad_norm": 1.076759934425354,
+      "learning_rate": 9.852417661182253e-05,
+      "loss": 0.7557607293128967,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9096008629989213,
+      "grad_norm": 1.1000943183898926,
+      "learning_rate": 9.851824930982733e-05,
+      "loss": 0.7025325298309326,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9104638619201726,
+      "grad_norm": 1.0235389471054077,
+      "learning_rate": 9.851231030787504e-05,
+      "loss": 0.6967638731002808,
+      "step": 2110
+    },
+    {
+      "epoch": 0.911326860841424,
+      "grad_norm": 1.2311428785324097,
+      "learning_rate": 9.850635960739782e-05,
+      "loss": 0.7890311479568481,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9121898597626753,
+      "grad_norm": 1.173770785331726,
+      "learning_rate": 9.850039720983064e-05,
+      "loss": 0.6824729442596436,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9130528586839266,
+      "grad_norm": 1.2165251970291138,
+      "learning_rate": 9.849442311661131e-05,
+      "loss": 0.7157722115516663,
+      "step": 2116
+    },
+    {
+      "epoch": 0.913915857605178,
+      "grad_norm": 1.251520037651062,
+      "learning_rate": 9.848843732918047e-05,
+      "loss": 0.7164111733436584,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9147788565264293,
+      "grad_norm": 1.1330616474151611,
+      "learning_rate": 9.848243984898153e-05,
+      "loss": 0.6830796003341675,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9156418554476807,
+      "grad_norm": 1.1052451133728027,
+      "learning_rate": 9.847643067746081e-05,
+      "loss": 0.7149171829223633,
+      "step": 2122
+    },
+    {
+      "epoch": 0.916504854368932,
+      "grad_norm": 1.0389703512191772,
+      "learning_rate": 9.847040981606735e-05,
+      "loss": 0.7213287949562073,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9173678532901833,
+      "grad_norm": 1.151895523071289,
+      "learning_rate": 9.846437726625308e-05,
+      "loss": 0.703312873840332,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9182308522114347,
+      "grad_norm": 1.1624029874801636,
+      "learning_rate": 9.845833302947269e-05,
+      "loss": 0.8059951066970825,
+      "step": 2128
+    },
+    {
+      "epoch": 0.919093851132686,
+      "grad_norm": 1.2247511148452759,
+      "learning_rate": 9.845227710718377e-05,
+      "loss": 0.735476553440094,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9199568500539375,
+      "grad_norm": 1.1649967432022095,
+      "learning_rate": 9.844620950084661e-05,
+      "loss": 0.8130571842193604,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9208198489751888,
+      "grad_norm": 1.1915717124938965,
+      "learning_rate": 9.844013021192447e-05,
+      "loss": 0.6908681392669678,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9216828478964402,
+      "grad_norm": 1.211199164390564,
+      "learning_rate": 9.843403924188328e-05,
+      "loss": 0.652173638343811,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9225458468176915,
+      "grad_norm": 1.267242193222046,
+      "learning_rate": 9.842793659219186e-05,
+      "loss": 0.6672233939170837,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9234088457389428,
+      "grad_norm": 1.7038761377334595,
+      "learning_rate": 9.842182226432186e-05,
+      "loss": 0.7274143695831299,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9242718446601942,
+      "grad_norm": 1.151426076889038,
+      "learning_rate": 9.841569625974769e-05,
+      "loss": 0.707998514175415,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9251348435814455,
+      "grad_norm": 1.0414749383926392,
+      "learning_rate": 9.840955857994663e-05,
+      "loss": 0.6639942526817322,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9259978425026969,
+      "grad_norm": 1.1045551300048828,
+      "learning_rate": 9.840340922639875e-05,
+      "loss": 0.7096070647239685,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9268608414239482,
+      "grad_norm": 1.1193996667861938,
+      "learning_rate": 9.839724820058695e-05,
+      "loss": 0.71246737241745,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9277238403451996,
+      "grad_norm": 1.2609117031097412,
+      "learning_rate": 9.839107550399692e-05,
+      "loss": 0.7351052761077881,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9285868392664509,
+      "grad_norm": 1.1127749681472778,
+      "learning_rate": 9.838489113811717e-05,
+      "loss": 0.7125008702278137,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9294498381877022,
+      "grad_norm": 1.4771665334701538,
+      "learning_rate": 9.837869510443905e-05,
+      "loss": 0.7891085743904114,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9303128371089536,
+      "grad_norm": 1.0701470375061035,
+      "learning_rate": 9.83724874044567e-05,
+      "loss": 0.7370918393135071,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9311758360302049,
+      "grad_norm": 1.2045820951461792,
+      "learning_rate": 9.836626803966708e-05,
+      "loss": 0.7543193697929382,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.1944687366485596,
+      "learning_rate": 9.836003701156996e-05,
+      "loss": 0.6903626322746277,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9329018338727076,
+      "grad_norm": 1.1035394668579102,
+      "learning_rate": 9.83537943216679e-05,
+      "loss": 0.6923050880432129,
+      "step": 2162
+    },
+    {
+      "epoch": 0.933764832793959,
+      "grad_norm": 1.3766103982925415,
+      "learning_rate": 9.834753997146633e-05,
+      "loss": 0.8099993467330933,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9346278317152104,
+      "grad_norm": 1.0874576568603516,
+      "learning_rate": 9.834127396247346e-05,
+      "loss": 0.7896219491958618,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9354908306364617,
+      "grad_norm": 1.1633459329605103,
+      "learning_rate": 9.833499629620028e-05,
+      "loss": 0.7659367918968201,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9363538295577131,
+      "grad_norm": 1.405423641204834,
+      "learning_rate": 9.832870697416066e-05,
+      "loss": 0.7474715709686279,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9372168284789644,
+      "grad_norm": 1.1794270277023315,
+      "learning_rate": 9.83224059978712e-05,
+      "loss": 0.7018750905990601,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9380798274002158,
+      "grad_norm": 1.082466959953308,
+      "learning_rate": 9.831609336885139e-05,
+      "loss": 0.6477128267288208,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9389428263214671,
+      "grad_norm": 0.9830297827720642,
+      "learning_rate": 9.830976908862346e-05,
+      "loss": 0.6516029238700867,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9398058252427185,
+      "grad_norm": 1.112289309501648,
+      "learning_rate": 9.83034331587125e-05,
+      "loss": 0.6793080568313599,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9406688241639698,
+      "grad_norm": 1.1212154626846313,
+      "learning_rate": 9.829708558064638e-05,
+      "loss": 0.7220560312271118,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9415318230852211,
+      "grad_norm": 1.246471643447876,
+      "learning_rate": 9.829072635595582e-05,
+      "loss": 0.7621577978134155,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9423948220064725,
+      "grad_norm": 1.1732288599014282,
+      "learning_rate": 9.828435548617428e-05,
+      "loss": 0.6988050937652588,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9432578209277238,
+      "grad_norm": 1.1767359972000122,
+      "learning_rate": 9.827797297283808e-05,
+      "loss": 0.7909345626831055,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9441208198489752,
+      "grad_norm": 1.007215976715088,
+      "learning_rate": 9.827157881748633e-05,
+      "loss": 0.6651493310928345,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9449838187702265,
+      "grad_norm": 1.184118628501892,
+      "learning_rate": 9.826517302166096e-05,
+      "loss": 0.7742198705673218,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9458468176914779,
+      "grad_norm": 1.1227030754089355,
+      "learning_rate": 9.82587555869067e-05,
+      "loss": 0.7614272236824036,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9467098166127292,
+      "grad_norm": 1.1030738353729248,
+      "learning_rate": 9.825232651477109e-05,
+      "loss": 0.7302108407020569,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9475728155339805,
+      "grad_norm": 1.5209563970565796,
+      "learning_rate": 9.824588580680445e-05,
+      "loss": 0.7408936023712158,
+      "step": 2196
+    },
+    {
+      "epoch": 0.948435814455232,
+      "grad_norm": 1.1853431463241577,
+      "learning_rate": 9.823943346455995e-05,
+      "loss": 0.8050196766853333,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "grad_norm": 1.285551905632019,
+      "learning_rate": 9.823296948959354e-05,
+      "loss": 0.726731538772583,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "eval_loss": 0.7400083541870117,
+      "eval_runtime": 838.7633,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9501618122977347,
+      "grad_norm": 1.2159804105758667,
+      "learning_rate": 9.822649388346396e-05,
+      "loss": 0.7061861753463745,
+      "step": 2202
+    },
+    {
+      "epoch": 0.951024811218986,
+      "grad_norm": 1.1189815998077393,
+      "learning_rate": 9.822000664773278e-05,
+      "loss": 0.824293851852417,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9518878101402374,
+      "grad_norm": 0.9968506097793579,
+      "learning_rate": 9.821350778396439e-05,
+      "loss": 0.7230689525604248,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9527508090614887,
+      "grad_norm": 1.101393222808838,
+      "learning_rate": 9.820699729372593e-05,
+      "loss": 0.6633790135383606,
+      "step": 2208
+    },
+    {
+      "epoch": 0.95361380798274,
+      "grad_norm": 1.084307312965393,
+      "learning_rate": 9.820047517858741e-05,
+      "loss": 0.6548227071762085,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9544768069039914,
+      "grad_norm": 1.1291496753692627,
+      "learning_rate": 9.819394144012158e-05,
+      "loss": 0.6434295177459717,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9553398058252427,
+      "grad_norm": 2.303483724594116,
+      "learning_rate": 9.818739607990404e-05,
+      "loss": 0.8122013211250305,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9562028047464941,
+      "grad_norm": 1.244541883468628,
+      "learning_rate": 9.818083909951315e-05,
+      "loss": 0.6120955348014832,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9570658036677454,
+      "grad_norm": 1.1051502227783203,
+      "learning_rate": 9.817427050053012e-05,
+      "loss": 0.7569628953933716,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9579288025889967,
+      "grad_norm": 1.3161911964416504,
+      "learning_rate": 9.816769028453893e-05,
+      "loss": 0.7588980197906494,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9587918015102481,
+      "grad_norm": 1.2526928186416626,
+      "learning_rate": 9.816109845312636e-05,
+      "loss": 0.743370771408081,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9596548004314994,
+      "grad_norm": 1.196215033531189,
+      "learning_rate": 9.815449500788202e-05,
+      "loss": 0.80516517162323,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9605177993527508,
+      "grad_norm": 1.3204832077026367,
+      "learning_rate": 9.81478799503983e-05,
+      "loss": 0.7331632971763611,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9613807982740021,
+      "grad_norm": 1.2442790269851685,
+      "learning_rate": 9.814125328227035e-05,
+      "loss": 0.6810339093208313,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9622437971952535,
+      "grad_norm": 1.1739158630371094,
+      "learning_rate": 9.813461500509623e-05,
+      "loss": 0.7269608378410339,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9631067961165048,
+      "grad_norm": 1.1016241312026978,
+      "learning_rate": 9.812796512047666e-05,
+      "loss": 0.7159227132797241,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9639697950377561,
+      "grad_norm": 1.1654549837112427,
+      "learning_rate": 9.812130363001529e-05,
+      "loss": 0.7145758867263794,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9648327939590076,
+      "grad_norm": 1.0521938800811768,
+      "learning_rate": 9.811463053531846e-05,
+      "loss": 0.6873835325241089,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9656957928802589,
+      "grad_norm": 1.2285776138305664,
+      "learning_rate": 9.810794583799538e-05,
+      "loss": 0.6683027148246765,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9665587918015103,
+      "grad_norm": 0.9826368689537048,
+      "learning_rate": 9.810124953965805e-05,
+      "loss": 0.663141131401062,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9674217907227616,
+      "grad_norm": 1.0102336406707764,
+      "learning_rate": 9.809454164192123e-05,
+      "loss": 0.6688778400421143,
+      "step": 2242
+    },
+    {
+      "epoch": 0.968284789644013,
+      "grad_norm": 0.9781575202941895,
+      "learning_rate": 9.808782214640251e-05,
+      "loss": 0.671240508556366,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9691477885652643,
+      "grad_norm": 1.137385368347168,
+      "learning_rate": 9.808109105472225e-05,
+      "loss": 0.7673252820968628,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9700107874865156,
+      "grad_norm": 1.3359450101852417,
+      "learning_rate": 9.807434836850362e-05,
+      "loss": 0.8085554838180542,
+      "step": 2248
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 1.150694727897644,
+      "learning_rate": 9.806759408937262e-05,
+      "loss": 0.6750202178955078,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9717367853290183,
+      "grad_norm": 1.080316424369812,
+      "learning_rate": 9.8060828218958e-05,
+      "loss": 0.5928418040275574,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9725997842502697,
+      "grad_norm": 1.0489610433578491,
+      "learning_rate": 9.805405075889129e-05,
+      "loss": 0.6947471499443054,
+      "step": 2254
+    },
+    {
+      "epoch": 0.973462783171521,
+      "grad_norm": 1.199119210243225,
+      "learning_rate": 9.80472617108069e-05,
+      "loss": 0.7273924350738525,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9743257820927724,
+      "grad_norm": 0.9830896258354187,
+      "learning_rate": 9.804046107634193e-05,
+      "loss": 0.668613612651825,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9751887810140237,
+      "grad_norm": 1.1655877828598022,
+      "learning_rate": 9.803364885713633e-05,
+      "loss": 0.7255772352218628,
+      "step": 2260
+    },
+    {
+      "epoch": 0.976051779935275,
+      "grad_norm": 1.0289462804794312,
+      "learning_rate": 9.802682505483284e-05,
+      "loss": 0.6865858435630798,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9769147788565264,
+      "grad_norm": 1.1020126342773438,
+      "learning_rate": 9.801998967107702e-05,
+      "loss": 0.6680663228034973,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 1.0846027135849,
+      "learning_rate": 9.801314270751716e-05,
+      "loss": 0.7476468086242676,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9786407766990292,
+      "grad_norm": 1.1834278106689453,
+      "learning_rate": 9.800628416580437e-05,
+      "loss": 0.7173228859901428,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9795037756202805,
+      "grad_norm": 1.0082321166992188,
+      "learning_rate": 9.799941404759257e-05,
+      "loss": 0.6768618226051331,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9803667745415319,
+      "grad_norm": 1.1573102474212646,
+      "learning_rate": 9.799253235453846e-05,
+      "loss": 0.7877311110496521,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9812297734627832,
+      "grad_norm": 1.2030092477798462,
+      "learning_rate": 9.798563908830152e-05,
+      "loss": 0.6843966841697693,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9820927723840345,
+      "grad_norm": 1.080560564994812,
+      "learning_rate": 9.797873425054404e-05,
+      "loss": 0.6084349155426025,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9829557713052859,
+      "grad_norm": 1.1488033533096313,
+      "learning_rate": 9.797181784293107e-05,
+      "loss": 0.6784868836402893,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9838187702265372,
+      "grad_norm": 1.144773006439209,
+      "learning_rate": 9.796488986713049e-05,
+      "loss": 0.6111500263214111,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9846817691477886,
+      "grad_norm": 1.1356254816055298,
+      "learning_rate": 9.795795032481297e-05,
+      "loss": 0.7634603977203369,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9855447680690399,
+      "grad_norm": 1.139446496963501,
+      "learning_rate": 9.79509992176519e-05,
+      "loss": 0.6470903158187866,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9864077669902913,
+      "grad_norm": 1.1564133167266846,
+      "learning_rate": 9.794403654732354e-05,
+      "loss": 0.71474289894104,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9872707659115426,
+      "grad_norm": 1.0605655908584595,
+      "learning_rate": 9.793706231550689e-05,
+      "loss": 0.6413968801498413,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9881337648327939,
+      "grad_norm": 1.1660863161087036,
+      "learning_rate": 9.793007652388378e-05,
+      "loss": 0.6595705151557922,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9889967637540453,
+      "grad_norm": 1.1661914587020874,
+      "learning_rate": 9.792307917413877e-05,
+      "loss": 0.7144066095352173,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9898597626752966,
+      "grad_norm": 1.2250727415084839,
+      "learning_rate": 9.791607026795926e-05,
+      "loss": 0.7956018447875977,
+      "step": 2294
+    },
+    {
+      "epoch": 0.990722761596548,
+      "grad_norm": 0.9593650698661804,
+      "learning_rate": 9.790904980703541e-05,
+      "loss": 0.5786031484603882,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9915857605177993,
+      "grad_norm": 1.0697777271270752,
+      "learning_rate": 9.790201779306018e-05,
+      "loss": 0.6817460656166077,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "grad_norm": 1.2658408880233765,
+      "learning_rate": 9.789497422772931e-05,
+      "loss": 0.6845148801803589,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "eval_loss": 0.7360134720802307,
+      "eval_runtime": 825.8559,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2300
+    },
+    {
+      "epoch": 0.993311758360302,
+      "grad_norm": 1.0869544744491577,
+      "learning_rate": 9.788791911274131e-05,
+      "loss": 0.6689707040786743,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9941747572815534,
+      "grad_norm": 1.0049875974655151,
+      "learning_rate": 9.788085244979748e-05,
+      "loss": 0.6470839381217957,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9950377562028048,
+      "grad_norm": 1.4471046924591064,
+      "learning_rate": 9.787377424060196e-05,
+      "loss": 0.8018016815185547,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9959007551240561,
+      "grad_norm": 1.3750648498535156,
+      "learning_rate": 9.786668448686159e-05,
+      "loss": 0.8078322410583496,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9967637540453075,
+      "grad_norm": 1.098283290863037,
+      "learning_rate": 9.7859583190286e-05,
+      "loss": 0.6740908622741699,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9976267529665588,
+      "grad_norm": 1.0312142372131348,
+      "learning_rate": 9.785247035258773e-05,
+      "loss": 0.6844386458396912,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9984897518878102,
+      "grad_norm": 1.4482319355010986,
+      "learning_rate": 9.784534597548193e-05,
+      "loss": 0.7547479271888733,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9993527508090615,
+      "grad_norm": 1.0897735357284546,
+      "learning_rate": 9.783821006068663e-05,
+      "loss": 0.7713987231254578,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7887805700302124,
+      "learning_rate": 9.783106260992264e-05,
+      "loss": 0.6868399381637573,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0008629989212514,
+      "grad_norm": 1.0021663904190063,
+      "learning_rate": 9.782390362491353e-05,
+      "loss": 0.696355938911438,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0017259978425026,
+      "grad_norm": 1.3650091886520386,
+      "learning_rate": 9.781673310738564e-05,
+      "loss": 0.7315902709960938,
+      "step": 2322
+    },
+    {
+      "epoch": 1.002588996763754,
+      "grad_norm": 1.2247875928878784,
+      "learning_rate": 9.780955105906813e-05,
+      "loss": 0.7667547464370728,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0034519956850054,
+      "grad_norm": 1.1917641162872314,
+      "learning_rate": 9.780235748169291e-05,
+      "loss": 0.7380582094192505,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0043149946062568,
+      "grad_norm": 1.146159052848816,
+      "learning_rate": 9.779515237699466e-05,
+      "loss": 0.6780733466148376,
+      "step": 2328
+    },
+    {
+      "epoch": 1.005177993527508,
+      "grad_norm": 1.4877620935440063,
+      "learning_rate": 9.778793574671089e-05,
+      "loss": 0.8116409182548523,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0060409924487594,
+      "grad_norm": 1.3488554954528809,
+      "learning_rate": 9.778070759258185e-05,
+      "loss": 0.65245521068573,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0069039913700109,
+      "grad_norm": 1.096686840057373,
+      "learning_rate": 9.777346791635056e-05,
+      "loss": 0.584334135055542,
+      "step": 2334
+    },
+    {
+      "epoch": 1.007766990291262,
+      "grad_norm": 1.063981056213379,
+      "learning_rate": 9.776621671976285e-05,
+      "loss": 0.6594714522361755,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0086299892125135,
+      "grad_norm": 1.2811797857284546,
+      "learning_rate": 9.775895400456729e-05,
+      "loss": 0.6963099837303162,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0094929881337649,
+      "grad_norm": 0.9491971731185913,
+      "learning_rate": 9.775167977251529e-05,
+      "loss": 0.6246375441551208,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0103559870550163,
+      "grad_norm": 1.198683500289917,
+      "learning_rate": 9.774439402536098e-05,
+      "loss": 0.6352161169052124,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0112189859762675,
+      "grad_norm": 1.2178813219070435,
+      "learning_rate": 9.773709676486129e-05,
+      "loss": 0.6884894967079163,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0120819848975189,
+      "grad_norm": 1.1905319690704346,
+      "learning_rate": 9.77297879927759e-05,
+      "loss": 0.5945048928260803,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0129449838187703,
+      "grad_norm": 1.210302710533142,
+      "learning_rate": 9.77224677108673e-05,
+      "loss": 0.6533923149108887,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0138079827400215,
+      "grad_norm": 1.15508234500885,
+      "learning_rate": 9.771513592090076e-05,
+      "loss": 0.6619405746459961,
+      "step": 2350
+    },
+    {
+      "epoch": 1.014670981661273,
+      "grad_norm": 1.0833762884140015,
+      "learning_rate": 9.77077926246443e-05,
+      "loss": 0.6581310033798218,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0155339805825243,
+      "grad_norm": 1.0330514907836914,
+      "learning_rate": 9.770043782386869e-05,
+      "loss": 0.6670945882797241,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0163969795037757,
+      "grad_norm": 1.1472195386886597,
+      "learning_rate": 9.769307152034755e-05,
+      "loss": 0.6801814436912537,
+      "step": 2356
+    },
+    {
+      "epoch": 1.017259978425027,
+      "grad_norm": 1.2204132080078125,
+      "learning_rate": 9.76856937158572e-05,
+      "loss": 0.7206465005874634,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0181229773462783,
+      "grad_norm": 1.2335962057113647,
+      "learning_rate": 9.76783044121768e-05,
+      "loss": 0.6884574294090271,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0189859762675297,
+      "grad_norm": 1.0858685970306396,
+      "learning_rate": 9.767090361108822e-05,
+      "loss": 0.655097484588623,
+      "step": 2362
+    },
+    {
+      "epoch": 1.019848975188781,
+      "grad_norm": 1.117907166481018,
+      "learning_rate": 9.766349131437614e-05,
+      "loss": 0.6478561758995056,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0207119741100323,
+      "grad_norm": 1.1995056867599487,
+      "learning_rate": 9.765606752382797e-05,
+      "loss": 0.6310951113700867,
+      "step": 2366
+    },
+    {
+      "epoch": 1.0215749730312838,
+      "grad_norm": 1.354819655418396,
+      "learning_rate": 9.764863224123397e-05,
+      "loss": 0.6373490691184998,
+      "step": 2368
+    },
+    {
+      "epoch": 1.0224379719525352,
+      "grad_norm": 1.2194819450378418,
+      "learning_rate": 9.764118546838712e-05,
+      "loss": 0.6595761179924011,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0233009708737864,
+      "grad_norm": 1.1485158205032349,
+      "learning_rate": 9.763372720708314e-05,
+      "loss": 0.6299480199813843,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0241639697950378,
+      "grad_norm": 1.1153662204742432,
+      "learning_rate": 9.76262574591206e-05,
+      "loss": 0.6788742542266846,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0250269687162892,
+      "grad_norm": 1.0690817832946777,
+      "learning_rate": 9.761877622630075e-05,
+      "loss": 0.559056282043457,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0258899676375404,
+      "grad_norm": 1.1896858215332031,
+      "learning_rate": 9.76112835104277e-05,
+      "loss": 0.7164555191993713,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0267529665587918,
+      "grad_norm": 1.1792134046554565,
+      "learning_rate": 9.760377931330824e-05,
+      "loss": 0.6791456937789917,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0276159654800432,
+      "grad_norm": 1.361763834953308,
+      "learning_rate": 9.759626363675202e-05,
+      "loss": 0.7421526312828064,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0284789644012946,
+      "grad_norm": 1.1960662603378296,
+      "learning_rate": 9.758873648257138e-05,
+      "loss": 0.7032517194747925,
+      "step": 2384
+    },
+    {
+      "epoch": 1.0293419633225458,
+      "grad_norm": 1.323255181312561,
+      "learning_rate": 9.758119785258146e-05,
+      "loss": 0.6953344345092773,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0302049622437972,
+      "grad_norm": 1.2343937158584595,
+      "learning_rate": 9.75736477486002e-05,
+      "loss": 0.7287829518318176,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0310679611650486,
+      "grad_norm": 1.0712062120437622,
+      "learning_rate": 9.756608617244823e-05,
+      "loss": 0.6812928915023804,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0319309600862998,
+      "grad_norm": 1.264726996421814,
+      "learning_rate": 9.755851312594902e-05,
+      "loss": 0.7173963785171509,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0327939590075512,
+      "grad_norm": 1.1563389301300049,
+      "learning_rate": 9.755092861092874e-05,
+      "loss": 0.6509836316108704,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0336569579288026,
+      "grad_norm": 1.326188087463379,
+      "learning_rate": 9.75433326292164e-05,
+      "loss": 0.718437135219574,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0345199568500538,
+      "grad_norm": 1.2263392210006714,
+      "learning_rate": 9.753572518264373e-05,
+      "loss": 0.664516031742096,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "grad_norm": 1.2874478101730347,
+      "learning_rate": 9.752810627304523e-05,
+      "loss": 0.6754546165466309,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "eval_loss": 0.7346783876419067,
+      "eval_runtime": 825.8356,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0362459546925566,
+      "grad_norm": 1.304518461227417,
+      "learning_rate": 9.752047590225814e-05,
+      "loss": 0.708167314529419,
+      "step": 2402
+    },
+    {
+      "epoch": 1.037108953613808,
+      "grad_norm": 1.21835196018219,
+      "learning_rate": 9.751283407212252e-05,
+      "loss": 0.7213018536567688,
+      "step": 2404
+    },
+    {
+      "epoch": 1.0379719525350593,
+      "grad_norm": 1.2947579622268677,
+      "learning_rate": 9.750518078448116e-05,
+      "loss": 0.6739231944084167,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 1.138368010520935,
+      "learning_rate": 9.749751604117959e-05,
+      "loss": 0.7671371102333069,
+      "step": 2408
+    },
+    {
+      "epoch": 1.039697950377562,
+      "grad_norm": 1.1255316734313965,
+      "learning_rate": 9.748983984406616e-05,
+      "loss": 0.6593353748321533,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0405609492988135,
+      "grad_norm": 1.2362794876098633,
+      "learning_rate": 9.748215219499193e-05,
+      "loss": 0.713056743144989,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0414239482200647,
+      "grad_norm": 1.2021304368972778,
+      "learning_rate": 9.747445309581076e-05,
+      "loss": 0.6989394426345825,
+      "step": 2414
+    },
+    {
+      "epoch": 1.042286947141316,
+      "grad_norm": 1.127341389656067,
+      "learning_rate": 9.746674254837925e-05,
+      "loss": 0.623035192489624,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0431499460625675,
+      "grad_norm": 1.1254255771636963,
+      "learning_rate": 9.745902055455672e-05,
+      "loss": 0.6490310430526733,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0440129449838187,
+      "grad_norm": 1.2142512798309326,
+      "learning_rate": 9.745128711620535e-05,
+      "loss": 0.6210505366325378,
+      "step": 2420
+    },
+    {
+      "epoch": 1.04487594390507,
+      "grad_norm": 1.1306709051132202,
+      "learning_rate": 9.744354223519001e-05,
+      "loss": 0.6243326663970947,
+      "step": 2422
+    },
+    {
+      "epoch": 1.0457389428263215,
+      "grad_norm": 1.202621340751648,
+      "learning_rate": 9.743578591337833e-05,
+      "loss": 0.6312575340270996,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0466019417475727,
+      "grad_norm": 1.0556137561798096,
+      "learning_rate": 9.742801815264072e-05,
+      "loss": 0.6190714836120605,
+      "step": 2426
+    },
+    {
+      "epoch": 1.0474649406688241,
+      "grad_norm": 1.5071814060211182,
+      "learning_rate": 9.742023895485033e-05,
+      "loss": 0.7496671080589294,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0483279395900755,
+      "grad_norm": 1.1448185443878174,
+      "learning_rate": 9.74124483218831e-05,
+      "loss": 0.6288847327232361,
+      "step": 2430
+    },
+    {
+      "epoch": 1.049190938511327,
+      "grad_norm": 1.0072964429855347,
+      "learning_rate": 9.74046462556177e-05,
+      "loss": 0.6938424706459045,
+      "step": 2432
+    },
+    {
+      "epoch": 1.0500539374325781,
+      "grad_norm": 1.198836326599121,
+      "learning_rate": 9.739683275793554e-05,
+      "loss": 0.7529639601707458,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0509169363538295,
+      "grad_norm": 1.2170380353927612,
+      "learning_rate": 9.738900783072083e-05,
+      "loss": 0.7193960547447205,
+      "step": 2436
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 1.0384336709976196,
+      "learning_rate": 9.738117147586051e-05,
+      "loss": 0.6488150954246521,
+      "step": 2438
+    },
+    {
+      "epoch": 1.0526429341963321,
+      "grad_norm": 1.1562492847442627,
+      "learning_rate": 9.737332369524429e-05,
+      "loss": 0.6672247052192688,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0535059331175836,
+      "grad_norm": 1.1553802490234375,
+      "learning_rate": 9.736546449076462e-05,
+      "loss": 0.651012122631073,
+      "step": 2442
+    },
+    {
+      "epoch": 1.054368932038835,
+      "grad_norm": 1.136838674545288,
+      "learning_rate": 9.735759386431671e-05,
+      "loss": 0.681246817111969,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0552319309600864,
+      "grad_norm": 1.177869200706482,
+      "learning_rate": 9.734971181779854e-05,
+      "loss": 0.7331019639968872,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0560949298813376,
+      "grad_norm": 1.2983415126800537,
+      "learning_rate": 9.73418183531108e-05,
+      "loss": 0.7316935658454895,
+      "step": 2448
+    },
+    {
+      "epoch": 1.056957928802589,
+      "grad_norm": 1.1189959049224854,
+      "learning_rate": 9.733391347215699e-05,
+      "loss": 0.6513608694076538,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0578209277238404,
+      "grad_norm": 1.3256937265396118,
+      "learning_rate": 9.732599717684334e-05,
+      "loss": 0.6469123363494873,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0586839266450916,
+      "grad_norm": 0.9766954779624939,
+      "learning_rate": 9.731806946907881e-05,
+      "loss": 0.5860153436660767,
+      "step": 2454
+    },
+    {
+      "epoch": 1.059546925566343,
+      "grad_norm": 1.1360441446304321,
+      "learning_rate": 9.731013035077514e-05,
+      "loss": 0.6783872842788696,
+      "step": 2456
+    },
+    {
+      "epoch": 1.0604099244875944,
+      "grad_norm": 1.3483080863952637,
+      "learning_rate": 9.73021798238468e-05,
+      "loss": 0.7894490361213684,
+      "step": 2458
+    },
+    {
+      "epoch": 1.0612729234088458,
+      "grad_norm": 1.1412159204483032,
+      "learning_rate": 9.729421789021105e-05,
+      "loss": 0.7098361849784851,
+      "step": 2460
+    },
+    {
+      "epoch": 1.062135922330097,
+      "grad_norm": 1.2670512199401855,
+      "learning_rate": 9.728624455178785e-05,
+      "loss": 0.7138097286224365,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0629989212513484,
+      "grad_norm": 1.2577205896377563,
+      "learning_rate": 9.727825981049994e-05,
+      "loss": 0.7136340141296387,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0638619201725998,
+      "grad_norm": 1.098997712135315,
+      "learning_rate": 9.727026366827283e-05,
+      "loss": 0.6273087859153748,
+      "step": 2466
+    },
+    {
+      "epoch": 1.064724919093851,
+      "grad_norm": 1.1550003290176392,
+      "learning_rate": 9.72622561270347e-05,
+      "loss": 0.7076449394226074,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0655879180151024,
+      "grad_norm": 1.0861129760742188,
+      "learning_rate": 9.725423718871659e-05,
+      "loss": 0.6856985092163086,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0664509169363539,
+      "grad_norm": 1.2742570638656616,
+      "learning_rate": 9.724620685525219e-05,
+      "loss": 0.7311446070671082,
+      "step": 2472
+    },
+    {
+      "epoch": 1.0673139158576053,
+      "grad_norm": 1.504339337348938,
+      "learning_rate": 9.7238165128578e-05,
+      "loss": 0.7728472948074341,
+      "step": 2474
+    },
+    {
+      "epoch": 1.0681769147788565,
+      "grad_norm": 1.2232030630111694,
+      "learning_rate": 9.723011201063323e-05,
+      "loss": 0.6259000301361084,
+      "step": 2476
+    },
+    {
+      "epoch": 1.0690399137001079,
+      "grad_norm": 1.3238953351974487,
+      "learning_rate": 9.722204750335989e-05,
+      "loss": 0.7132798433303833,
+      "step": 2478
+    },
+    {
+      "epoch": 1.0699029126213593,
+      "grad_norm": 1.2286484241485596,
+      "learning_rate": 9.721397160870265e-05,
+      "loss": 0.6067986488342285,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0707659115426105,
+      "grad_norm": 1.3101580142974854,
+      "learning_rate": 9.7205884328609e-05,
+      "loss": 0.6884648203849792,
+      "step": 2482
+    },
+    {
+      "epoch": 1.0716289104638619,
+      "grad_norm": 1.437833309173584,
+      "learning_rate": 9.719778566502915e-05,
+      "loss": 0.7164831757545471,
+      "step": 2484
+    },
+    {
+      "epoch": 1.0724919093851133,
+      "grad_norm": 1.0980390310287476,
+      "learning_rate": 9.718967561991606e-05,
+      "loss": 0.6326627731323242,
+      "step": 2486
+    },
+    {
+      "epoch": 1.0733549083063647,
+      "grad_norm": 0.9919570088386536,
+      "learning_rate": 9.718155419522543e-05,
+      "loss": 0.5712647438049316,
+      "step": 2488
+    },
+    {
+      "epoch": 1.074217907227616,
+      "grad_norm": 1.0773793458938599,
+      "learning_rate": 9.717342139291569e-05,
+      "loss": 0.6923331022262573,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0750809061488673,
+      "grad_norm": 1.1553318500518799,
+      "learning_rate": 9.716527721494805e-05,
+      "loss": 0.6737485527992249,
+      "step": 2492
+    },
+    {
+      "epoch": 1.0759439050701187,
+      "grad_norm": 1.112865924835205,
+      "learning_rate": 9.715712166328644e-05,
+      "loss": 0.6422063708305359,
+      "step": 2494
+    },
+    {
+      "epoch": 1.07680690399137,
+      "grad_norm": 1.0651512145996094,
+      "learning_rate": 9.71489547398975e-05,
+      "loss": 0.6399799585342407,
+      "step": 2496
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 1.1394895315170288,
+      "learning_rate": 9.71407764467507e-05,
+      "loss": 0.6218312382698059,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "grad_norm": 1.3323278427124023,
+      "learning_rate": 9.713258678581815e-05,
+      "loss": 0.65947425365448,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "eval_loss": 0.7294915914535522,
+      "eval_runtime": 836.4643,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0793959007551241,
+      "grad_norm": 1.3122533559799194,
+      "learning_rate": 9.712438575907479e-05,
+      "loss": 0.66681307554245,
+      "step": 2502
+    },
+    {
+      "epoch": 1.0802588996763753,
+      "grad_norm": 1.3631974458694458,
+      "learning_rate": 9.711617336849824e-05,
+      "loss": 0.6926553249359131,
+      "step": 2504
+    },
+    {
+      "epoch": 1.0811218985976268,
+      "grad_norm": 1.2383692264556885,
+      "learning_rate": 9.710794961606886e-05,
+      "loss": 0.6706596612930298,
+      "step": 2506
+    },
+    {
+      "epoch": 1.0819848975188782,
+      "grad_norm": 1.1660504341125488,
+      "learning_rate": 9.709971450376982e-05,
+      "loss": 0.6011359095573425,
+      "step": 2508
+    },
+    {
+      "epoch": 1.0828478964401294,
+      "grad_norm": 1.3021246194839478,
+      "learning_rate": 9.709146803358693e-05,
+      "loss": 0.689009428024292,
+      "step": 2510
+    },
+    {
+      "epoch": 1.0837108953613808,
+      "grad_norm": 1.0553252696990967,
+      "learning_rate": 9.708321020750883e-05,
+      "loss": 0.6216740012168884,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0845738942826322,
+      "grad_norm": 1.1862177848815918,
+      "learning_rate": 9.707494102752683e-05,
+      "loss": 0.6439616680145264,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0854368932038836,
+      "grad_norm": 1.1166558265686035,
+      "learning_rate": 9.706666049563504e-05,
+      "loss": 0.6825361251831055,
+      "step": 2516
+    },
+    {
+      "epoch": 1.0862998921251348,
+      "grad_norm": 1.128812551498413,
+      "learning_rate": 9.705836861383022e-05,
+      "loss": 0.6471308469772339,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0871628910463862,
+      "grad_norm": 1.324802279472351,
+      "learning_rate": 9.705006538411197e-05,
+      "loss": 0.6535511016845703,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0880258899676376,
+      "grad_norm": 1.3586591482162476,
+      "learning_rate": 9.704175080848255e-05,
+      "loss": 0.678022563457489,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 1.1334232091903687,
+      "learning_rate": 9.703342488894697e-05,
+      "loss": 0.5904073119163513,
+      "step": 2524
+    },
+    {
+      "epoch": 1.0897518878101402,
+      "grad_norm": 1.182215929031372,
+      "learning_rate": 9.702508762751303e-05,
+      "loss": 0.6072601079940796,
+      "step": 2526
+    },
+    {
+      "epoch": 1.0906148867313916,
+      "grad_norm": 1.2525769472122192,
+      "learning_rate": 9.701673902619119e-05,
+      "loss": 0.6752429604530334,
+      "step": 2528
+    },
+    {
+      "epoch": 1.091477885652643,
+      "grad_norm": 1.2811026573181152,
+      "learning_rate": 9.70083790869947e-05,
+      "loss": 0.7192611694335938,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0923408845738942,
+      "grad_norm": 1.1434061527252197,
+      "learning_rate": 9.700000781193949e-05,
+      "loss": 0.6991751194000244,
+      "step": 2532
+    },
+    {
+      "epoch": 1.0932038834951456,
+      "grad_norm": 1.26658296585083,
+      "learning_rate": 9.699162520304428e-05,
+      "loss": 0.7010498046875,
+      "step": 2534
+    },
+    {
+      "epoch": 1.094066882416397,
+      "grad_norm": 1.2083529233932495,
+      "learning_rate": 9.698323126233048e-05,
+      "loss": 0.6945615410804749,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0949298813376482,
+      "grad_norm": 1.2064114809036255,
+      "learning_rate": 9.697482599182227e-05,
+      "loss": 0.689496636390686,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0957928802588996,
+      "grad_norm": 1.17113196849823,
+      "learning_rate": 9.696640939354655e-05,
+      "loss": 0.6966884136199951,
+      "step": 2540
+    },
+    {
+      "epoch": 1.096655879180151,
+      "grad_norm": 1.1551356315612793,
+      "learning_rate": 9.695798146953291e-05,
+      "loss": 0.6524666547775269,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0975188781014025,
+      "grad_norm": 1.0966521501541138,
+      "learning_rate": 9.694954222181374e-05,
+      "loss": 0.5769726037979126,
+      "step": 2544
+    },
+    {
+      "epoch": 1.0983818770226537,
+      "grad_norm": 1.320669174194336,
+      "learning_rate": 9.694109165242411e-05,
+      "loss": 0.6986333727836609,
+      "step": 2546
+    },
+    {
+      "epoch": 1.099244875943905,
+      "grad_norm": 1.1678471565246582,
+      "learning_rate": 9.693262976340183e-05,
+      "loss": 0.7014455199241638,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1001078748651565,
+      "grad_norm": 1.1527155637741089,
+      "learning_rate": 9.692415655678745e-05,
+      "loss": 0.5786730647087097,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1009708737864077,
+      "grad_norm": 1.2511217594146729,
+      "learning_rate": 9.691567203462428e-05,
+      "loss": 0.755945086479187,
+      "step": 2552
+    },
+    {
+      "epoch": 1.101833872707659,
+      "grad_norm": 1.0910221338272095,
+      "learning_rate": 9.690717619895828e-05,
+      "loss": 0.6487610936164856,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1026968716289105,
+      "grad_norm": 1.3682197332382202,
+      "learning_rate": 9.68986690518382e-05,
+      "loss": 0.6346665024757385,
+      "step": 2556
+    },
+    {
+      "epoch": 1.103559870550162,
+      "grad_norm": 1.2161564826965332,
+      "learning_rate": 9.68901505953155e-05,
+      "loss": 0.647144615650177,
+      "step": 2558
+    },
+    {
+      "epoch": 1.104422869471413,
+      "grad_norm": 1.2559088468551636,
+      "learning_rate": 9.688162083144436e-05,
+      "loss": 0.6461755633354187,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1052858683926645,
+      "grad_norm": 1.1260225772857666,
+      "learning_rate": 9.68730797622817e-05,
+      "loss": 0.6888763904571533,
+      "step": 2562
+    },
+    {
+      "epoch": 1.106148867313916,
+      "grad_norm": 1.1499868631362915,
+      "learning_rate": 9.686452738988716e-05,
+      "loss": 0.6318327784538269,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1070118662351671,
+      "grad_norm": 1.1775025129318237,
+      "learning_rate": 9.68559637163231e-05,
+      "loss": 0.6538340449333191,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1078748651564185,
+      "grad_norm": 1.264656901359558,
+      "learning_rate": 9.684738874365465e-05,
+      "loss": 0.6706726551055908,
+      "step": 2568
+    },
+    {
+      "epoch": 1.10873786407767,
+      "grad_norm": 1.1130270957946777,
+      "learning_rate": 9.683880247394956e-05,
+      "loss": 0.7573289275169373,
+      "step": 2570
+    },
+    {
+      "epoch": 1.1096008629989211,
+      "grad_norm": 1.2781217098236084,
+      "learning_rate": 9.683020490927841e-05,
+      "loss": 0.6603306531906128,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1104638619201725,
+      "grad_norm": 1.251062035560608,
+      "learning_rate": 9.682159605171446e-05,
+      "loss": 0.6523779630661011,
+      "step": 2574
+    },
+    {
+      "epoch": 1.111326860841424,
+      "grad_norm": 1.2631903886795044,
+      "learning_rate": 9.68129759033337e-05,
+      "loss": 0.677550733089447,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1121898597626754,
+      "grad_norm": 1.0212602615356445,
+      "learning_rate": 9.680434446621483e-05,
+      "loss": 0.666540801525116,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1130528586839266,
+      "grad_norm": 1.221718430519104,
+      "learning_rate": 9.679570174243929e-05,
+      "loss": 0.7467086315155029,
+      "step": 2580
+    },
+    {
+      "epoch": 1.113915857605178,
+      "grad_norm": 1.1175146102905273,
+      "learning_rate": 9.678704773409124e-05,
+      "loss": 0.673492968082428,
+      "step": 2582
+    },
+    {
+      "epoch": 1.1147788565264294,
+      "grad_norm": 1.1274988651275635,
+      "learning_rate": 9.677838244325754e-05,
+      "loss": 0.5724847912788391,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1156418554476808,
+      "grad_norm": 1.297654390335083,
+      "learning_rate": 9.676970587202779e-05,
+      "loss": 0.634985625743866,
+      "step": 2586
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 1.1908061504364014,
+      "learning_rate": 9.67610180224943e-05,
+      "loss": 0.6813890933990479,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1173678532901834,
+      "grad_norm": 1.1787118911743164,
+      "learning_rate": 9.675231889675212e-05,
+      "loss": 0.6738101243972778,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1182308522114348,
+      "grad_norm": 1.1327152252197266,
+      "learning_rate": 9.674360849689899e-05,
+      "loss": 0.6662938594818115,
+      "step": 2592
+    },
+    {
+      "epoch": 1.119093851132686,
+      "grad_norm": 1.1246299743652344,
+      "learning_rate": 9.67348868250354e-05,
+      "loss": 0.677423357963562,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1199568500539374,
+      "grad_norm": 1.1411702632904053,
+      "learning_rate": 9.672615388326454e-05,
+      "loss": 0.6257020235061646,
+      "step": 2596
+    },
+    {
+      "epoch": 1.1208198489751888,
+      "grad_norm": 1.1607346534729004,
+      "learning_rate": 9.671740967369231e-05,
+      "loss": 0.7029470205307007,
+      "step": 2598
+    },
+    {
+      "epoch": 1.12168284789644,
+      "grad_norm": 1.047934651374817,
+      "learning_rate": 9.670865419842735e-05,
+      "loss": 0.6040990352630615,
+      "step": 2600
+    },
+    {
+      "epoch": 1.12168284789644,
+      "eval_loss": 0.7287447452545166,
+      "eval_runtime": 831.8138,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2600
+    },
+    {
+      "epoch": 1.1225458468176914,
+      "grad_norm": 1.0571783781051636,
+      "learning_rate": 9.669988745958098e-05,
+      "loss": 0.6872879266738892,
+      "step": 2602
+    },
+    {
+      "epoch": 1.1234088457389428,
+      "grad_norm": 1.0940403938293457,
+      "learning_rate": 9.669110945926727e-05,
+      "loss": 0.6173275113105774,
+      "step": 2604
+    },
+    {
+      "epoch": 1.1242718446601943,
+      "grad_norm": 1.1180222034454346,
+      "learning_rate": 9.6682320199603e-05,
+      "loss": 0.6142664551734924,
+      "step": 2606
+    },
+    {
+      "epoch": 1.1251348435814454,
+      "grad_norm": 1.251098871231079,
+      "learning_rate": 9.667351968270768e-05,
+      "loss": 0.7225713133811951,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1259978425026969,
+      "grad_norm": 1.2333059310913086,
+      "learning_rate": 9.666470791070347e-05,
+      "loss": 0.6830481290817261,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1268608414239483,
+      "grad_norm": 1.2904123067855835,
+      "learning_rate": 9.665588488571531e-05,
+      "loss": 0.6835238337516785,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1277238403451997,
+      "grad_norm": 1.4779443740844727,
+      "learning_rate": 9.664705060987085e-05,
+      "loss": 0.6993544697761536,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1285868392664509,
+      "grad_norm": 1.267021656036377,
+      "learning_rate": 9.663820508530042e-05,
+      "loss": 0.6819394826889038,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1294498381877023,
+      "grad_norm": 1.0755046606063843,
+      "learning_rate": 9.662934831413708e-05,
+      "loss": 0.6200308799743652,
+      "step": 2618
+    },
+    {
+      "epoch": 1.1303128371089537,
+      "grad_norm": 1.2043033838272095,
+      "learning_rate": 9.66204802985166e-05,
+      "loss": 0.7138658165931702,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1311758360302049,
+      "grad_norm": 1.2641128301620483,
+      "learning_rate": 9.661160104057744e-05,
+      "loss": 0.7238332033157349,
+      "step": 2622
+    },
+    {
+      "epoch": 1.1320388349514563,
+      "grad_norm": 1.0422990322113037,
+      "learning_rate": 9.660271054246085e-05,
+      "loss": 0.6401007175445557,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1329018338727077,
+      "grad_norm": 1.2298130989074707,
+      "learning_rate": 9.659380880631068e-05,
+      "loss": 0.6305879950523376,
+      "step": 2626
+    },
+    {
+      "epoch": 1.133764832793959,
+      "grad_norm": 1.254573106765747,
+      "learning_rate": 9.658489583427357e-05,
+      "loss": 0.5751872062683105,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1346278317152103,
+      "grad_norm": 1.15679132938385,
+      "learning_rate": 9.657597162849883e-05,
+      "loss": 0.6629595756530762,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1354908306364617,
+      "grad_norm": 1.193629264831543,
+      "learning_rate": 9.656703619113849e-05,
+      "loss": 0.6832258105278015,
+      "step": 2632
+    },
+    {
+      "epoch": 1.1363538295577131,
+      "grad_norm": 1.2330398559570312,
+      "learning_rate": 9.655808952434732e-05,
+      "loss": 0.6823931932449341,
+      "step": 2634
+    },
+    {
+      "epoch": 1.1372168284789643,
+      "grad_norm": 1.3926069736480713,
+      "learning_rate": 9.654913163028272e-05,
+      "loss": 0.6671395897865295,
+      "step": 2636
+    },
+    {
+      "epoch": 1.1380798274002157,
+      "grad_norm": 1.2995383739471436,
+      "learning_rate": 9.65401625111049e-05,
+      "loss": 0.662219226360321,
+      "step": 2638
+    },
+    {
+      "epoch": 1.1389428263214671,
+      "grad_norm": 1.1584415435791016,
+      "learning_rate": 9.653118216897667e-05,
+      "loss": 0.6419461965560913,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1398058252427186,
+      "grad_norm": 1.3584444522857666,
+      "learning_rate": 9.652219060606364e-05,
+      "loss": 0.6844428181648254,
+      "step": 2642
+    },
+    {
+      "epoch": 1.1406688241639698,
+      "grad_norm": 1.2619858980178833,
+      "learning_rate": 9.651318782453407e-05,
+      "loss": 0.6761999130249023,
+      "step": 2644
+    },
+    {
+      "epoch": 1.1415318230852212,
+      "grad_norm": 1.2202177047729492,
+      "learning_rate": 9.650417382655892e-05,
+      "loss": 0.7258123159408569,
+      "step": 2646
+    },
+    {
+      "epoch": 1.1423948220064726,
+      "grad_norm": 1.0225578546524048,
+      "learning_rate": 9.64951486143119e-05,
+      "loss": 0.5709917545318604,
+      "step": 2648
+    },
+    {
+      "epoch": 1.1432578209277238,
+      "grad_norm": 1.0961005687713623,
+      "learning_rate": 9.64861121899694e-05,
+      "loss": 0.7353709936141968,
+      "step": 2650
+    },
+    {
+      "epoch": 1.1441208198489752,
+      "grad_norm": 0.9393200874328613,
+      "learning_rate": 9.64770645557105e-05,
+      "loss": 0.7083833813667297,
+      "step": 2652
+    },
+    {
+      "epoch": 1.1449838187702266,
+      "grad_norm": 1.089733362197876,
+      "learning_rate": 9.646800571371701e-05,
+      "loss": 0.6405303478240967,
+      "step": 2654
+    },
+    {
+      "epoch": 1.1458468176914778,
+      "grad_norm": 1.1581398248672485,
+      "learning_rate": 9.645893566617341e-05,
+      "loss": 0.617733359336853,
+      "step": 2656
+    },
+    {
+      "epoch": 1.1467098166127292,
+      "grad_norm": 1.1336220502853394,
+      "learning_rate": 9.644985441526693e-05,
+      "loss": 0.6948232650756836,
+      "step": 2658
+    },
+    {
+      "epoch": 1.1475728155339806,
+      "grad_norm": 1.028591513633728,
+      "learning_rate": 9.644076196318743e-05,
+      "loss": 0.6392919421195984,
+      "step": 2660
+    },
+    {
+      "epoch": 1.148435814455232,
+      "grad_norm": 1.2088173627853394,
+      "learning_rate": 9.643165831212756e-05,
+      "loss": 0.6384931206703186,
+      "step": 2662
+    },
+    {
+      "epoch": 1.1492988133764832,
+      "grad_norm": 0.9905838966369629,
+      "learning_rate": 9.642254346428259e-05,
+      "loss": 0.5909387469291687,
+      "step": 2664
+    },
+    {
+      "epoch": 1.1501618122977346,
+      "grad_norm": 1.1670424938201904,
+      "learning_rate": 9.641341742185054e-05,
+      "loss": 0.700904130935669,
+      "step": 2666
+    },
+    {
+      "epoch": 1.151024811218986,
+      "grad_norm": 1.164746880531311,
+      "learning_rate": 9.640428018703211e-05,
+      "loss": 0.6036953330039978,
+      "step": 2668
+    },
+    {
+      "epoch": 1.1518878101402374,
+      "grad_norm": 1.00895357131958,
+      "learning_rate": 9.639513176203072e-05,
+      "loss": 0.6158549785614014,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1527508090614886,
+      "grad_norm": 1.2236207723617554,
+      "learning_rate": 9.638597214905244e-05,
+      "loss": 0.7280431389808655,
+      "step": 2672
+    },
+    {
+      "epoch": 1.15361380798274,
+      "grad_norm": 1.0930732488632202,
+      "learning_rate": 9.637680135030609e-05,
+      "loss": 0.6434410810470581,
+      "step": 2674
+    },
+    {
+      "epoch": 1.1544768069039915,
+      "grad_norm": 1.366325855255127,
+      "learning_rate": 9.636761936800315e-05,
+      "loss": 0.6771191954612732,
+      "step": 2676
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 2.249835968017578,
+      "learning_rate": 9.635842620435785e-05,
+      "loss": 0.6382432579994202,
+      "step": 2678
+    },
+    {
+      "epoch": 1.156202804746494,
+      "grad_norm": 1.274483561515808,
+      "learning_rate": 9.634922186158704e-05,
+      "loss": 0.6733657121658325,
+      "step": 2680
+    },
+    {
+      "epoch": 1.1570658036677455,
+      "grad_norm": 1.3715782165527344,
+      "learning_rate": 9.634000634191033e-05,
+      "loss": 0.7601200938224792,
+      "step": 2682
+    },
+    {
+      "epoch": 1.1579288025889967,
+      "grad_norm": 1.1312460899353027,
+      "learning_rate": 9.633077964754998e-05,
+      "loss": 0.5959098935127258,
+      "step": 2684
+    },
+    {
+      "epoch": 1.158791801510248,
+      "grad_norm": 1.063233494758606,
+      "learning_rate": 9.632154178073098e-05,
+      "loss": 0.615096926689148,
+      "step": 2686
+    },
+    {
+      "epoch": 1.1596548004314995,
+      "grad_norm": 1.074887752532959,
+      "learning_rate": 9.6312292743681e-05,
+      "loss": 0.6553784608840942,
+      "step": 2688
+    },
+    {
+      "epoch": 1.160517799352751,
+      "grad_norm": 1.17954683303833,
+      "learning_rate": 9.630303253863042e-05,
+      "loss": 0.6963269114494324,
+      "step": 2690
+    },
+    {
+      "epoch": 1.161380798274002,
+      "grad_norm": 1.2549158334732056,
+      "learning_rate": 9.629376116781226e-05,
+      "loss": 0.6633999347686768,
+      "step": 2692
+    },
+    {
+      "epoch": 1.1622437971952535,
+      "grad_norm": 1.237196683883667,
+      "learning_rate": 9.628447863346229e-05,
+      "loss": 0.6727597117424011,
+      "step": 2694
+    },
+    {
+      "epoch": 1.163106796116505,
+      "grad_norm": 1.4047741889953613,
+      "learning_rate": 9.627518493781895e-05,
+      "loss": 0.7018169164657593,
+      "step": 2696
+    },
+    {
+      "epoch": 1.1639697950377563,
+      "grad_norm": 1.15412175655365,
+      "learning_rate": 9.626588008312339e-05,
+      "loss": 0.6566510796546936,
+      "step": 2698
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "grad_norm": 1.194110631942749,
+      "learning_rate": 9.625656407161938e-05,
+      "loss": 0.7017931342124939,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "eval_loss": 0.72404944896698,
+      "eval_runtime": 824.2877,
+      "eval_samples_per_second": 2.499,
+      "eval_steps_per_second": 2.499,
+      "step": 2700
+    },
+    {
+      "epoch": 1.165695792880259,
+      "grad_norm": 1.1512954235076904,
+      "learning_rate": 9.62472369055535e-05,
+      "loss": 0.7058532238006592,
+      "step": 2702
+    },
+    {
+      "epoch": 1.1665587918015103,
+      "grad_norm": 1.4743460416793823,
+      "learning_rate": 9.623789858717492e-05,
+      "loss": 0.7458376288414001,
+      "step": 2704
+    },
+    {
+      "epoch": 1.1674217907227615,
+      "grad_norm": 1.0526450872421265,
+      "learning_rate": 9.622854911873551e-05,
+      "loss": 0.6526424288749695,
+      "step": 2706
+    },
+    {
+      "epoch": 1.168284789644013,
+      "grad_norm": 1.0408949851989746,
+      "learning_rate": 9.62191885024899e-05,
+      "loss": 0.6055451035499573,
+      "step": 2708
+    },
+    {
+      "epoch": 1.1691477885652644,
+      "grad_norm": 1.214052677154541,
+      "learning_rate": 9.620981674069534e-05,
+      "loss": 0.666462779045105,
+      "step": 2710
+    },
+    {
+      "epoch": 1.1700107874865155,
+      "grad_norm": 1.2526963949203491,
+      "learning_rate": 9.62004338356118e-05,
+      "loss": 0.7208677530288696,
+      "step": 2712
+    },
+    {
+      "epoch": 1.170873786407767,
+      "grad_norm": 1.2097395658493042,
+      "learning_rate": 9.61910397895019e-05,
+      "loss": 0.6269269585609436,
+      "step": 2714
+    },
+    {
+      "epoch": 1.1717367853290184,
+      "grad_norm": 1.035733699798584,
+      "learning_rate": 9.618163460463096e-05,
+      "loss": 0.6063299179077148,
+      "step": 2716
+    },
+    {
+      "epoch": 1.1725997842502698,
+      "grad_norm": 1.2881336212158203,
+      "learning_rate": 9.617221828326706e-05,
+      "loss": 0.7084641456604004,
+      "step": 2718
+    },
+    {
+      "epoch": 1.173462783171521,
+      "grad_norm": 1.2808257341384888,
+      "learning_rate": 9.616279082768085e-05,
+      "loss": 0.6515718102455139,
+      "step": 2720
+    },
+    {
+      "epoch": 1.1743257820927724,
+      "grad_norm": 1.1174249649047852,
+      "learning_rate": 9.615335224014574e-05,
+      "loss": 0.6745902299880981,
+      "step": 2722
+    },
+    {
+      "epoch": 1.1751887810140238,
+      "grad_norm": 1.1222290992736816,
+      "learning_rate": 9.61439025229378e-05,
+      "loss": 0.6824861168861389,
+      "step": 2724
+    },
+    {
+      "epoch": 1.176051779935275,
+      "grad_norm": 1.156187891960144,
+      "learning_rate": 9.613444167833579e-05,
+      "loss": 0.6020817160606384,
+      "step": 2726
+    },
+    {
+      "epoch": 1.1769147788565264,
+      "grad_norm": 1.2502888441085815,
+      "learning_rate": 9.612496970862113e-05,
+      "loss": 0.6430572271347046,
+      "step": 2728
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 1.6157052516937256,
+      "learning_rate": 9.611548661607797e-05,
+      "loss": 0.7195103168487549,
+      "step": 2730
+    },
+    {
+      "epoch": 1.1786407766990292,
+      "grad_norm": 1.3196234703063965,
+      "learning_rate": 9.610599240299308e-05,
+      "loss": 0.710418164730072,
+      "step": 2732
+    },
+    {
+      "epoch": 1.1795037756202804,
+      "grad_norm": 1.2117719650268555,
+      "learning_rate": 9.6096487071656e-05,
+      "loss": 0.650095522403717,
+      "step": 2734
+    },
+    {
+      "epoch": 1.1803667745415318,
+      "grad_norm": 2.2723631858825684,
+      "learning_rate": 9.608697062435884e-05,
+      "loss": 0.7150285243988037,
+      "step": 2736
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 1.0523152351379395,
+      "learning_rate": 9.607744306339647e-05,
+      "loss": 0.6271997690200806,
+      "step": 2738
+    },
+    {
+      "epoch": 1.1820927723840344,
+      "grad_norm": 1.2567129135131836,
+      "learning_rate": 9.606790439106646e-05,
+      "loss": 0.659799337387085,
+      "step": 2740
+    },
+    {
+      "epoch": 1.1829557713052858,
+      "grad_norm": 1.1448849439620972,
+      "learning_rate": 9.605835460966896e-05,
+      "loss": 0.7097326517105103,
+      "step": 2742
+    },
+    {
+      "epoch": 1.1838187702265373,
+      "grad_norm": 1.3698351383209229,
+      "learning_rate": 9.604879372150688e-05,
+      "loss": 0.6692788004875183,
+      "step": 2744
+    },
+    {
+      "epoch": 1.1846817691477884,
+      "grad_norm": 1.1424204111099243,
+      "learning_rate": 9.603922172888577e-05,
+      "loss": 0.6453474760055542,
+      "step": 2746
+    },
+    {
+      "epoch": 1.1855447680690399,
+      "grad_norm": 1.2701301574707031,
+      "learning_rate": 9.60296386341139e-05,
+      "loss": 0.6197408437728882,
+      "step": 2748
+    },
+    {
+      "epoch": 1.1864077669902913,
+      "grad_norm": 1.1171884536743164,
+      "learning_rate": 9.602004443950217e-05,
+      "loss": 0.6525035500526428,
+      "step": 2750
+    },
+    {
+      "epoch": 1.1872707659115427,
+      "grad_norm": 1.132633090019226,
+      "learning_rate": 9.60104391473642e-05,
+      "loss": 0.6459797620773315,
+      "step": 2752
+    },
+    {
+      "epoch": 1.1881337648327939,
+      "grad_norm": 1.1362011432647705,
+      "learning_rate": 9.600082276001623e-05,
+      "loss": 0.6175895929336548,
+      "step": 2754
+    },
+    {
+      "epoch": 1.1889967637540453,
+      "grad_norm": 1.1216920614242554,
+      "learning_rate": 9.599119527977722e-05,
+      "loss": 0.6782001852989197,
+      "step": 2756
+    },
+    {
+      "epoch": 1.1898597626752967,
+      "grad_norm": 1.024833083152771,
+      "learning_rate": 9.59815567089688e-05,
+      "loss": 0.6387497782707214,
+      "step": 2758
+    },
+    {
+      "epoch": 1.190722761596548,
+      "grad_norm": 1.3539615869522095,
+      "learning_rate": 9.597190704991527e-05,
+      "loss": 0.6974280476570129,
+      "step": 2760
+    },
+    {
+      "epoch": 1.1915857605177993,
+      "grad_norm": 1.002167820930481,
+      "learning_rate": 9.59622463049436e-05,
+      "loss": 0.6039770245552063,
+      "step": 2762
+    },
+    {
+      "epoch": 1.1924487594390507,
+      "grad_norm": 1.150636076927185,
+      "learning_rate": 9.595257447638343e-05,
+      "loss": 0.6543651819229126,
+      "step": 2764
+    },
+    {
+      "epoch": 1.1933117583603021,
+      "grad_norm": 1.413001537322998,
+      "learning_rate": 9.594289156656709e-05,
+      "loss": 0.7388567328453064,
+      "step": 2766
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 1.2041958570480347,
+      "learning_rate": 9.593319757782955e-05,
+      "loss": 0.7128161191940308,
+      "step": 2768
+    },
+    {
+      "epoch": 1.1950377562028047,
+      "grad_norm": 1.2151374816894531,
+      "learning_rate": 9.592349251250847e-05,
+      "loss": 0.6593868136405945,
+      "step": 2770
+    },
+    {
+      "epoch": 1.1959007551240561,
+      "grad_norm": 1.6225353479385376,
+      "learning_rate": 9.59137763729442e-05,
+      "loss": 0.6128022074699402,
+      "step": 2772
+    },
+    {
+      "epoch": 1.1967637540453073,
+      "grad_norm": 1.1904171705245972,
+      "learning_rate": 9.590404916147976e-05,
+      "loss": 0.6537567377090454,
+      "step": 2774
+    },
+    {
+      "epoch": 1.1976267529665587,
+      "grad_norm": 1.159506916999817,
+      "learning_rate": 9.589431088046079e-05,
+      "loss": 0.6325924396514893,
+      "step": 2776
+    },
+    {
+      "epoch": 1.1984897518878102,
+      "grad_norm": 1.1314560174942017,
+      "learning_rate": 9.588456153223565e-05,
+      "loss": 0.6438963413238525,
+      "step": 2778
+    },
+    {
+      "epoch": 1.1993527508090616,
+      "grad_norm": 1.0182478427886963,
+      "learning_rate": 9.587480111915532e-05,
+      "loss": 0.6703475117683411,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2002157497303128,
+      "grad_norm": 1.5506517887115479,
+      "learning_rate": 9.586502964357352e-05,
+      "loss": 0.7694190144538879,
+      "step": 2782
+    },
+    {
+      "epoch": 1.2010787486515642,
+      "grad_norm": 1.0769826173782349,
+      "learning_rate": 9.585524710784659e-05,
+      "loss": 0.6117724180221558,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2019417475728156,
+      "grad_norm": 1.207175850868225,
+      "learning_rate": 9.584545351433352e-05,
+      "loss": 0.6375927329063416,
+      "step": 2786
+    },
+    {
+      "epoch": 1.202804746494067,
+      "grad_norm": 1.1291100978851318,
+      "learning_rate": 9.583564886539602e-05,
+      "loss": 0.6526537537574768,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2036677454153182,
+      "grad_norm": 1.3076926469802856,
+      "learning_rate": 9.582583316339843e-05,
+      "loss": 0.7080113887786865,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2045307443365696,
+      "grad_norm": 1.166823148727417,
+      "learning_rate": 9.581600641070773e-05,
+      "loss": 0.6415606141090393,
+      "step": 2792
+    },
+    {
+      "epoch": 1.205393743257821,
+      "grad_norm": 1.1248292922973633,
+      "learning_rate": 9.580616860969364e-05,
+      "loss": 0.6538968086242676,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2062567421790722,
+      "grad_norm": 1.406731128692627,
+      "learning_rate": 9.579631976272849e-05,
+      "loss": 0.694507896900177,
+      "step": 2796
+    },
+    {
+      "epoch": 1.2071197411003236,
+      "grad_norm": 1.185895562171936,
+      "learning_rate": 9.578645987218728e-05,
+      "loss": 0.7064871788024902,
+      "step": 2798
+    },
+    {
+      "epoch": 1.207982740021575,
+      "grad_norm": 1.0860708951950073,
+      "learning_rate": 9.577658894044768e-05,
+      "loss": 0.5790154933929443,
+      "step": 2800
+    },
+    {
+      "epoch": 1.207982740021575,
+      "eval_loss": 0.7220398187637329,
+      "eval_runtime": 831.5812,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2088457389428262,
+      "grad_norm": 1.2761744260787964,
+      "learning_rate": 9.576670696989001e-05,
+      "loss": 0.6645335555076599,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2097087378640776,
+      "grad_norm": 1.3502917289733887,
+      "learning_rate": 9.575681396289728e-05,
+      "loss": 0.6320346593856812,
+      "step": 2804
+    },
+    {
+      "epoch": 1.210571736785329,
+      "grad_norm": 1.2638593912124634,
+      "learning_rate": 9.574690992185514e-05,
+      "loss": 0.6297362446784973,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2114347357065804,
+      "grad_norm": 1.0787774324417114,
+      "learning_rate": 9.573699484915189e-05,
+      "loss": 0.7569923996925354,
+      "step": 2808
+    },
+    {
+      "epoch": 1.2122977346278316,
+      "grad_norm": 1.097254991531372,
+      "learning_rate": 9.572706874717851e-05,
+      "loss": 0.6712964773178101,
+      "step": 2810
+    },
+    {
+      "epoch": 1.213160733549083,
+      "grad_norm": 1.0995726585388184,
+      "learning_rate": 9.571713161832865e-05,
+      "loss": 0.6775450706481934,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2140237324703345,
+      "grad_norm": 1.7386797666549683,
+      "learning_rate": 9.570718346499859e-05,
+      "loss": 0.6691418290138245,
+      "step": 2814
+    },
+    {
+      "epoch": 1.2148867313915859,
+      "grad_norm": 1.2281049489974976,
+      "learning_rate": 9.569722428958728e-05,
+      "loss": 0.6406403183937073,
+      "step": 2816
+    },
+    {
+      "epoch": 1.215749730312837,
+      "grad_norm": 1.1326305866241455,
+      "learning_rate": 9.568725409449635e-05,
+      "loss": 0.7008584141731262,
+      "step": 2818
+    },
+    {
+      "epoch": 1.2166127292340885,
+      "grad_norm": 1.3039381504058838,
+      "learning_rate": 9.567727288213005e-05,
+      "loss": 0.7438415288925171,
+      "step": 2820
+    },
+    {
+      "epoch": 1.2174757281553399,
+      "grad_norm": 1.219136357307434,
+      "learning_rate": 9.566728065489531e-05,
+      "loss": 0.6937777996063232,
+      "step": 2822
+    },
+    {
+      "epoch": 1.218338727076591,
+      "grad_norm": 1.321285605430603,
+      "learning_rate": 9.56572774152017e-05,
+      "loss": 0.7586468458175659,
+      "step": 2824
+    },
+    {
+      "epoch": 1.2192017259978425,
+      "grad_norm": 1.3538484573364258,
+      "learning_rate": 9.564726316546147e-05,
+      "loss": 0.6356340050697327,
+      "step": 2826
+    },
+    {
+      "epoch": 1.220064724919094,
+      "grad_norm": 1.1015816926956177,
+      "learning_rate": 9.563723790808952e-05,
+      "loss": 0.6304616928100586,
+      "step": 2828
+    },
+    {
+      "epoch": 1.220927723840345,
+      "grad_norm": 0.9658390283584595,
+      "learning_rate": 9.562720164550337e-05,
+      "loss": 0.6459048986434937,
+      "step": 2830
+    },
+    {
+      "epoch": 1.2217907227615965,
+      "grad_norm": 1.3981443643569946,
+      "learning_rate": 9.561715438012323e-05,
+      "loss": 0.7034182548522949,
+      "step": 2832
+    },
+    {
+      "epoch": 1.222653721682848,
+      "grad_norm": 1.1737093925476074,
+      "learning_rate": 9.560709611437195e-05,
+      "loss": 0.6629860997200012,
+      "step": 2834
+    },
+    {
+      "epoch": 1.2235167206040993,
+      "grad_norm": 1.2071303129196167,
+      "learning_rate": 9.559702685067506e-05,
+      "loss": 0.7166537642478943,
+      "step": 2836
+    },
+    {
+      "epoch": 1.2243797195253505,
+      "grad_norm": 1.2520930767059326,
+      "learning_rate": 9.558694659146068e-05,
+      "loss": 0.6711867451667786,
+      "step": 2838
+    },
+    {
+      "epoch": 1.225242718446602,
+      "grad_norm": 1.2304543256759644,
+      "learning_rate": 9.557685533915964e-05,
+      "loss": 0.7135099768638611,
+      "step": 2840
+    },
+    {
+      "epoch": 1.2261057173678533,
+      "grad_norm": 1.1509613990783691,
+      "learning_rate": 9.556675309620539e-05,
+      "loss": 0.6279205083847046,
+      "step": 2842
+    },
+    {
+      "epoch": 1.2269687162891048,
+      "grad_norm": 1.2024176120758057,
+      "learning_rate": 9.555663986503406e-05,
+      "loss": 0.6848570704460144,
+      "step": 2844
+    },
+    {
+      "epoch": 1.227831715210356,
+      "grad_norm": 1.360257863998413,
+      "learning_rate": 9.55465156480844e-05,
+      "loss": 0.7109206914901733,
+      "step": 2846
+    },
+    {
+      "epoch": 1.2286947141316074,
+      "grad_norm": 1.1556848287582397,
+      "learning_rate": 9.55363804477978e-05,
+      "loss": 0.5983237028121948,
+      "step": 2848
+    },
+    {
+      "epoch": 1.2295577130528588,
+      "grad_norm": 1.131439447402954,
+      "learning_rate": 9.552623426661836e-05,
+      "loss": 0.6387401223182678,
+      "step": 2850
+    },
+    {
+      "epoch": 1.23042071197411,
+      "grad_norm": 1.1854766607284546,
+      "learning_rate": 9.551607710699273e-05,
+      "loss": 0.6584593057632446,
+      "step": 2852
+    },
+    {
+      "epoch": 1.2312837108953614,
+      "grad_norm": 1.1811782121658325,
+      "learning_rate": 9.55059089713703e-05,
+      "loss": 0.6284405589103699,
+      "step": 2854
+    },
+    {
+      "epoch": 1.2321467098166128,
+      "grad_norm": 1.1038708686828613,
+      "learning_rate": 9.549572986220307e-05,
+      "loss": 0.6183417439460754,
+      "step": 2856
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 1.3645228147506714,
+      "learning_rate": 9.548553978194567e-05,
+      "loss": 0.7266764640808105,
+      "step": 2858
+    },
+    {
+      "epoch": 1.2338727076591154,
+      "grad_norm": 1.1896798610687256,
+      "learning_rate": 9.547533873305542e-05,
+      "loss": 0.6580738425254822,
+      "step": 2860
+    },
+    {
+      "epoch": 1.2347357065803668,
+      "grad_norm": 1.1351168155670166,
+      "learning_rate": 9.546512671799224e-05,
+      "loss": 0.669810950756073,
+      "step": 2862
+    },
+    {
+      "epoch": 1.2355987055016182,
+      "grad_norm": 1.1937936544418335,
+      "learning_rate": 9.545490373921869e-05,
+      "loss": 0.730329155921936,
+      "step": 2864
+    },
+    {
+      "epoch": 1.2364617044228694,
+      "grad_norm": 1.094347357749939,
+      "learning_rate": 9.544466979920002e-05,
+      "loss": 0.6867446303367615,
+      "step": 2866
+    },
+    {
+      "epoch": 1.2373247033441208,
+      "grad_norm": 1.4509600400924683,
+      "learning_rate": 9.54344249004041e-05,
+      "loss": 0.735468327999115,
+      "step": 2868
+    },
+    {
+      "epoch": 1.2381877022653722,
+      "grad_norm": 1.1195857524871826,
+      "learning_rate": 9.542416904530142e-05,
+      "loss": 0.6316167116165161,
+      "step": 2870
+    },
+    {
+      "epoch": 1.2390507011866236,
+      "grad_norm": 1.0227746963500977,
+      "learning_rate": 9.541390223636517e-05,
+      "loss": 0.6637873649597168,
+      "step": 2872
+    },
+    {
+      "epoch": 1.2399137001078748,
+      "grad_norm": 1.1871414184570312,
+      "learning_rate": 9.540362447607111e-05,
+      "loss": 0.6579877734184265,
+      "step": 2874
+    },
+    {
+      "epoch": 1.2407766990291262,
+      "grad_norm": 1.3265694379806519,
+      "learning_rate": 9.539333576689768e-05,
+      "loss": 0.7563158273696899,
+      "step": 2876
+    },
+    {
+      "epoch": 1.2416396979503777,
+      "grad_norm": 1.0074095726013184,
+      "learning_rate": 9.538303611132598e-05,
+      "loss": 0.6373291015625,
+      "step": 2878
+    },
+    {
+      "epoch": 1.2425026968716288,
+      "grad_norm": 1.421204924583435,
+      "learning_rate": 9.537272551183971e-05,
+      "loss": 0.6989303827285767,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2433656957928803,
+      "grad_norm": 1.256486415863037,
+      "learning_rate": 9.536240397092522e-05,
+      "loss": 0.7590261697769165,
+      "step": 2882
+    },
+    {
+      "epoch": 1.2442286947141317,
+      "grad_norm": 1.1513274908065796,
+      "learning_rate": 9.53520714910715e-05,
+      "loss": 0.6652005314826965,
+      "step": 2884
+    },
+    {
+      "epoch": 1.2450916936353829,
+      "grad_norm": 1.2129496335983276,
+      "learning_rate": 9.53417280747702e-05,
+      "loss": 0.6510938405990601,
+      "step": 2886
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 1.1393296718597412,
+      "learning_rate": 9.533137372451557e-05,
+      "loss": 0.6405928730964661,
+      "step": 2888
+    },
+    {
+      "epoch": 1.2468176914778857,
+      "grad_norm": 1.1467162370681763,
+      "learning_rate": 9.532100844280451e-05,
+      "loss": 0.6819307208061218,
+      "step": 2890
+    },
+    {
+      "epoch": 1.247680690399137,
+      "grad_norm": 1.1516613960266113,
+      "learning_rate": 9.531063223213659e-05,
+      "loss": 0.6333560347557068,
+      "step": 2892
+    },
+    {
+      "epoch": 1.2485436893203883,
+      "grad_norm": 1.463189959526062,
+      "learning_rate": 9.530024509501396e-05,
+      "loss": 0.7259839177131653,
+      "step": 2894
+    },
+    {
+      "epoch": 1.2494066882416397,
+      "grad_norm": 1.0450547933578491,
+      "learning_rate": 9.528984703394144e-05,
+      "loss": 0.661693811416626,
+      "step": 2896
+    },
+    {
+      "epoch": 1.250269687162891,
+      "grad_norm": 1.0525774955749512,
+      "learning_rate": 9.527943805142648e-05,
+      "loss": 0.5889303684234619,
+      "step": 2898
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "grad_norm": 1.130048394203186,
+      "learning_rate": 9.526901814997914e-05,
+      "loss": 0.6380610466003418,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "eval_loss": 0.7182289958000183,
+      "eval_runtime": 840.8156,
+      "eval_samples_per_second": 2.45,
+      "eval_steps_per_second": 2.45,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2519956850053937,
+      "grad_norm": 1.0194777250289917,
+      "learning_rate": 9.525858733211217e-05,
+      "loss": 0.6397356986999512,
+      "step": 2902
+    },
+    {
+      "epoch": 1.2528586839266451,
+      "grad_norm": 1.3902885913848877,
+      "learning_rate": 9.524814560034087e-05,
+      "loss": 0.7931577563285828,
+      "step": 2904
+    },
+    {
+      "epoch": 1.2537216828478965,
+      "grad_norm": 1.258169412612915,
+      "learning_rate": 9.523769295718325e-05,
+      "loss": 0.6997087597846985,
+      "step": 2906
+    },
+    {
+      "epoch": 1.2545846817691477,
+      "grad_norm": 1.3403711318969727,
+      "learning_rate": 9.522722940515991e-05,
+      "loss": 0.6946825385093689,
+      "step": 2908
+    },
+    {
+      "epoch": 1.2554476806903991,
+      "grad_norm": 1.1724950075149536,
+      "learning_rate": 9.521675494679409e-05,
+      "loss": 0.6309347748756409,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2563106796116505,
+      "grad_norm": 1.210274338722229,
+      "learning_rate": 9.520626958461163e-05,
+      "loss": 0.6453542113304138,
+      "step": 2912
+    },
+    {
+      "epoch": 1.2571736785329017,
+      "grad_norm": 1.233112096786499,
+      "learning_rate": 9.519577332114106e-05,
+      "loss": 0.6747867465019226,
+      "step": 2914
+    },
+    {
+      "epoch": 1.2580366774541532,
+      "grad_norm": 1.1785951852798462,
+      "learning_rate": 9.51852661589135e-05,
+      "loss": 0.6729235053062439,
+      "step": 2916
+    },
+    {
+      "epoch": 1.2588996763754046,
+      "grad_norm": 1.0781456232070923,
+      "learning_rate": 9.51747481004627e-05,
+      "loss": 0.6581047177314758,
+      "step": 2918
+    },
+    {
+      "epoch": 1.2597626752966558,
+      "grad_norm": 1.1122184991836548,
+      "learning_rate": 9.516421914832506e-05,
+      "loss": 0.6576398015022278,
+      "step": 2920
+    },
+    {
+      "epoch": 1.2606256742179072,
+      "grad_norm": 1.3378703594207764,
+      "learning_rate": 9.515367930503956e-05,
+      "loss": 0.7026275396347046,
+      "step": 2922
+    },
+    {
+      "epoch": 1.2614886731391586,
+      "grad_norm": 1.127657175064087,
+      "learning_rate": 9.514312857314785e-05,
+      "loss": 0.6111703515052795,
+      "step": 2924
+    },
+    {
+      "epoch": 1.26235167206041,
+      "grad_norm": 1.2058607339859009,
+      "learning_rate": 9.513256695519419e-05,
+      "loss": 0.6698760986328125,
+      "step": 2926
+    },
+    {
+      "epoch": 1.2632146709816614,
+      "grad_norm": 1.2206004858016968,
+      "learning_rate": 9.512199445372546e-05,
+      "loss": 0.6793603897094727,
+      "step": 2928
+    },
+    {
+      "epoch": 1.2640776699029126,
+      "grad_norm": 1.167311668395996,
+      "learning_rate": 9.511141107129119e-05,
+      "loss": 0.709406316280365,
+      "step": 2930
+    },
+    {
+      "epoch": 1.264940668824164,
+      "grad_norm": 1.3347517251968384,
+      "learning_rate": 9.51008168104435e-05,
+      "loss": 0.685570240020752,
+      "step": 2932
+    },
+    {
+      "epoch": 1.2658036677454154,
+      "grad_norm": 1.186468482017517,
+      "learning_rate": 9.509021167373714e-05,
+      "loss": 0.6775280833244324,
+      "step": 2934
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 1.3934271335601807,
+      "learning_rate": 9.507959566372953e-05,
+      "loss": 0.6587507128715515,
+      "step": 2936
+    },
+    {
+      "epoch": 1.267529665587918,
+      "grad_norm": 1.3766281604766846,
+      "learning_rate": 9.506896878298061e-05,
+      "loss": 0.692707896232605,
+      "step": 2938
+    },
+    {
+      "epoch": 1.2683926645091694,
+      "grad_norm": 1.034883975982666,
+      "learning_rate": 9.505833103405307e-05,
+      "loss": 0.6481655836105347,
+      "step": 2940
+    },
+    {
+      "epoch": 1.2692556634304206,
+      "grad_norm": 1.113044261932373,
+      "learning_rate": 9.504768241951212e-05,
+      "loss": 0.5867606997489929,
+      "step": 2942
+    },
+    {
+      "epoch": 1.270118662351672,
+      "grad_norm": 1.0769455432891846,
+      "learning_rate": 9.503702294192563e-05,
+      "loss": 0.6147398948669434,
+      "step": 2944
+    },
+    {
+      "epoch": 1.2709816612729234,
+      "grad_norm": 1.100616216659546,
+      "learning_rate": 9.502635260386408e-05,
+      "loss": 0.650299608707428,
+      "step": 2946
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 1.1982520818710327,
+      "learning_rate": 9.501567140790057e-05,
+      "loss": 0.6896426677703857,
+      "step": 2948
+    },
+    {
+      "epoch": 1.272707659115426,
+      "grad_norm": 1.276911735534668,
+      "learning_rate": 9.500497935661085e-05,
+      "loss": 0.6780133247375488,
+      "step": 2950
+    },
+    {
+      "epoch": 1.2735706580366775,
+      "grad_norm": 1.131345272064209,
+      "learning_rate": 9.499427645257323e-05,
+      "loss": 0.6906163692474365,
+      "step": 2952
+    },
+    {
+      "epoch": 1.2744336569579289,
+      "grad_norm": 1.125251293182373,
+      "learning_rate": 9.498356269836868e-05,
+      "loss": 0.6418634057044983,
+      "step": 2954
+    },
+    {
+      "epoch": 1.2752966558791803,
+      "grad_norm": 1.2508443593978882,
+      "learning_rate": 9.497283809658078e-05,
+      "loss": 0.690779983997345,
+      "step": 2956
+    },
+    {
+      "epoch": 1.2761596548004315,
+      "grad_norm": 0.9558717608451843,
+      "learning_rate": 9.49621026497957e-05,
+      "loss": 0.645685613155365,
+      "step": 2958
+    },
+    {
+      "epoch": 1.2770226537216829,
+      "grad_norm": 1.1781591176986694,
+      "learning_rate": 9.495135636060226e-05,
+      "loss": 0.6313972473144531,
+      "step": 2960
+    },
+    {
+      "epoch": 1.2778856526429343,
+      "grad_norm": 0.9713736772537231,
+      "learning_rate": 9.494059923159189e-05,
+      "loss": 0.5853794813156128,
+      "step": 2962
+    },
+    {
+      "epoch": 1.2787486515641855,
+      "grad_norm": 1.3793153762817383,
+      "learning_rate": 9.492983126535859e-05,
+      "loss": 0.6966362595558167,
+      "step": 2964
+    },
+    {
+      "epoch": 1.279611650485437,
+      "grad_norm": 1.250563383102417,
+      "learning_rate": 9.491905246449903e-05,
+      "loss": 0.7281587719917297,
+      "step": 2966
+    },
+    {
+      "epoch": 1.2804746494066883,
+      "grad_norm": 1.3268930912017822,
+      "learning_rate": 9.490826283161247e-05,
+      "loss": 0.7322974801063538,
+      "step": 2968
+    },
+    {
+      "epoch": 1.2813376483279395,
+      "grad_norm": 1.3876935243606567,
+      "learning_rate": 9.489746236930077e-05,
+      "loss": 0.7875081896781921,
+      "step": 2970
+    },
+    {
+      "epoch": 1.282200647249191,
+      "grad_norm": 1.02049720287323,
+      "learning_rate": 9.488665108016842e-05,
+      "loss": 0.601652979850769,
+      "step": 2972
+    },
+    {
+      "epoch": 1.2830636461704423,
+      "grad_norm": 1.0868756771087646,
+      "learning_rate": 9.487582896682252e-05,
+      "loss": 0.6255780458450317,
+      "step": 2974
+    },
+    {
+      "epoch": 1.2839266450916935,
+      "grad_norm": 1.00597083568573,
+      "learning_rate": 9.486499603187275e-05,
+      "loss": 0.5967418551445007,
+      "step": 2976
+    },
+    {
+      "epoch": 1.284789644012945,
+      "grad_norm": 1.3068969249725342,
+      "learning_rate": 9.485415227793145e-05,
+      "loss": 0.719573438167572,
+      "step": 2978
+    },
+    {
+      "epoch": 1.2856526429341963,
+      "grad_norm": 1.2478386163711548,
+      "learning_rate": 9.484329770761353e-05,
+      "loss": 0.665522038936615,
+      "step": 2980
+    },
+    {
+      "epoch": 1.2865156418554478,
+      "grad_norm": 1.1220918893814087,
+      "learning_rate": 9.483243232353652e-05,
+      "loss": 0.6798994541168213,
+      "step": 2982
+    },
+    {
+      "epoch": 1.287378640776699,
+      "grad_norm": 1.1369609832763672,
+      "learning_rate": 9.482155612832057e-05,
+      "loss": 0.672002911567688,
+      "step": 2984
+    },
+    {
+      "epoch": 1.2882416396979504,
+      "grad_norm": 1.2077715396881104,
+      "learning_rate": 9.48106691245884e-05,
+      "loss": 0.6755662560462952,
+      "step": 2986
+    },
+    {
+      "epoch": 1.2891046386192018,
+      "grad_norm": 1.3299074172973633,
+      "learning_rate": 9.479977131496539e-05,
+      "loss": 0.7504231333732605,
+      "step": 2988
+    },
+    {
+      "epoch": 1.2899676375404532,
+      "grad_norm": 1.08982253074646,
+      "learning_rate": 9.478886270207948e-05,
+      "loss": 0.6837405562400818,
+      "step": 2990
+    },
+    {
+      "epoch": 1.2908306364617044,
+      "grad_norm": 1.1015464067459106,
+      "learning_rate": 9.477794328856124e-05,
+      "loss": 0.6552648544311523,
+      "step": 2992
+    },
+    {
+      "epoch": 1.2916936353829558,
+      "grad_norm": 1.1793229579925537,
+      "learning_rate": 9.476701307704384e-05,
+      "loss": 0.6160483360290527,
+      "step": 2994
+    },
+    {
+      "epoch": 1.2925566343042072,
+      "grad_norm": 1.4781638383865356,
+      "learning_rate": 9.475607207016306e-05,
+      "loss": 0.7406170964241028,
+      "step": 2996
+    },
+    {
+      "epoch": 1.2934196332254584,
+      "grad_norm": 1.2391449213027954,
+      "learning_rate": 9.474512027055724e-05,
+      "loss": 0.5950883626937866,
+      "step": 2998
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "grad_norm": 1.166535496711731,
+      "learning_rate": 9.473415768086738e-05,
+      "loss": 0.6587089896202087,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "eval_loss": 0.7160727977752686,
+      "eval_runtime": 830.8186,
+      "eval_samples_per_second": 2.479,
+      "eval_steps_per_second": 2.479,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2951456310679612,
+      "grad_norm": 0.9833650588989258,
+      "learning_rate": 9.472318430373706e-05,
+      "loss": 0.5802602767944336,
+      "step": 3002
+    },
+    {
+      "epoch": 1.2960086299892124,
+      "grad_norm": 1.047341227531433,
+      "learning_rate": 9.471220014181246e-05,
+      "loss": 0.6255064606666565,
+      "step": 3004
+    },
+    {
+      "epoch": 1.2968716289104638,
+      "grad_norm": 1.2761468887329102,
+      "learning_rate": 9.470120519774237e-05,
+      "loss": 0.7038446068763733,
+      "step": 3006
+    },
+    {
+      "epoch": 1.2977346278317152,
+      "grad_norm": 1.2568607330322266,
+      "learning_rate": 9.469019947417815e-05,
+      "loss": 0.7063520550727844,
+      "step": 3008
+    },
+    {
+      "epoch": 1.2985976267529666,
+      "grad_norm": 1.2337472438812256,
+      "learning_rate": 9.467918297377378e-05,
+      "loss": 0.6931142807006836,
+      "step": 3010
+    },
+    {
+      "epoch": 1.2994606256742178,
+      "grad_norm": 1.1557739973068237,
+      "learning_rate": 9.466815569918588e-05,
+      "loss": 0.6660345196723938,
+      "step": 3012
+    },
+    {
+      "epoch": 1.3003236245954692,
+      "grad_norm": 1.2341985702514648,
+      "learning_rate": 9.465711765307357e-05,
+      "loss": 0.6522690653800964,
+      "step": 3014
+    },
+    {
+      "epoch": 1.3011866235167207,
+      "grad_norm": 1.0495452880859375,
+      "learning_rate": 9.464606883809867e-05,
+      "loss": 0.6304616332054138,
+      "step": 3016
+    },
+    {
+      "epoch": 1.302049622437972,
+      "grad_norm": 1.1920377016067505,
+      "learning_rate": 9.463500925692554e-05,
+      "loss": 0.7271291613578796,
+      "step": 3018
+    },
+    {
+      "epoch": 1.3029126213592233,
+      "grad_norm": 1.0697529315948486,
+      "learning_rate": 9.462393891222112e-05,
+      "loss": 0.5725325345993042,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3037756202804747,
+      "grad_norm": 1.0559576749801636,
+      "learning_rate": 9.461285780665502e-05,
+      "loss": 0.5725870728492737,
+      "step": 3022
+    },
+    {
+      "epoch": 1.304638619201726,
+      "grad_norm": 1.1314871311187744,
+      "learning_rate": 9.460176594289937e-05,
+      "loss": 0.6498687863349915,
+      "step": 3024
+    },
+    {
+      "epoch": 1.3055016181229773,
+      "grad_norm": 1.0620028972625732,
+      "learning_rate": 9.459066332362892e-05,
+      "loss": 0.5473160147666931,
+      "step": 3026
+    },
+    {
+      "epoch": 1.3063646170442287,
+      "grad_norm": 1.245997667312622,
+      "learning_rate": 9.457954995152103e-05,
+      "loss": 0.6976684927940369,
+      "step": 3028
+    },
+    {
+      "epoch": 1.30722761596548,
+      "grad_norm": 1.0973089933395386,
+      "learning_rate": 9.456842582925565e-05,
+      "loss": 0.6469701528549194,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3080906148867313,
+      "grad_norm": 1.1558068990707397,
+      "learning_rate": 9.455729095951527e-05,
+      "loss": 0.6712765097618103,
+      "step": 3032
+    },
+    {
+      "epoch": 1.3089536138079827,
+      "grad_norm": 1.0367828607559204,
+      "learning_rate": 9.454614534498506e-05,
+      "loss": 0.6504483222961426,
+      "step": 3034
+    },
+    {
+      "epoch": 1.309816612729234,
+      "grad_norm": 1.3485349416732788,
+      "learning_rate": 9.453498898835272e-05,
+      "loss": 0.7000672221183777,
+      "step": 3036
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 1.2435184717178345,
+      "learning_rate": 9.452382189230853e-05,
+      "loss": 0.6946983337402344,
+      "step": 3038
+    },
+    {
+      "epoch": 1.3115426105717367,
+      "grad_norm": 1.0728851556777954,
+      "learning_rate": 9.451264405954541e-05,
+      "loss": 0.5928027033805847,
+      "step": 3040
+    },
+    {
+      "epoch": 1.3124056094929881,
+      "grad_norm": 1.2320033311843872,
+      "learning_rate": 9.450145549275884e-05,
+      "loss": 0.6589536070823669,
+      "step": 3042
+    },
+    {
+      "epoch": 1.3132686084142395,
+      "grad_norm": 1.143538236618042,
+      "learning_rate": 9.449025619464689e-05,
+      "loss": 0.6516439914703369,
+      "step": 3044
+    },
+    {
+      "epoch": 1.314131607335491,
+      "grad_norm": 1.3419140577316284,
+      "learning_rate": 9.447904616791023e-05,
+      "loss": 0.6889938116073608,
+      "step": 3046
+    },
+    {
+      "epoch": 1.3149946062567421,
+      "grad_norm": 1.4128848314285278,
+      "learning_rate": 9.44678254152521e-05,
+      "loss": 0.7203409075737,
+      "step": 3048
+    },
+    {
+      "epoch": 1.3158576051779936,
+      "grad_norm": 1.1409581899642944,
+      "learning_rate": 9.445659393937833e-05,
+      "loss": 0.656524658203125,
+      "step": 3050
+    },
+    {
+      "epoch": 1.316720604099245,
+      "grad_norm": 1.0904145240783691,
+      "learning_rate": 9.444535174299734e-05,
+      "loss": 0.6888052821159363,
+      "step": 3052
+    },
+    {
+      "epoch": 1.3175836030204962,
+      "grad_norm": 1.1533669233322144,
+      "learning_rate": 9.443409882882014e-05,
+      "loss": 0.6149572730064392,
+      "step": 3054
+    },
+    {
+      "epoch": 1.3184466019417476,
+      "grad_norm": 1.0775257349014282,
+      "learning_rate": 9.442283519956031e-05,
+      "loss": 0.6445736885070801,
+      "step": 3056
+    },
+    {
+      "epoch": 1.319309600862999,
+      "grad_norm": 1.1126797199249268,
+      "learning_rate": 9.441156085793403e-05,
+      "loss": 0.6723806858062744,
+      "step": 3058
+    },
+    {
+      "epoch": 1.3201725997842502,
+      "grad_norm": 1.1584032773971558,
+      "learning_rate": 9.440027580666008e-05,
+      "loss": 0.5901219248771667,
+      "step": 3060
+    },
+    {
+      "epoch": 1.3210355987055016,
+      "grad_norm": 1.0951672792434692,
+      "learning_rate": 9.438898004845975e-05,
+      "loss": 0.6816843748092651,
+      "step": 3062
+    },
+    {
+      "epoch": 1.321898597626753,
+      "grad_norm": 1.2165061235427856,
+      "learning_rate": 9.4377673586057e-05,
+      "loss": 0.6883957386016846,
+      "step": 3064
+    },
+    {
+      "epoch": 1.3227615965480042,
+      "grad_norm": 1.1537514925003052,
+      "learning_rate": 9.436635642217832e-05,
+      "loss": 0.6828808188438416,
+      "step": 3066
+    },
+    {
+      "epoch": 1.3236245954692556,
+      "grad_norm": 1.2186168432235718,
+      "learning_rate": 9.435502855955281e-05,
+      "loss": 0.6086717844009399,
+      "step": 3068
+    },
+    {
+      "epoch": 1.324487594390507,
+      "grad_norm": 1.0345975160598755,
+      "learning_rate": 9.43436900009121e-05,
+      "loss": 0.6000884771347046,
+      "step": 3070
+    },
+    {
+      "epoch": 1.3253505933117584,
+      "grad_norm": 1.1357595920562744,
+      "learning_rate": 9.433234074899044e-05,
+      "loss": 0.6812195181846619,
+      "step": 3072
+    },
+    {
+      "epoch": 1.3262135922330098,
+      "grad_norm": 1.1914640665054321,
+      "learning_rate": 9.432098080652467e-05,
+      "loss": 0.6666595935821533,
+      "step": 3074
+    },
+    {
+      "epoch": 1.327076591154261,
+      "grad_norm": 1.0881390571594238,
+      "learning_rate": 9.430961017625418e-05,
+      "loss": 0.6455721259117126,
+      "step": 3076
+    },
+    {
+      "epoch": 1.3279395900755124,
+      "grad_norm": 1.1859697103500366,
+      "learning_rate": 9.429822886092093e-05,
+      "loss": 0.5861079692840576,
+      "step": 3078
+    },
+    {
+      "epoch": 1.3288025889967638,
+      "grad_norm": 1.083703637123108,
+      "learning_rate": 9.42868368632695e-05,
+      "loss": 0.6173920035362244,
+      "step": 3080
+    },
+    {
+      "epoch": 1.329665587918015,
+      "grad_norm": 1.2820687294006348,
+      "learning_rate": 9.427543418604701e-05,
+      "loss": 0.817629873752594,
+      "step": 3082
+    },
+    {
+      "epoch": 1.3305285868392664,
+      "grad_norm": 1.1267426013946533,
+      "learning_rate": 9.426402083200315e-05,
+      "loss": 0.7454650402069092,
+      "step": 3084
+    },
+    {
+      "epoch": 1.3313915857605179,
+      "grad_norm": 1.0095305442810059,
+      "learning_rate": 9.425259680389022e-05,
+      "loss": 0.625328779220581,
+      "step": 3086
+    },
+    {
+      "epoch": 1.332254584681769,
+      "grad_norm": 1.183151364326477,
+      "learning_rate": 9.424116210446305e-05,
+      "loss": 0.5977230072021484,
+      "step": 3088
+    },
+    {
+      "epoch": 1.3331175836030205,
+      "grad_norm": 1.1312501430511475,
+      "learning_rate": 9.42297167364791e-05,
+      "loss": 0.6251013278961182,
+      "step": 3090
+    },
+    {
+      "epoch": 1.3339805825242719,
+      "grad_norm": 1.3226121664047241,
+      "learning_rate": 9.421826070269832e-05,
+      "loss": 0.7469011545181274,
+      "step": 3092
+    },
+    {
+      "epoch": 1.334843581445523,
+      "grad_norm": 1.1389979124069214,
+      "learning_rate": 9.420679400588334e-05,
+      "loss": 0.7012571096420288,
+      "step": 3094
+    },
+    {
+      "epoch": 1.3357065803667745,
+      "grad_norm": 1.2065027952194214,
+      "learning_rate": 9.419531664879925e-05,
+      "loss": 0.7112689018249512,
+      "step": 3096
+    },
+    {
+      "epoch": 1.3365695792880259,
+      "grad_norm": 1.2532200813293457,
+      "learning_rate": 9.418382863421379e-05,
+      "loss": 0.6219019293785095,
+      "step": 3098
+    },
+    {
+      "epoch": 1.3374325782092773,
+      "grad_norm": 1.1859256029129028,
+      "learning_rate": 9.417232996489724e-05,
+      "loss": 0.6905092597007751,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3374325782092773,
+      "eval_loss": 0.7127581834793091,
+      "eval_runtime": 821.1346,
+      "eval_samples_per_second": 2.509,
+      "eval_steps_per_second": 2.509,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3382955771305287,
+      "grad_norm": 1.172967791557312,
+      "learning_rate": 9.416082064362244e-05,
+      "loss": 0.6473057866096497,
+      "step": 3102
+    },
+    {
+      "epoch": 1.33915857605178,
+      "grad_norm": 1.2966243028640747,
+      "learning_rate": 9.414930067316482e-05,
+      "loss": 0.7377036213874817,
+      "step": 3104
+    },
+    {
+      "epoch": 1.3400215749730313,
+      "grad_norm": 1.0844253301620483,
+      "learning_rate": 9.413777005630237e-05,
+      "loss": 0.6842042207717896,
+      "step": 3106
+    },
+    {
+      "epoch": 1.3408845738942827,
+      "grad_norm": 0.930098831653595,
+      "learning_rate": 9.412622879581565e-05,
+      "loss": 0.662327229976654,
+      "step": 3108
+    },
+    {
+      "epoch": 1.341747572815534,
+      "grad_norm": 1.1382031440734863,
+      "learning_rate": 9.411467689448777e-05,
+      "loss": 0.6979284882545471,
+      "step": 3110
+    },
+    {
+      "epoch": 1.3426105717367853,
+      "grad_norm": 1.1555129289627075,
+      "learning_rate": 9.410311435510444e-05,
+      "loss": 0.7032427191734314,
+      "step": 3112
+    },
+    {
+      "epoch": 1.3434735706580367,
+      "grad_norm": 1.147678017616272,
+      "learning_rate": 9.409154118045389e-05,
+      "loss": 0.6282536387443542,
+      "step": 3114
+    },
+    {
+      "epoch": 1.344336569579288,
+      "grad_norm": 1.0925829410552979,
+      "learning_rate": 9.407995737332696e-05,
+      "loss": 0.6572934985160828,
+      "step": 3116
+    },
+    {
+      "epoch": 1.3451995685005393,
+      "grad_norm": 1.3632638454437256,
+      "learning_rate": 9.4068362936517e-05,
+      "loss": 0.7233583927154541,
+      "step": 3118
+    },
+    {
+      "epoch": 1.3460625674217908,
+      "grad_norm": 1.3994121551513672,
+      "learning_rate": 9.405675787282e-05,
+      "loss": 0.6465377807617188,
+      "step": 3120
+    },
+    {
+      "epoch": 1.346925566343042,
+      "grad_norm": 1.1862388849258423,
+      "learning_rate": 9.404514218503444e-05,
+      "loss": 0.6348391771316528,
+      "step": 3122
+    },
+    {
+      "epoch": 1.3477885652642934,
+      "grad_norm": 1.1724227666854858,
+      "learning_rate": 9.403351587596141e-05,
+      "loss": 0.6559093594551086,
+      "step": 3124
+    },
+    {
+      "epoch": 1.3486515641855448,
+      "grad_norm": 1.175567865371704,
+      "learning_rate": 9.402187894840452e-05,
+      "loss": 0.6693678498268127,
+      "step": 3126
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 1.1273375749588013,
+      "learning_rate": 9.401023140516999e-05,
+      "loss": 0.6766137480735779,
+      "step": 3128
+    },
+    {
+      "epoch": 1.3503775620280476,
+      "grad_norm": 1.2515393495559692,
+      "learning_rate": 9.399857324906652e-05,
+      "loss": 0.7133797407150269,
+      "step": 3130
+    },
+    {
+      "epoch": 1.3512405609492988,
+      "grad_norm": 1.1575038433074951,
+      "learning_rate": 9.398690448290547e-05,
+      "loss": 0.6572166681289673,
+      "step": 3132
+    },
+    {
+      "epoch": 1.3521035598705502,
+      "grad_norm": 1.2350742816925049,
+      "learning_rate": 9.397522510950072e-05,
+      "loss": 0.6389561891555786,
+      "step": 3134
+    },
+    {
+      "epoch": 1.3529665587918016,
+      "grad_norm": 0.9648056626319885,
+      "learning_rate": 9.396353513166867e-05,
+      "loss": 0.5810164213180542,
+      "step": 3136
+    },
+    {
+      "epoch": 1.3538295577130528,
+      "grad_norm": 1.147218108177185,
+      "learning_rate": 9.39518345522283e-05,
+      "loss": 0.7157397866249084,
+      "step": 3138
+    },
+    {
+      "epoch": 1.3546925566343042,
+      "grad_norm": 1.1948896646499634,
+      "learning_rate": 9.394012337400117e-05,
+      "loss": 0.6749072670936584,
+      "step": 3140
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 1.0945216417312622,
+      "learning_rate": 9.392840159981136e-05,
+      "loss": 0.63704514503479,
+      "step": 3142
+    },
+    {
+      "epoch": 1.3564185544768068,
+      "grad_norm": 1.0654480457305908,
+      "learning_rate": 9.391666923248555e-05,
+      "loss": 0.6804014444351196,
+      "step": 3144
+    },
+    {
+      "epoch": 1.3572815533980582,
+      "grad_norm": 1.3977962732315063,
+      "learning_rate": 9.390492627485292e-05,
+      "loss": 0.7055573463439941,
+      "step": 3146
+    },
+    {
+      "epoch": 1.3581445523193096,
+      "grad_norm": 1.1607128381729126,
+      "learning_rate": 9.389317272974526e-05,
+      "loss": 0.7383730411529541,
+      "step": 3148
+    },
+    {
+      "epoch": 1.3590075512405608,
+      "grad_norm": 1.192213773727417,
+      "learning_rate": 9.388140859999686e-05,
+      "loss": 0.7339736223220825,
+      "step": 3150
+    },
+    {
+      "epoch": 1.3598705501618122,
+      "grad_norm": 1.1597269773483276,
+      "learning_rate": 9.386963388844459e-05,
+      "loss": 0.6564369797706604,
+      "step": 3152
+    },
+    {
+      "epoch": 1.3607335490830637,
+      "grad_norm": 1.2374637126922607,
+      "learning_rate": 9.385784859792788e-05,
+      "loss": 0.7393131256103516,
+      "step": 3154
+    },
+    {
+      "epoch": 1.361596548004315,
+      "grad_norm": 1.1975921392440796,
+      "learning_rate": 9.384605273128868e-05,
+      "loss": 0.6794763207435608,
+      "step": 3156
+    },
+    {
+      "epoch": 1.3624595469255665,
+      "grad_norm": 1.158282995223999,
+      "learning_rate": 9.383424629137154e-05,
+      "loss": 0.5906237363815308,
+      "step": 3158
+    },
+    {
+      "epoch": 1.3633225458468177,
+      "grad_norm": 1.0605288743972778,
+      "learning_rate": 9.38224292810235e-05,
+      "loss": 0.6813581585884094,
+      "step": 3160
+    },
+    {
+      "epoch": 1.364185544768069,
+      "grad_norm": 0.9797501564025879,
+      "learning_rate": 9.38106017030942e-05,
+      "loss": 0.639259934425354,
+      "step": 3162
+    },
+    {
+      "epoch": 1.3650485436893205,
+      "grad_norm": 1.1023533344268799,
+      "learning_rate": 9.379876356043578e-05,
+      "loss": 0.667378306388855,
+      "step": 3164
+    },
+    {
+      "epoch": 1.3659115426105717,
+      "grad_norm": 1.1105194091796875,
+      "learning_rate": 9.378691485590299e-05,
+      "loss": 0.671601414680481,
+      "step": 3166
+    },
+    {
+      "epoch": 1.366774541531823,
+      "grad_norm": 0.9441736936569214,
+      "learning_rate": 9.377505559235308e-05,
+      "loss": 0.623394787311554,
+      "step": 3168
+    },
+    {
+      "epoch": 1.3676375404530745,
+      "grad_norm": 1.308560848236084,
+      "learning_rate": 9.376318577264585e-05,
+      "loss": 0.6558915972709656,
+      "step": 3170
+    },
+    {
+      "epoch": 1.3685005393743257,
+      "grad_norm": 1.065693974494934,
+      "learning_rate": 9.375130539964363e-05,
+      "loss": 0.6431660652160645,
+      "step": 3172
+    },
+    {
+      "epoch": 1.369363538295577,
+      "grad_norm": 1.027817726135254,
+      "learning_rate": 9.373941447621137e-05,
+      "loss": 0.6541599631309509,
+      "step": 3174
+    },
+    {
+      "epoch": 1.3702265372168285,
+      "grad_norm": 1.1651864051818848,
+      "learning_rate": 9.372751300521649e-05,
+      "loss": 0.6710350513458252,
+      "step": 3176
+    },
+    {
+      "epoch": 1.3710895361380797,
+      "grad_norm": 1.1066170930862427,
+      "learning_rate": 9.371560098952896e-05,
+      "loss": 0.6525241732597351,
+      "step": 3178
+    },
+    {
+      "epoch": 1.3719525350593311,
+      "grad_norm": 1.0942546129226685,
+      "learning_rate": 9.370367843202133e-05,
+      "loss": 0.6362038850784302,
+      "step": 3180
+    },
+    {
+      "epoch": 1.3728155339805825,
+      "grad_norm": 1.175432562828064,
+      "learning_rate": 9.369174533556868e-05,
+      "loss": 0.6017009615898132,
+      "step": 3182
+    },
+    {
+      "epoch": 1.373678532901834,
+      "grad_norm": 1.3605095148086548,
+      "learning_rate": 9.367980170304857e-05,
+      "loss": 0.6853877305984497,
+      "step": 3184
+    },
+    {
+      "epoch": 1.3745415318230851,
+      "grad_norm": 1.224565029144287,
+      "learning_rate": 9.366784753734121e-05,
+      "loss": 0.6434065699577332,
+      "step": 3186
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 1.1151230335235596,
+      "learning_rate": 9.365588284132927e-05,
+      "loss": 0.596630334854126,
+      "step": 3188
+    },
+    {
+      "epoch": 1.376267529665588,
+      "grad_norm": 1.2674700021743774,
+      "learning_rate": 9.364390761789799e-05,
+      "loss": 0.6716568470001221,
+      "step": 3190
+    },
+    {
+      "epoch": 1.3771305285868394,
+      "grad_norm": 1.3221330642700195,
+      "learning_rate": 9.363192186993512e-05,
+      "loss": 0.6230685710906982,
+      "step": 3192
+    },
+    {
+      "epoch": 1.3779935275080906,
+      "grad_norm": 1.2978439331054688,
+      "learning_rate": 9.3619925600331e-05,
+      "loss": 0.6495190858840942,
+      "step": 3194
+    },
+    {
+      "epoch": 1.378856526429342,
+      "grad_norm": 1.1542012691497803,
+      "learning_rate": 9.360791881197846e-05,
+      "loss": 0.6534912586212158,
+      "step": 3196
+    },
+    {
+      "epoch": 1.3797195253505934,
+      "grad_norm": 1.1294280290603638,
+      "learning_rate": 9.359590150777288e-05,
+      "loss": 0.5706549286842346,
+      "step": 3198
+    },
+    {
+      "epoch": 1.3805825242718446,
+      "grad_norm": 1.0622307062149048,
+      "learning_rate": 9.358387369061218e-05,
+      "loss": 0.6193811893463135,
+      "step": 3200
+    },
+    {
+      "epoch": 1.3805825242718446,
+      "eval_loss": 0.7103806734085083,
+      "eval_runtime": 831.4047,
+      "eval_samples_per_second": 2.478,
+      "eval_steps_per_second": 2.478,
+      "step": 3200
+    },
+    {
+      "epoch": 1.381445523193096,
+      "grad_norm": 1.1722023487091064,
+      "learning_rate": 9.357183536339683e-05,
+      "loss": 0.6021195650100708,
+      "step": 3202
+    },
+    {
+      "epoch": 1.3823085221143474,
+      "grad_norm": 1.2733373641967773,
+      "learning_rate": 9.355978652902978e-05,
+      "loss": 0.7195428013801575,
+      "step": 3204
+    },
+    {
+      "epoch": 1.3831715210355986,
+      "grad_norm": 1.019609808921814,
+      "learning_rate": 9.354772719041658e-05,
+      "loss": 0.6314851641654968,
+      "step": 3206
+    },
+    {
+      "epoch": 1.38403451995685,
+      "grad_norm": 1.36198890209198,
+      "learning_rate": 9.35356573504653e-05,
+      "loss": 0.722943902015686,
+      "step": 3208
+    },
+    {
+      "epoch": 1.3848975188781014,
+      "grad_norm": 1.1291067600250244,
+      "learning_rate": 9.35235770120865e-05,
+      "loss": 0.7058439254760742,
+      "step": 3210
+    },
+    {
+      "epoch": 1.3857605177993528,
+      "grad_norm": 1.2160366773605347,
+      "learning_rate": 9.351148617819332e-05,
+      "loss": 0.6186655759811401,
+      "step": 3212
+    },
+    {
+      "epoch": 1.386623516720604,
+      "grad_norm": 1.1663079261779785,
+      "learning_rate": 9.349938485170139e-05,
+      "loss": 0.5912084579467773,
+      "step": 3214
+    },
+    {
+      "epoch": 1.3874865156418554,
+      "grad_norm": 1.2145130634307861,
+      "learning_rate": 9.34872730355289e-05,
+      "loss": 0.6494051218032837,
+      "step": 3216
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 1.018477439880371,
+      "learning_rate": 9.347515073259657e-05,
+      "loss": 0.6258159279823303,
+      "step": 3218
+    },
+    {
+      "epoch": 1.3892125134843583,
+      "grad_norm": 1.2019606828689575,
+      "learning_rate": 9.346301794582761e-05,
+      "loss": 0.7076614499092102,
+      "step": 3220
+    },
+    {
+      "epoch": 1.3900755124056094,
+      "grad_norm": 1.072648048400879,
+      "learning_rate": 9.345087467814782e-05,
+      "loss": 0.5958849191665649,
+      "step": 3222
+    },
+    {
+      "epoch": 1.3909385113268609,
+      "grad_norm": 1.087908387184143,
+      "learning_rate": 9.343872093248546e-05,
+      "loss": 0.5659268498420715,
+      "step": 3224
+    },
+    {
+      "epoch": 1.3918015102481123,
+      "grad_norm": 1.2628878355026245,
+      "learning_rate": 9.342655671177138e-05,
+      "loss": 0.6369889974594116,
+      "step": 3226
+    },
+    {
+      "epoch": 1.3926645091693635,
+      "grad_norm": 1.1580097675323486,
+      "learning_rate": 9.341438201893891e-05,
+      "loss": 0.64048171043396,
+      "step": 3228
+    },
+    {
+      "epoch": 1.3935275080906149,
+      "grad_norm": 1.2230379581451416,
+      "learning_rate": 9.340219685692395e-05,
+      "loss": 0.7039018273353577,
+      "step": 3230
+    },
+    {
+      "epoch": 1.3943905070118663,
+      "grad_norm": 1.1932404041290283,
+      "learning_rate": 9.339000122866486e-05,
+      "loss": 0.6461225152015686,
+      "step": 3232
+    },
+    {
+      "epoch": 1.3952535059331175,
+      "grad_norm": 1.4337340593338013,
+      "learning_rate": 9.337779513710257e-05,
+      "loss": 0.6538844108581543,
+      "step": 3234
+    },
+    {
+      "epoch": 1.396116504854369,
+      "grad_norm": 1.2300320863723755,
+      "learning_rate": 9.336557858518055e-05,
+      "loss": 0.6788939237594604,
+      "step": 3236
+    },
+    {
+      "epoch": 1.3969795037756203,
+      "grad_norm": 1.1031018495559692,
+      "learning_rate": 9.335335157584473e-05,
+      "loss": 0.6180470585823059,
+      "step": 3238
+    },
+    {
+      "epoch": 1.3978425026968715,
+      "grad_norm": 1.1876481771469116,
+      "learning_rate": 9.334111411204363e-05,
+      "loss": 0.7097930908203125,
+      "step": 3240
+    },
+    {
+      "epoch": 1.398705501618123,
+      "grad_norm": 1.1510672569274902,
+      "learning_rate": 9.332886619672823e-05,
+      "loss": 0.6521070003509521,
+      "step": 3242
+    },
+    {
+      "epoch": 1.3995685005393743,
+      "grad_norm": 1.2715339660644531,
+      "learning_rate": 9.331660783285208e-05,
+      "loss": 0.6738741397857666,
+      "step": 3244
+    },
+    {
+      "epoch": 1.4004314994606257,
+      "grad_norm": 1.1528351306915283,
+      "learning_rate": 9.330433902337124e-05,
+      "loss": 0.6261277794837952,
+      "step": 3246
+    },
+    {
+      "epoch": 1.4012944983818771,
+      "grad_norm": 1.1715319156646729,
+      "learning_rate": 9.329205977124424e-05,
+      "loss": 0.7071060538291931,
+      "step": 3248
+    },
+    {
+      "epoch": 1.4021574973031283,
+      "grad_norm": 1.1527135372161865,
+      "learning_rate": 9.32797700794322e-05,
+      "loss": 0.6339676380157471,
+      "step": 3250
+    },
+    {
+      "epoch": 1.4030204962243797,
+      "grad_norm": 1.1546231508255005,
+      "learning_rate": 9.32674699508987e-05,
+      "loss": 0.6572067737579346,
+      "step": 3252
+    },
+    {
+      "epoch": 1.4038834951456312,
+      "grad_norm": 1.1085712909698486,
+      "learning_rate": 9.325515938860988e-05,
+      "loss": 0.6182453036308289,
+      "step": 3254
+    },
+    {
+      "epoch": 1.4047464940668823,
+      "grad_norm": 1.2346831560134888,
+      "learning_rate": 9.324283839553437e-05,
+      "loss": 0.6558552980422974,
+      "step": 3256
+    },
+    {
+      "epoch": 1.4056094929881338,
+      "grad_norm": 1.1873912811279297,
+      "learning_rate": 9.323050697464331e-05,
+      "loss": 0.6406651735305786,
+      "step": 3258
+    },
+    {
+      "epoch": 1.4064724919093852,
+      "grad_norm": 1.5326859951019287,
+      "learning_rate": 9.32181651289104e-05,
+      "loss": 0.6869475841522217,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4073354908306364,
+      "grad_norm": 1.2141598463058472,
+      "learning_rate": 9.320581286131178e-05,
+      "loss": 0.6805770397186279,
+      "step": 3262
+    },
+    {
+      "epoch": 1.4081984897518878,
+      "grad_norm": 1.232690453529358,
+      "learning_rate": 9.319345017482615e-05,
+      "loss": 0.6193398833274841,
+      "step": 3264
+    },
+    {
+      "epoch": 1.4090614886731392,
+      "grad_norm": 1.0616317987442017,
+      "learning_rate": 9.318107707243475e-05,
+      "loss": 0.5868768095970154,
+      "step": 3266
+    },
+    {
+      "epoch": 1.4099244875943904,
+      "grad_norm": 1.091909408569336,
+      "learning_rate": 9.316869355712127e-05,
+      "loss": 0.6360217928886414,
+      "step": 3268
+    },
+    {
+      "epoch": 1.4107874865156418,
+      "grad_norm": 1.0247875452041626,
+      "learning_rate": 9.315629963187194e-05,
+      "loss": 0.6082828044891357,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4116504854368932,
+      "grad_norm": 1.2529716491699219,
+      "learning_rate": 9.314389529967551e-05,
+      "loss": 0.6340673565864563,
+      "step": 3272
+    },
+    {
+      "epoch": 1.4125134843581446,
+      "grad_norm": 0.9816965460777283,
+      "learning_rate": 9.31314805635232e-05,
+      "loss": 0.6333131790161133,
+      "step": 3274
+    },
+    {
+      "epoch": 1.413376483279396,
+      "grad_norm": 1.2140034437179565,
+      "learning_rate": 9.311905542640882e-05,
+      "loss": 0.7256702780723572,
+      "step": 3276
+    },
+    {
+      "epoch": 1.4142394822006472,
+      "grad_norm": 1.2171046733856201,
+      "learning_rate": 9.310661989132859e-05,
+      "loss": 0.7062730193138123,
+      "step": 3278
+    },
+    {
+      "epoch": 1.4151024811218986,
+      "grad_norm": 1.0224506855010986,
+      "learning_rate": 9.309417396128131e-05,
+      "loss": 0.6192088723182678,
+      "step": 3280
+    },
+    {
+      "epoch": 1.41596548004315,
+      "grad_norm": 1.1904140710830688,
+      "learning_rate": 9.308171763926823e-05,
+      "loss": 0.6522074341773987,
+      "step": 3282
+    },
+    {
+      "epoch": 1.4168284789644012,
+      "grad_norm": 1.1734132766723633,
+      "learning_rate": 9.306925092829317e-05,
+      "loss": 0.6472417712211609,
+      "step": 3284
+    },
+    {
+      "epoch": 1.4176914778856526,
+      "grad_norm": 1.2349748611450195,
+      "learning_rate": 9.30567738313624e-05,
+      "loss": 0.6990411877632141,
+      "step": 3286
+    },
+    {
+      "epoch": 1.418554476806904,
+      "grad_norm": 1.1732749938964844,
+      "learning_rate": 9.304428635148472e-05,
+      "loss": 0.6181937456130981,
+      "step": 3288
+    },
+    {
+      "epoch": 1.4194174757281552,
+      "grad_norm": 1.3659436702728271,
+      "learning_rate": 9.303178849167142e-05,
+      "loss": 0.7471824288368225,
+      "step": 3290
+    },
+    {
+      "epoch": 1.4202804746494067,
+      "grad_norm": 1.0865631103515625,
+      "learning_rate": 9.301928025493634e-05,
+      "loss": 0.5541732907295227,
+      "step": 3292
+    },
+    {
+      "epoch": 1.421143473570658,
+      "grad_norm": 1.2114793062210083,
+      "learning_rate": 9.300676164429574e-05,
+      "loss": 0.6934989094734192,
+      "step": 3294
+    },
+    {
+      "epoch": 1.4220064724919093,
+      "grad_norm": 1.1783509254455566,
+      "learning_rate": 9.299423266276846e-05,
+      "loss": 0.6887713670730591,
+      "step": 3296
+    },
+    {
+      "epoch": 1.4228694714131607,
+      "grad_norm": 1.3001595735549927,
+      "learning_rate": 9.298169331337577e-05,
+      "loss": 0.6956127882003784,
+      "step": 3298
+    },
+    {
+      "epoch": 1.423732470334412,
+      "grad_norm": 1.1657462120056152,
+      "learning_rate": 9.296914359914151e-05,
+      "loss": 0.690113365650177,
+      "step": 3300
+    },
+    {
+      "epoch": 1.423732470334412,
+      "eval_loss": 0.7086145281791687,
+      "eval_runtime": 838.887,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 3300
+    },
+    {
+      "epoch": 1.4245954692556635,
+      "grad_norm": 1.1632269620895386,
+      "learning_rate": 9.295658352309198e-05,
+      "loss": 0.6801355481147766,
+      "step": 3302
+    },
+    {
+      "epoch": 1.425458468176915,
+      "grad_norm": 1.3355532884597778,
+      "learning_rate": 9.2944013088256e-05,
+      "loss": 0.6801818013191223,
+      "step": 3304
+    },
+    {
+      "epoch": 1.426321467098166,
+      "grad_norm": 1.0911835432052612,
+      "learning_rate": 9.293143229766484e-05,
+      "loss": 0.633465051651001,
+      "step": 3306
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 1.1873823404312134,
+      "learning_rate": 9.29188411543523e-05,
+      "loss": 0.6677361130714417,
+      "step": 3308
+    },
+    {
+      "epoch": 1.428047464940669,
+      "grad_norm": 1.2115873098373413,
+      "learning_rate": 9.290623966135472e-05,
+      "loss": 0.6741982102394104,
+      "step": 3310
+    },
+    {
+      "epoch": 1.42891046386192,
+      "grad_norm": 1.2389469146728516,
+      "learning_rate": 9.289362782171084e-05,
+      "loss": 0.7169438600540161,
+      "step": 3312
+    },
+    {
+      "epoch": 1.4297734627831715,
+      "grad_norm": 1.2088500261306763,
+      "learning_rate": 9.288100563846201e-05,
+      "loss": 0.7112365961074829,
+      "step": 3314
+    },
+    {
+      "epoch": 1.430636461704423,
+      "grad_norm": 1.1279292106628418,
+      "learning_rate": 9.286837311465196e-05,
+      "loss": 0.6910304427146912,
+      "step": 3316
+    },
+    {
+      "epoch": 1.4314994606256741,
+      "grad_norm": 1.1812585592269897,
+      "learning_rate": 9.285573025332698e-05,
+      "loss": 0.7288362979888916,
+      "step": 3318
+    },
+    {
+      "epoch": 1.4323624595469255,
+      "grad_norm": 1.5522732734680176,
+      "learning_rate": 9.284307705753587e-05,
+      "loss": 0.6918224692344666,
+      "step": 3320
+    },
+    {
+      "epoch": 1.433225458468177,
+      "grad_norm": 1.2825061082839966,
+      "learning_rate": 9.283041353032984e-05,
+      "loss": 0.6756533980369568,
+      "step": 3322
+    },
+    {
+      "epoch": 1.4340884573894281,
+      "grad_norm": 1.0962797403335571,
+      "learning_rate": 9.281773967476269e-05,
+      "loss": 0.6005351543426514,
+      "step": 3324
+    },
+    {
+      "epoch": 1.4349514563106796,
+      "grad_norm": 1.249868631362915,
+      "learning_rate": 9.280505549389063e-05,
+      "loss": 0.6576942205429077,
+      "step": 3326
+    },
+    {
+      "epoch": 1.435814455231931,
+      "grad_norm": 1.2587590217590332,
+      "learning_rate": 9.279236099077239e-05,
+      "loss": 0.6686707735061646,
+      "step": 3328
+    },
+    {
+      "epoch": 1.4366774541531824,
+      "grad_norm": 1.0961607694625854,
+      "learning_rate": 9.277965616846922e-05,
+      "loss": 0.5858182311058044,
+      "step": 3330
+    },
+    {
+      "epoch": 1.4375404530744338,
+      "grad_norm": 1.199666142463684,
+      "learning_rate": 9.276694103004481e-05,
+      "loss": 0.6661836504936218,
+      "step": 3332
+    },
+    {
+      "epoch": 1.438403451995685,
+      "grad_norm": 1.2444990873336792,
+      "learning_rate": 9.275421557856536e-05,
+      "loss": 0.6922421455383301,
+      "step": 3334
+    },
+    {
+      "epoch": 1.4392664509169364,
+      "grad_norm": 1.3198295831680298,
+      "learning_rate": 9.274147981709957e-05,
+      "loss": 0.750586986541748,
+      "step": 3336
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 1.0782018899917603,
+      "learning_rate": 9.27287337487186e-05,
+      "loss": 0.6817078590393066,
+      "step": 3338
+    },
+    {
+      "epoch": 1.440992448759439,
+      "grad_norm": 1.1463382244110107,
+      "learning_rate": 9.271597737649611e-05,
+      "loss": 0.6163688898086548,
+      "step": 3340
+    },
+    {
+      "epoch": 1.4418554476806904,
+      "grad_norm": 1.0618164539337158,
+      "learning_rate": 9.270321070350825e-05,
+      "loss": 0.6624670028686523,
+      "step": 3342
+    },
+    {
+      "epoch": 1.4427184466019418,
+      "grad_norm": 1.1389427185058594,
+      "learning_rate": 9.269043373283362e-05,
+      "loss": 0.6802749037742615,
+      "step": 3344
+    },
+    {
+      "epoch": 1.443581445523193,
+      "grad_norm": 1.082173466682434,
+      "learning_rate": 9.267764646755335e-05,
+      "loss": 0.6428707838058472,
+      "step": 3346
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 1.1741442680358887,
+      "learning_rate": 9.266484891075103e-05,
+      "loss": 0.6853904128074646,
+      "step": 3348
+    },
+    {
+      "epoch": 1.4453074433656958,
+      "grad_norm": 1.1328803300857544,
+      "learning_rate": 9.265204106551273e-05,
+      "loss": 0.6889524459838867,
+      "step": 3350
+    },
+    {
+      "epoch": 1.446170442286947,
+      "grad_norm": 1.0607874393463135,
+      "learning_rate": 9.2639222934927e-05,
+      "loss": 0.6037911176681519,
+      "step": 3352
+    },
+    {
+      "epoch": 1.4470334412081984,
+      "grad_norm": 1.1584129333496094,
+      "learning_rate": 9.262639452208487e-05,
+      "loss": 0.6671991944313049,
+      "step": 3354
+    },
+    {
+      "epoch": 1.4478964401294498,
+      "grad_norm": 1.1881403923034668,
+      "learning_rate": 9.261355583007987e-05,
+      "loss": 0.7165482044219971,
+      "step": 3356
+    },
+    {
+      "epoch": 1.4487594390507013,
+      "grad_norm": 1.2093054056167603,
+      "learning_rate": 9.260070686200797e-05,
+      "loss": 0.7108280062675476,
+      "step": 3358
+    },
+    {
+      "epoch": 1.4496224379719527,
+      "grad_norm": 1.1105186939239502,
+      "learning_rate": 9.258784762096768e-05,
+      "loss": 0.718152642250061,
+      "step": 3360
+    },
+    {
+      "epoch": 1.4504854368932039,
+      "grad_norm": 1.028255820274353,
+      "learning_rate": 9.25749781100599e-05,
+      "loss": 0.6449500322341919,
+      "step": 3362
+    },
+    {
+      "epoch": 1.4513484358144553,
+      "grad_norm": 1.235719919204712,
+      "learning_rate": 9.25620983323881e-05,
+      "loss": 0.6895458102226257,
+      "step": 3364
+    },
+    {
+      "epoch": 1.4522114347357067,
+      "grad_norm": 1.3693572282791138,
+      "learning_rate": 9.254920829105816e-05,
+      "loss": 0.7243028879165649,
+      "step": 3366
+    },
+    {
+      "epoch": 1.4530744336569579,
+      "grad_norm": 1.1795028448104858,
+      "learning_rate": 9.253630798917843e-05,
+      "loss": 0.6177746057510376,
+      "step": 3368
+    },
+    {
+      "epoch": 1.4539374325782093,
+      "grad_norm": 1.0759676694869995,
+      "learning_rate": 9.25233974298598e-05,
+      "loss": 0.6603507399559021,
+      "step": 3370
+    },
+    {
+      "epoch": 1.4548004314994607,
+      "grad_norm": 0.9495601058006287,
+      "learning_rate": 9.251047661621556e-05,
+      "loss": 0.5989331603050232,
+      "step": 3372
+    },
+    {
+      "epoch": 1.455663430420712,
+      "grad_norm": 1.1334787607192993,
+      "learning_rate": 9.249754555136154e-05,
+      "loss": 0.5743664503097534,
+      "step": 3374
+    },
+    {
+      "epoch": 1.4565264293419633,
+      "grad_norm": 1.114450216293335,
+      "learning_rate": 9.248460423841598e-05,
+      "loss": 0.5905263423919678,
+      "step": 3376
+    },
+    {
+      "epoch": 1.4573894282632147,
+      "grad_norm": 1.2400883436203003,
+      "learning_rate": 9.247165268049963e-05,
+      "loss": 0.6751091480255127,
+      "step": 3378
+    },
+    {
+      "epoch": 1.458252427184466,
+      "grad_norm": 1.2845438718795776,
+      "learning_rate": 9.245869088073569e-05,
+      "loss": 0.6872944235801697,
+      "step": 3380
+    },
+    {
+      "epoch": 1.4591154261057173,
+      "grad_norm": 1.054388165473938,
+      "learning_rate": 9.244571884224987e-05,
+      "loss": 0.6467806100845337,
+      "step": 3382
+    },
+    {
+      "epoch": 1.4599784250269687,
+      "grad_norm": 1.0959769487380981,
+      "learning_rate": 9.243273656817029e-05,
+      "loss": 0.6486450433731079,
+      "step": 3384
+    },
+    {
+      "epoch": 1.4608414239482201,
+      "grad_norm": 1.6487104892730713,
+      "learning_rate": 9.241974406162756e-05,
+      "loss": 0.7024005651473999,
+      "step": 3386
+    },
+    {
+      "epoch": 1.4617044228694713,
+      "grad_norm": 1.3018934726715088,
+      "learning_rate": 9.240674132575481e-05,
+      "loss": 0.7081685662269592,
+      "step": 3388
+    },
+    {
+      "epoch": 1.4625674217907227,
+      "grad_norm": 0.9295057058334351,
+      "learning_rate": 9.239372836368754e-05,
+      "loss": 0.5842618346214294,
+      "step": 3390
+    },
+    {
+      "epoch": 1.4634304207119742,
+      "grad_norm": 1.0556540489196777,
+      "learning_rate": 9.238070517856378e-05,
+      "loss": 0.6355351209640503,
+      "step": 3392
+    },
+    {
+      "epoch": 1.4642934196332256,
+      "grad_norm": 1.2247962951660156,
+      "learning_rate": 9.236767177352403e-05,
+      "loss": 0.6340579390525818,
+      "step": 3394
+    },
+    {
+      "epoch": 1.4651564185544768,
+      "grad_norm": 1.1625045537948608,
+      "learning_rate": 9.235462815171123e-05,
+      "loss": 0.6405525803565979,
+      "step": 3396
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 1.009028673171997,
+      "learning_rate": 9.234157431627077e-05,
+      "loss": 0.5884681940078735,
+      "step": 3398
+    },
+    {
+      "epoch": 1.4668824163969796,
+      "grad_norm": 1.174851894378662,
+      "learning_rate": 9.232851027035056e-05,
+      "loss": 0.6532622575759888,
+      "step": 3400
+    },
+    {
+      "epoch": 1.4668824163969796,
+      "eval_loss": 0.7067069411277771,
+      "eval_runtime": 829.9411,
+      "eval_samples_per_second": 2.482,
+      "eval_steps_per_second": 2.482,
+      "step": 3400
+    },
+    {
+      "epoch": 1.4677454153182308,
+      "grad_norm": 0.9843166470527649,
+      "learning_rate": 9.231543601710091e-05,
+      "loss": 0.6505036354064941,
+      "step": 3402
+    },
+    {
+      "epoch": 1.4686084142394822,
+      "grad_norm": 1.1260782480239868,
+      "learning_rate": 9.230235155967461e-05,
+      "loss": 0.5809159874916077,
+      "step": 3404
+    },
+    {
+      "epoch": 1.4694714131607336,
+      "grad_norm": 1.3986146450042725,
+      "learning_rate": 9.228925690122696e-05,
+      "loss": 0.6431854367256165,
+      "step": 3406
+    },
+    {
+      "epoch": 1.4703344120819848,
+      "grad_norm": 1.145108938217163,
+      "learning_rate": 9.227615204491562e-05,
+      "loss": 0.6126174330711365,
+      "step": 3408
+    },
+    {
+      "epoch": 1.4711974110032362,
+      "grad_norm": 1.0973509550094604,
+      "learning_rate": 9.22630369939008e-05,
+      "loss": 0.6271735429763794,
+      "step": 3410
+    },
+    {
+      "epoch": 1.4720604099244876,
+      "grad_norm": 1.0548614263534546,
+      "learning_rate": 9.224991175134515e-05,
+      "loss": 0.5894950032234192,
+      "step": 3412
+    },
+    {
+      "epoch": 1.472923408845739,
+      "grad_norm": 1.3288220167160034,
+      "learning_rate": 9.223677632041374e-05,
+      "loss": 0.6621272563934326,
+      "step": 3414
+    },
+    {
+      "epoch": 1.4737864077669902,
+      "grad_norm": 1.0312891006469727,
+      "learning_rate": 9.222363070427411e-05,
+      "loss": 0.6911466717720032,
+      "step": 3416
+    },
+    {
+      "epoch": 1.4746494066882416,
+      "grad_norm": 1.1039631366729736,
+      "learning_rate": 9.221047490609629e-05,
+      "loss": 0.5945214033126831,
+      "step": 3418
+    },
+    {
+      "epoch": 1.475512405609493,
+      "grad_norm": 1.1555835008621216,
+      "learning_rate": 9.219730892905273e-05,
+      "loss": 0.6231427192687988,
+      "step": 3420
+    },
+    {
+      "epoch": 1.4763754045307445,
+      "grad_norm": 1.0504940748214722,
+      "learning_rate": 9.218413277631835e-05,
+      "loss": 0.6413036584854126,
+      "step": 3422
+    },
+    {
+      "epoch": 1.4772384034519956,
+      "grad_norm": 1.0855275392532349,
+      "learning_rate": 9.217094645107053e-05,
+      "loss": 0.663884699344635,
+      "step": 3424
+    },
+    {
+      "epoch": 1.478101402373247,
+      "grad_norm": 1.068493366241455,
+      "learning_rate": 9.215774995648907e-05,
+      "loss": 0.6161345839500427,
+      "step": 3426
+    },
+    {
+      "epoch": 1.4789644012944985,
+      "grad_norm": 1.0962451696395874,
+      "learning_rate": 9.214454329575626e-05,
+      "loss": 0.6693077087402344,
+      "step": 3428
+    },
+    {
+      "epoch": 1.4798274002157497,
+      "grad_norm": 1.070582389831543,
+      "learning_rate": 9.213132647205682e-05,
+      "loss": 0.6362645030021667,
+      "step": 3430
+    },
+    {
+      "epoch": 1.480690399137001,
+      "grad_norm": 1.201202154159546,
+      "learning_rate": 9.211809948857793e-05,
+      "loss": 0.7378971576690674,
+      "step": 3432
+    },
+    {
+      "epoch": 1.4815533980582525,
+      "grad_norm": 1.1988805532455444,
+      "learning_rate": 9.210486234850921e-05,
+      "loss": 0.686621367931366,
+      "step": 3434
+    },
+    {
+      "epoch": 1.4824163969795037,
+      "grad_norm": 1.2886041402816772,
+      "learning_rate": 9.209161505504274e-05,
+      "loss": 0.6873248219490051,
+      "step": 3436
+    },
+    {
+      "epoch": 1.483279395900755,
+      "grad_norm": 1.0028132200241089,
+      "learning_rate": 9.207835761137307e-05,
+      "loss": 0.624991774559021,
+      "step": 3438
+    },
+    {
+      "epoch": 1.4841423948220065,
+      "grad_norm": 1.139201045036316,
+      "learning_rate": 9.206509002069715e-05,
+      "loss": 0.6064573526382446,
+      "step": 3440
+    },
+    {
+      "epoch": 1.4850053937432577,
+      "grad_norm": 1.0371732711791992,
+      "learning_rate": 9.205181228621438e-05,
+      "loss": 0.6139655709266663,
+      "step": 3442
+    },
+    {
+      "epoch": 1.485868392664509,
+      "grad_norm": 1.1208233833312988,
+      "learning_rate": 9.203852441112665e-05,
+      "loss": 0.6176325082778931,
+      "step": 3444
+    },
+    {
+      "epoch": 1.4867313915857605,
+      "grad_norm": 1.2438154220581055,
+      "learning_rate": 9.20252263986383e-05,
+      "loss": 0.6920754313468933,
+      "step": 3446
+    },
+    {
+      "epoch": 1.487594390507012,
+      "grad_norm": 1.2117916345596313,
+      "learning_rate": 9.201191825195604e-05,
+      "loss": 0.6859762072563171,
+      "step": 3448
+    },
+    {
+      "epoch": 1.4884573894282633,
+      "grad_norm": 1.2274118661880493,
+      "learning_rate": 9.199859997428908e-05,
+      "loss": 0.7358171939849854,
+      "step": 3450
+    },
+    {
+      "epoch": 1.4893203883495145,
+      "grad_norm": 1.2810527086257935,
+      "learning_rate": 9.198527156884907e-05,
+      "loss": 0.6878830194473267,
+      "step": 3452
+    },
+    {
+      "epoch": 1.490183387270766,
+      "grad_norm": 1.2344684600830078,
+      "learning_rate": 9.197193303885008e-05,
+      "loss": 0.6995986104011536,
+      "step": 3454
+    },
+    {
+      "epoch": 1.4910463861920173,
+      "grad_norm": 1.233686923980713,
+      "learning_rate": 9.195858438750866e-05,
+      "loss": 0.6423236131668091,
+      "step": 3456
+    },
+    {
+      "epoch": 1.4919093851132685,
+      "grad_norm": 1.3485876321792603,
+      "learning_rate": 9.194522561804377e-05,
+      "loss": 0.7513291835784912,
+      "step": 3458
+    },
+    {
+      "epoch": 1.49277238403452,
+      "grad_norm": 0.9913640022277832,
+      "learning_rate": 9.19318567336768e-05,
+      "loss": 0.6667183637619019,
+      "step": 3460
+    },
+    {
+      "epoch": 1.4936353829557714,
+      "grad_norm": 1.146099328994751,
+      "learning_rate": 9.191847773763161e-05,
+      "loss": 0.643630862236023,
+      "step": 3462
+    },
+    {
+      "epoch": 1.4944983818770226,
+      "grad_norm": 1.0300167798995972,
+      "learning_rate": 9.190508863313448e-05,
+      "loss": 0.655176043510437,
+      "step": 3464
+    },
+    {
+      "epoch": 1.495361380798274,
+      "grad_norm": 1.285008430480957,
+      "learning_rate": 9.189168942341416e-05,
+      "loss": 0.6792252659797668,
+      "step": 3466
+    },
+    {
+      "epoch": 1.4962243797195254,
+      "grad_norm": 1.1455490589141846,
+      "learning_rate": 9.187828011170175e-05,
+      "loss": 0.7081018686294556,
+      "step": 3468
+    },
+    {
+      "epoch": 1.4970873786407766,
+      "grad_norm": 1.2560410499572754,
+      "learning_rate": 9.186486070123091e-05,
+      "loss": 0.6613248586654663,
+      "step": 3470
+    },
+    {
+      "epoch": 1.497950377562028,
+      "grad_norm": 1.2391352653503418,
+      "learning_rate": 9.185143119523761e-05,
+      "loss": 0.6299360990524292,
+      "step": 3472
+    },
+    {
+      "epoch": 1.4988133764832794,
+      "grad_norm": 1.134585976600647,
+      "learning_rate": 9.183799159696035e-05,
+      "loss": 0.5910941362380981,
+      "step": 3474
+    },
+    {
+      "epoch": 1.4996763754045308,
+      "grad_norm": 1.1744983196258545,
+      "learning_rate": 9.182454190964005e-05,
+      "loss": 0.7077166438102722,
+      "step": 3476
+    },
+    {
+      "epoch": 1.5005393743257822,
+      "grad_norm": 1.1142816543579102,
+      "learning_rate": 9.181108213651999e-05,
+      "loss": 0.6168769598007202,
+      "step": 3478
+    },
+    {
+      "epoch": 1.5014023732470334,
+      "grad_norm": 1.1463894844055176,
+      "learning_rate": 9.179761228084597e-05,
+      "loss": 0.6587874293327332,
+      "step": 3480
+    },
+    {
+      "epoch": 1.5022653721682848,
+      "grad_norm": 1.3945616483688354,
+      "learning_rate": 9.178413234586616e-05,
+      "loss": 0.7696266770362854,
+      "step": 3482
+    },
+    {
+      "epoch": 1.5031283710895362,
+      "grad_norm": 1.3266242742538452,
+      "learning_rate": 9.177064233483121e-05,
+      "loss": 0.7142559885978699,
+      "step": 3484
+    },
+    {
+      "epoch": 1.5039913700107874,
+      "grad_norm": 1.1857004165649414,
+      "learning_rate": 9.175714225099417e-05,
+      "loss": 0.7085118293762207,
+      "step": 3486
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 1.1070430278778076,
+      "learning_rate": 9.174363209761049e-05,
+      "loss": 0.6643039584159851,
+      "step": 3488
+    },
+    {
+      "epoch": 1.5057173678532902,
+      "grad_norm": 1.0877243280410767,
+      "learning_rate": 9.173011187793814e-05,
+      "loss": 0.6282537579536438,
+      "step": 3490
+    },
+    {
+      "epoch": 1.5065803667745414,
+      "grad_norm": 1.2648247480392456,
+      "learning_rate": 9.171658159523741e-05,
+      "loss": 0.6811800599098206,
+      "step": 3492
+    },
+    {
+      "epoch": 1.5074433656957928,
+      "grad_norm": 1.1535227298736572,
+      "learning_rate": 9.170304125277111e-05,
+      "loss": 0.6442922353744507,
+      "step": 3494
+    },
+    {
+      "epoch": 1.5083063646170443,
+      "grad_norm": 1.1175323724746704,
+      "learning_rate": 9.16894908538044e-05,
+      "loss": 0.717490553855896,
+      "step": 3496
+    },
+    {
+      "epoch": 1.5091693635382954,
+      "grad_norm": 1.1923470497131348,
+      "learning_rate": 9.167593040160491e-05,
+      "loss": 0.6537191867828369,
+      "step": 3498
+    },
+    {
+      "epoch": 1.510032362459547,
+      "grad_norm": 1.3302042484283447,
+      "learning_rate": 9.166235989944268e-05,
+      "loss": 0.7314199805259705,
+      "step": 3500
+    },
+    {
+      "epoch": 1.510032362459547,
+      "eval_loss": 0.7047909498214722,
+      "eval_runtime": 821.8811,
+      "eval_samples_per_second": 2.506,
+      "eval_steps_per_second": 2.506,
+      "step": 3500
+    },
+    {
+      "epoch": 1.5108953613807983,
+      "grad_norm": 1.1480292081832886,
+      "learning_rate": 9.164877935059017e-05,
+      "loss": 0.6363071203231812,
+      "step": 3502
+    },
+    {
+      "epoch": 1.5117583603020495,
+      "grad_norm": 1.1068446636199951,
+      "learning_rate": 9.163518875832227e-05,
+      "loss": 0.6020049452781677,
+      "step": 3504
+    },
+    {
+      "epoch": 1.512621359223301,
+      "grad_norm": 1.2131398916244507,
+      "learning_rate": 9.16215881259163e-05,
+      "loss": 0.7053841352462769,
+      "step": 3506
+    },
+    {
+      "epoch": 1.5134843581445523,
+      "grad_norm": 1.1875312328338623,
+      "learning_rate": 9.160797745665199e-05,
+      "loss": 0.6994143724441528,
+      "step": 3508
+    },
+    {
+      "epoch": 1.5143473570658037,
+      "grad_norm": 1.301967740058899,
+      "learning_rate": 9.159435675381147e-05,
+      "loss": 0.7001288533210754,
+      "step": 3510
+    },
+    {
+      "epoch": 1.5152103559870551,
+      "grad_norm": 1.1110293865203857,
+      "learning_rate": 9.158072602067933e-05,
+      "loss": 0.6847118139266968,
+      "step": 3512
+    },
+    {
+      "epoch": 1.5160733549083063,
+      "grad_norm": 1.2346487045288086,
+      "learning_rate": 9.156708526054256e-05,
+      "loss": 0.711972713470459,
+      "step": 3514
+    },
+    {
+      "epoch": 1.5169363538295577,
+      "grad_norm": 1.079206109046936,
+      "learning_rate": 9.155343447669058e-05,
+      "loss": 0.6406894326210022,
+      "step": 3516
+    },
+    {
+      "epoch": 1.5177993527508091,
+      "grad_norm": 1.0806539058685303,
+      "learning_rate": 9.153977367241519e-05,
+      "loss": 0.6568963527679443,
+      "step": 3518
+    },
+    {
+      "epoch": 1.5186623516720603,
+      "grad_norm": 1.244635820388794,
+      "learning_rate": 9.152610285101064e-05,
+      "loss": 0.6841240525245667,
+      "step": 3520
+    },
+    {
+      "epoch": 1.5195253505933117,
+      "grad_norm": 1.1596359014511108,
+      "learning_rate": 9.151242201577359e-05,
+      "loss": 0.619236946105957,
+      "step": 3522
+    },
+    {
+      "epoch": 1.5203883495145631,
+      "grad_norm": 1.2404531240463257,
+      "learning_rate": 9.14987311700031e-05,
+      "loss": 0.6059816479682922,
+      "step": 3524
+    },
+    {
+      "epoch": 1.5212513484358143,
+      "grad_norm": 1.0919517278671265,
+      "learning_rate": 9.148503031700067e-05,
+      "loss": 0.606432318687439,
+      "step": 3526
+    },
+    {
+      "epoch": 1.522114347357066,
+      "grad_norm": 1.113590121269226,
+      "learning_rate": 9.147131946007019e-05,
+      "loss": 0.6507619023323059,
+      "step": 3528
+    },
+    {
+      "epoch": 1.5229773462783172,
+      "grad_norm": 1.309885025024414,
+      "learning_rate": 9.145759860251799e-05,
+      "loss": 0.7071492075920105,
+      "step": 3530
+    },
+    {
+      "epoch": 1.5238403451995683,
+      "grad_norm": 1.1807539463043213,
+      "learning_rate": 9.144386774765278e-05,
+      "loss": 0.6297637224197388,
+      "step": 3532
+    },
+    {
+      "epoch": 1.52470334412082,
+      "grad_norm": 1.1587409973144531,
+      "learning_rate": 9.143012689878568e-05,
+      "loss": 0.6588859558105469,
+      "step": 3534
+    },
+    {
+      "epoch": 1.5255663430420712,
+      "grad_norm": 1.2401988506317139,
+      "learning_rate": 9.141637605923026e-05,
+      "loss": 0.6460821628570557,
+      "step": 3536
+    },
+    {
+      "epoch": 1.5264293419633226,
+      "grad_norm": 1.0708749294281006,
+      "learning_rate": 9.140261523230244e-05,
+      "loss": 0.6757200956344604,
+      "step": 3538
+    },
+    {
+      "epoch": 1.527292340884574,
+      "grad_norm": 1.2433289289474487,
+      "learning_rate": 9.138884442132061e-05,
+      "loss": 0.63701993227005,
+      "step": 3540
+    },
+    {
+      "epoch": 1.5281553398058252,
+      "grad_norm": 1.376205563545227,
+      "learning_rate": 9.137506362960553e-05,
+      "loss": 0.6708084344863892,
+      "step": 3542
+    },
+    {
+      "epoch": 1.5290183387270766,
+      "grad_norm": 1.1187406778335571,
+      "learning_rate": 9.136127286048038e-05,
+      "loss": 0.6108639240264893,
+      "step": 3544
+    },
+    {
+      "epoch": 1.529881337648328,
+      "grad_norm": 1.0509378910064697,
+      "learning_rate": 9.134747211727074e-05,
+      "loss": 0.58659827709198,
+      "step": 3546
+    },
+    {
+      "epoch": 1.5307443365695792,
+      "grad_norm": 1.1126946210861206,
+      "learning_rate": 9.133366140330458e-05,
+      "loss": 0.6567258238792419,
+      "step": 3548
+    },
+    {
+      "epoch": 1.5316073354908306,
+      "grad_norm": 1.235511302947998,
+      "learning_rate": 9.131984072191232e-05,
+      "loss": 0.7188405990600586,
+      "step": 3550
+    },
+    {
+      "epoch": 1.532470334412082,
+      "grad_norm": 1.1522574424743652,
+      "learning_rate": 9.130601007642674e-05,
+      "loss": 0.7415963411331177,
+      "step": 3552
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 1.1485437154769897,
+      "learning_rate": 9.129216947018302e-05,
+      "loss": 0.6286011934280396,
+      "step": 3554
+    },
+    {
+      "epoch": 1.5341963322545846,
+      "grad_norm": 1.1583795547485352,
+      "learning_rate": 9.127831890651879e-05,
+      "loss": 0.6919670701026917,
+      "step": 3556
+    },
+    {
+      "epoch": 1.535059331175836,
+      "grad_norm": 1.1450461149215698,
+      "learning_rate": 9.126445838877404e-05,
+      "loss": 0.6787450313568115,
+      "step": 3558
+    },
+    {
+      "epoch": 1.5359223300970872,
+      "grad_norm": 1.0954205989837646,
+      "learning_rate": 9.125058792029118e-05,
+      "loss": 0.6718940138816833,
+      "step": 3560
+    },
+    {
+      "epoch": 1.5367853290183389,
+      "grad_norm": 1.2366043329238892,
+      "learning_rate": 9.1236707504415e-05,
+      "loss": 0.7359552383422852,
+      "step": 3562
+    },
+    {
+      "epoch": 1.53764832793959,
+      "grad_norm": 1.1417096853256226,
+      "learning_rate": 9.122281714449268e-05,
+      "loss": 0.654550313949585,
+      "step": 3564
+    },
+    {
+      "epoch": 1.5385113268608415,
+      "grad_norm": 1.006664752960205,
+      "learning_rate": 9.120891684387386e-05,
+      "loss": 0.7129375338554382,
+      "step": 3566
+    },
+    {
+      "epoch": 1.5393743257820929,
+      "grad_norm": 1.2887771129608154,
+      "learning_rate": 9.11950066059105e-05,
+      "loss": 0.728861391544342,
+      "step": 3568
+    },
+    {
+      "epoch": 1.540237324703344,
+      "grad_norm": 1.194985270500183,
+      "learning_rate": 9.118108643395702e-05,
+      "loss": 0.6207047700881958,
+      "step": 3570
+    },
+    {
+      "epoch": 1.5411003236245955,
+      "grad_norm": 1.1913846731185913,
+      "learning_rate": 9.116715633137019e-05,
+      "loss": 0.6507744789123535,
+      "step": 3572
+    },
+    {
+      "epoch": 1.541963322545847,
+      "grad_norm": 1.1695647239685059,
+      "learning_rate": 9.115321630150919e-05,
+      "loss": 0.6736630201339722,
+      "step": 3574
+    },
+    {
+      "epoch": 1.542826321467098,
+      "grad_norm": 1.2866519689559937,
+      "learning_rate": 9.113926634773558e-05,
+      "loss": 0.6998549699783325,
+      "step": 3576
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 1.1755015850067139,
+      "learning_rate": 9.112530647341337e-05,
+      "loss": 0.6986453533172607,
+      "step": 3578
+    },
+    {
+      "epoch": 1.544552319309601,
+      "grad_norm": 1.2085564136505127,
+      "learning_rate": 9.111133668190889e-05,
+      "loss": 0.6959735155105591,
+      "step": 3580
+    },
+    {
+      "epoch": 1.545415318230852,
+      "grad_norm": 1.2213492393493652,
+      "learning_rate": 9.109735697659089e-05,
+      "loss": 0.6336050033569336,
+      "step": 3582
+    },
+    {
+      "epoch": 1.5462783171521035,
+      "grad_norm": 1.1609338521957397,
+      "learning_rate": 9.108336736083054e-05,
+      "loss": 0.652992844581604,
+      "step": 3584
+    },
+    {
+      "epoch": 1.547141316073355,
+      "grad_norm": 1.1483949422836304,
+      "learning_rate": 9.106936783800136e-05,
+      "loss": 0.5593333840370178,
+      "step": 3586
+    },
+    {
+      "epoch": 1.5480043149946061,
+      "grad_norm": 1.0852224826812744,
+      "learning_rate": 9.105535841147925e-05,
+      "loss": 0.6488125920295715,
+      "step": 3588
+    },
+    {
+      "epoch": 1.5488673139158577,
+      "grad_norm": 1.1804794073104858,
+      "learning_rate": 9.104133908464254e-05,
+      "loss": 0.5877764225006104,
+      "step": 3590
+    },
+    {
+      "epoch": 1.549730312837109,
+      "grad_norm": 1.0295292139053345,
+      "learning_rate": 9.102730986087194e-05,
+      "loss": 0.5924050211906433,
+      "step": 3592
+    },
+    {
+      "epoch": 1.5505933117583603,
+      "grad_norm": 1.524871587753296,
+      "learning_rate": 9.101327074355051e-05,
+      "loss": 0.6224870085716248,
+      "step": 3594
+    },
+    {
+      "epoch": 1.5514563106796118,
+      "grad_norm": 1.3207813501358032,
+      "learning_rate": 9.099922173606373e-05,
+      "loss": 0.5918852090835571,
+      "step": 3596
+    },
+    {
+      "epoch": 1.552319309600863,
+      "grad_norm": 1.2524348497390747,
+      "learning_rate": 9.098516284179945e-05,
+      "loss": 0.6756706237792969,
+      "step": 3598
+    },
+    {
+      "epoch": 1.5531823085221144,
+      "grad_norm": 1.2434089183807373,
+      "learning_rate": 9.097109406414792e-05,
+      "loss": 0.7416443824768066,
+      "step": 3600
+    },
+    {
+      "epoch": 1.5531823085221144,
+      "eval_loss": 0.701831042766571,
+      "eval_runtime": 836.0199,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 3600
+    },
+    {
+      "epoch": 1.5540453074433658,
+      "grad_norm": 1.0528122186660767,
+      "learning_rate": 9.095701540650175e-05,
+      "loss": 0.5911263227462769,
+      "step": 3602
+    },
+    {
+      "epoch": 1.554908306364617,
+      "grad_norm": 1.1165716648101807,
+      "learning_rate": 9.094292687225595e-05,
+      "loss": 0.5925720930099487,
+      "step": 3604
+    },
+    {
+      "epoch": 1.5557713052858684,
+      "grad_norm": 1.1127078533172607,
+      "learning_rate": 9.09288284648079e-05,
+      "loss": 0.6038441061973572,
+      "step": 3606
+    },
+    {
+      "epoch": 1.5566343042071198,
+      "grad_norm": 1.195662498474121,
+      "learning_rate": 9.091472018755738e-05,
+      "loss": 0.6781054735183716,
+      "step": 3608
+    },
+    {
+      "epoch": 1.557497303128371,
+      "grad_norm": 1.1671538352966309,
+      "learning_rate": 9.090060204390652e-05,
+      "loss": 0.7171973586082458,
+      "step": 3610
+    },
+    {
+      "epoch": 1.5583603020496224,
+      "grad_norm": 1.2814654111862183,
+      "learning_rate": 9.088647403725987e-05,
+      "loss": 0.6465495228767395,
+      "step": 3612
+    },
+    {
+      "epoch": 1.5592233009708738,
+      "grad_norm": 1.2238248586654663,
+      "learning_rate": 9.087233617102433e-05,
+      "loss": 0.7105259299278259,
+      "step": 3614
+    },
+    {
+      "epoch": 1.560086299892125,
+      "grad_norm": 1.1868736743927002,
+      "learning_rate": 9.085818844860916e-05,
+      "loss": 0.6721329092979431,
+      "step": 3616
+    },
+    {
+      "epoch": 1.5609492988133766,
+      "grad_norm": 1.1358642578125,
+      "learning_rate": 9.084403087342606e-05,
+      "loss": 0.6684948801994324,
+      "step": 3618
+    },
+    {
+      "epoch": 1.5618122977346278,
+      "grad_norm": 1.1101531982421875,
+      "learning_rate": 9.082986344888901e-05,
+      "loss": 0.653505802154541,
+      "step": 3620
+    },
+    {
+      "epoch": 1.5626752966558792,
+      "grad_norm": 1.3343961238861084,
+      "learning_rate": 9.081568617841447e-05,
+      "loss": 0.673059344291687,
+      "step": 3622
+    },
+    {
+      "epoch": 1.5635382955771306,
+      "grad_norm": 1.0680615901947021,
+      "learning_rate": 9.080149906542119e-05,
+      "loss": 0.6429969668388367,
+      "step": 3624
+    },
+    {
+      "epoch": 1.5644012944983818,
+      "grad_norm": 1.0026768445968628,
+      "learning_rate": 9.078730211333038e-05,
+      "loss": 0.694593608379364,
+      "step": 3626
+    },
+    {
+      "epoch": 1.5652642934196332,
+      "grad_norm": 1.100269079208374,
+      "learning_rate": 9.077309532556551e-05,
+      "loss": 0.6931326985359192,
+      "step": 3628
+    },
+    {
+      "epoch": 1.5661272923408847,
+      "grad_norm": 1.1724740266799927,
+      "learning_rate": 9.07588787055525e-05,
+      "loss": 0.6182643175125122,
+      "step": 3630
+    },
+    {
+      "epoch": 1.5669902912621358,
+      "grad_norm": 1.0137040615081787,
+      "learning_rate": 9.074465225671965e-05,
+      "loss": 0.6307282447814941,
+      "step": 3632
+    },
+    {
+      "epoch": 1.5678532901833873,
+      "grad_norm": 1.0997183322906494,
+      "learning_rate": 9.073041598249756e-05,
+      "loss": 0.6049362421035767,
+      "step": 3634
+    },
+    {
+      "epoch": 1.5687162891046387,
+      "grad_norm": 1.1688816547393799,
+      "learning_rate": 9.07161698863193e-05,
+      "loss": 0.6583153009414673,
+      "step": 3636
+    },
+    {
+      "epoch": 1.5695792880258899,
+      "grad_norm": 1.2558770179748535,
+      "learning_rate": 9.07019139716202e-05,
+      "loss": 0.6346786022186279,
+      "step": 3638
+    },
+    {
+      "epoch": 1.5704422869471413,
+      "grad_norm": 1.037838101387024,
+      "learning_rate": 9.068764824183804e-05,
+      "loss": 0.6214386224746704,
+      "step": 3640
+    },
+    {
+      "epoch": 1.5713052858683927,
+      "grad_norm": 1.2514159679412842,
+      "learning_rate": 9.067337270041293e-05,
+      "loss": 0.6745668649673462,
+      "step": 3642
+    },
+    {
+      "epoch": 1.5721682847896439,
+      "grad_norm": 1.2718034982681274,
+      "learning_rate": 9.065908735078736e-05,
+      "loss": 0.6409499049186707,
+      "step": 3644
+    },
+    {
+      "epoch": 1.5730312837108955,
+      "grad_norm": 1.1754566431045532,
+      "learning_rate": 9.064479219640615e-05,
+      "loss": 0.5885956883430481,
+      "step": 3646
+    },
+    {
+      "epoch": 1.5738942826321467,
+      "grad_norm": 1.3022228479385376,
+      "learning_rate": 9.063048724071653e-05,
+      "loss": 0.6472923159599304,
+      "step": 3648
+    },
+    {
+      "epoch": 1.574757281553398,
+      "grad_norm": 1.2492727041244507,
+      "learning_rate": 9.061617248716808e-05,
+      "loss": 0.6977139711380005,
+      "step": 3650
+    },
+    {
+      "epoch": 1.5756202804746495,
+      "grad_norm": 1.0432982444763184,
+      "learning_rate": 9.060184793921274e-05,
+      "loss": 0.6403328776359558,
+      "step": 3652
+    },
+    {
+      "epoch": 1.5764832793959007,
+      "grad_norm": 1.0898420810699463,
+      "learning_rate": 9.05875136003048e-05,
+      "loss": 0.6726073026657104,
+      "step": 3654
+    },
+    {
+      "epoch": 1.5773462783171521,
+      "grad_norm": 1.1885254383087158,
+      "learning_rate": 9.057316947390091e-05,
+      "loss": 0.6861021518707275,
+      "step": 3656
+    },
+    {
+      "epoch": 1.5782092772384035,
+      "grad_norm": 1.0795915126800537,
+      "learning_rate": 9.055881556346012e-05,
+      "loss": 0.6786487102508545,
+      "step": 3658
+    },
+    {
+      "epoch": 1.5790722761596547,
+      "grad_norm": 1.0702277421951294,
+      "learning_rate": 9.054445187244379e-05,
+      "loss": 0.6757274866104126,
+      "step": 3660
+    },
+    {
+      "epoch": 1.5799352750809061,
+      "grad_norm": 0.9469023942947388,
+      "learning_rate": 9.053007840431566e-05,
+      "loss": 0.6596087217330933,
+      "step": 3662
+    },
+    {
+      "epoch": 1.5807982740021576,
+      "grad_norm": 0.979669988155365,
+      "learning_rate": 9.051569516254185e-05,
+      "loss": 0.6254066228866577,
+      "step": 3664
+    },
+    {
+      "epoch": 1.5816612729234087,
+      "grad_norm": 1.294577717781067,
+      "learning_rate": 9.050130215059081e-05,
+      "loss": 0.6463974118232727,
+      "step": 3666
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 1.1566381454467773,
+      "learning_rate": 9.04868993719333e-05,
+      "loss": 0.6496619582176208,
+      "step": 3668
+    },
+    {
+      "epoch": 1.5833872707659116,
+      "grad_norm": 1.0207090377807617,
+      "learning_rate": 9.047248683004254e-05,
+      "loss": 0.5987569093704224,
+      "step": 3670
+    },
+    {
+      "epoch": 1.5842502696871628,
+      "grad_norm": 1.1533408164978027,
+      "learning_rate": 9.045806452839404e-05,
+      "loss": 0.6392318606376648,
+      "step": 3672
+    },
+    {
+      "epoch": 1.5851132686084144,
+      "grad_norm": 1.0433615446090698,
+      "learning_rate": 9.044363247046565e-05,
+      "loss": 0.6197013854980469,
+      "step": 3674
+    },
+    {
+      "epoch": 1.5859762675296656,
+      "grad_norm": 1.2047847509384155,
+      "learning_rate": 9.042919065973763e-05,
+      "loss": 0.620261549949646,
+      "step": 3676
+    },
+    {
+      "epoch": 1.5868392664509168,
+      "grad_norm": 1.232506513595581,
+      "learning_rate": 9.041473909969252e-05,
+      "loss": 0.6751506924629211,
+      "step": 3678
+    },
+    {
+      "epoch": 1.5877022653721684,
+      "grad_norm": 1.1659306287765503,
+      "learning_rate": 9.040027779381528e-05,
+      "loss": 0.6619004011154175,
+      "step": 3680
+    },
+    {
+      "epoch": 1.5885652642934196,
+      "grad_norm": 1.0183755159378052,
+      "learning_rate": 9.038580674559316e-05,
+      "loss": 0.6120786666870117,
+      "step": 3682
+    },
+    {
+      "epoch": 1.589428263214671,
+      "grad_norm": 1.194318175315857,
+      "learning_rate": 9.037132595851581e-05,
+      "loss": 0.616247296333313,
+      "step": 3684
+    },
+    {
+      "epoch": 1.5902912621359224,
+      "grad_norm": 1.500475525856018,
+      "learning_rate": 9.035683543607522e-05,
+      "loss": 0.7229626774787903,
+      "step": 3686
+    },
+    {
+      "epoch": 1.5911542610571736,
+      "grad_norm": 1.047954797744751,
+      "learning_rate": 9.034233518176567e-05,
+      "loss": 0.6193928718566895,
+      "step": 3688
+    },
+    {
+      "epoch": 1.592017259978425,
+      "grad_norm": 1.3852946758270264,
+      "learning_rate": 9.032782519908386e-05,
+      "loss": 0.7276740074157715,
+      "step": 3690
+    },
+    {
+      "epoch": 1.5928802588996764,
+      "grad_norm": 1.2648470401763916,
+      "learning_rate": 9.031330549152882e-05,
+      "loss": 0.6635852456092834,
+      "step": 3692
+    },
+    {
+      "epoch": 1.5937432578209276,
+      "grad_norm": 1.1195251941680908,
+      "learning_rate": 9.029877606260186e-05,
+      "loss": 0.6905328631401062,
+      "step": 3694
+    },
+    {
+      "epoch": 1.594606256742179,
+      "grad_norm": 1.0823792219161987,
+      "learning_rate": 9.028423691580675e-05,
+      "loss": 0.6474795341491699,
+      "step": 3696
+    },
+    {
+      "epoch": 1.5954692556634305,
+      "grad_norm": 1.1743394136428833,
+      "learning_rate": 9.026968805464948e-05,
+      "loss": 0.6740748882293701,
+      "step": 3698
+    },
+    {
+      "epoch": 1.5963322545846816,
+      "grad_norm": 1.19667387008667,
+      "learning_rate": 9.02551294826385e-05,
+      "loss": 0.5662298798561096,
+      "step": 3700
+    },
+    {
+      "epoch": 1.5963322545846816,
+      "eval_loss": 0.6999035477638245,
+      "eval_runtime": 836.4944,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 3700
+    },
+    {
+      "epoch": 1.5971952535059333,
+      "grad_norm": 0.9947987794876099,
+      "learning_rate": 9.02405612032845e-05,
+      "loss": 0.6141433119773865,
+      "step": 3702
+    },
+    {
+      "epoch": 1.5980582524271845,
+      "grad_norm": 1.0307772159576416,
+      "learning_rate": 9.022598322010059e-05,
+      "loss": 0.6011156439781189,
+      "step": 3704
+    },
+    {
+      "epoch": 1.5989212513484357,
+      "grad_norm": 1.3516873121261597,
+      "learning_rate": 9.021139553660214e-05,
+      "loss": 0.7671760320663452,
+      "step": 3706
+    },
+    {
+      "epoch": 1.5997842502696873,
+      "grad_norm": 1.0319154262542725,
+      "learning_rate": 9.019679815630695e-05,
+      "loss": 0.5630212426185608,
+      "step": 3708
+    },
+    {
+      "epoch": 1.6006472491909385,
+      "grad_norm": 1.34807550907135,
+      "learning_rate": 9.01821910827351e-05,
+      "loss": 0.7143195271492004,
+      "step": 3710
+    },
+    {
+      "epoch": 1.60151024811219,
+      "grad_norm": 1.0618840456008911,
+      "learning_rate": 9.016757431940898e-05,
+      "loss": 0.6624904274940491,
+      "step": 3712
+    },
+    {
+      "epoch": 1.6023732470334413,
+      "grad_norm": 0.987445056438446,
+      "learning_rate": 9.015294786985342e-05,
+      "loss": 0.631399929523468,
+      "step": 3714
+    },
+    {
+      "epoch": 1.6032362459546925,
+      "grad_norm": 1.0872961282730103,
+      "learning_rate": 9.013831173759548e-05,
+      "loss": 0.646812379360199,
+      "step": 3716
+    },
+    {
+      "epoch": 1.604099244875944,
+      "grad_norm": 0.9798976182937622,
+      "learning_rate": 9.012366592616461e-05,
+      "loss": 0.6353676915168762,
+      "step": 3718
+    },
+    {
+      "epoch": 1.6049622437971953,
+      "grad_norm": 1.3835123777389526,
+      "learning_rate": 9.010901043909258e-05,
+      "loss": 0.5802496671676636,
+      "step": 3720
+    },
+    {
+      "epoch": 1.6058252427184465,
+      "grad_norm": 1.0870599746704102,
+      "learning_rate": 9.009434527991349e-05,
+      "loss": 0.6140236854553223,
+      "step": 3722
+    },
+    {
+      "epoch": 1.606688241639698,
+      "grad_norm": 1.0582176446914673,
+      "learning_rate": 9.007967045216379e-05,
+      "loss": 0.6562411785125732,
+      "step": 3724
+    },
+    {
+      "epoch": 1.6075512405609493,
+      "grad_norm": 1.0486522912979126,
+      "learning_rate": 9.006498595938224e-05,
+      "loss": 0.5995528697967529,
+      "step": 3726
+    },
+    {
+      "epoch": 1.6084142394822005,
+      "grad_norm": 1.1152576208114624,
+      "learning_rate": 9.005029180510994e-05,
+      "loss": 0.6296200752258301,
+      "step": 3728
+    },
+    {
+      "epoch": 1.6092772384034522,
+      "grad_norm": 1.2114241123199463,
+      "learning_rate": 9.003558799289031e-05,
+      "loss": 0.7050777077674866,
+      "step": 3730
+    },
+    {
+      "epoch": 1.6101402373247033,
+      "grad_norm": 1.1112302541732788,
+      "learning_rate": 9.002087452626913e-05,
+      "loss": 0.6323828101158142,
+      "step": 3732
+    },
+    {
+      "epoch": 1.6110032362459545,
+      "grad_norm": 1.102839469909668,
+      "learning_rate": 9.000615140879446e-05,
+      "loss": 0.7021811008453369,
+      "step": 3734
+    },
+    {
+      "epoch": 1.6118662351672062,
+      "grad_norm": 0.8860642910003662,
+      "learning_rate": 8.999141864401672e-05,
+      "loss": 0.5094663500785828,
+      "step": 3736
+    },
+    {
+      "epoch": 1.6127292340884574,
+      "grad_norm": 1.1873432397842407,
+      "learning_rate": 8.997667623548869e-05,
+      "loss": 0.6481591463088989,
+      "step": 3738
+    },
+    {
+      "epoch": 1.6135922330097088,
+      "grad_norm": 0.9258447885513306,
+      "learning_rate": 8.996192418676537e-05,
+      "loss": 0.6846200227737427,
+      "step": 3740
+    },
+    {
+      "epoch": 1.6144552319309602,
+      "grad_norm": 1.1108704805374146,
+      "learning_rate": 8.994716250140419e-05,
+      "loss": 0.7081074118614197,
+      "step": 3742
+    },
+    {
+      "epoch": 1.6153182308522114,
+      "grad_norm": 1.01946222782135,
+      "learning_rate": 8.993239118296486e-05,
+      "loss": 0.683017373085022,
+      "step": 3744
+    },
+    {
+      "epoch": 1.6161812297734628,
+      "grad_norm": 1.0794068574905396,
+      "learning_rate": 8.991761023500942e-05,
+      "loss": 0.6625820994377136,
+      "step": 3746
+    },
+    {
+      "epoch": 1.6170442286947142,
+      "grad_norm": 1.0697723627090454,
+      "learning_rate": 8.990281966110221e-05,
+      "loss": 0.6714970469474792,
+      "step": 3748
+    },
+    {
+      "epoch": 1.6179072276159654,
+      "grad_norm": 1.118213176727295,
+      "learning_rate": 8.988801946480991e-05,
+      "loss": 0.6702249646186829,
+      "step": 3750
+    },
+    {
+      "epoch": 1.6187702265372168,
+      "grad_norm": 1.3541207313537598,
+      "learning_rate": 8.987320964970156e-05,
+      "loss": 0.6783145070075989,
+      "step": 3752
+    },
+    {
+      "epoch": 1.6196332254584682,
+      "grad_norm": 1.2015430927276611,
+      "learning_rate": 8.985839021934843e-05,
+      "loss": 0.6774528622627258,
+      "step": 3754
+    },
+    {
+      "epoch": 1.6204962243797194,
+      "grad_norm": 1.2218681573867798,
+      "learning_rate": 8.984356117732419e-05,
+      "loss": 0.6122720241546631,
+      "step": 3756
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 1.0959049463272095,
+      "learning_rate": 8.982872252720478e-05,
+      "loss": 0.6658767461776733,
+      "step": 3758
+    },
+    {
+      "epoch": 1.6222222222222222,
+      "grad_norm": 1.0399792194366455,
+      "learning_rate": 8.981387427256849e-05,
+      "loss": 0.6943061947822571,
+      "step": 3760
+    },
+    {
+      "epoch": 1.6230852211434734,
+      "grad_norm": 1.1107885837554932,
+      "learning_rate": 8.979901641699591e-05,
+      "loss": 0.6100694537162781,
+      "step": 3762
+    },
+    {
+      "epoch": 1.623948220064725,
+      "grad_norm": 1.2478004693984985,
+      "learning_rate": 8.978414896406993e-05,
+      "loss": 0.7148857712745667,
+      "step": 3764
+    },
+    {
+      "epoch": 1.6248112189859762,
+      "grad_norm": 1.333235502243042,
+      "learning_rate": 8.976927191737577e-05,
+      "loss": 0.6626213788986206,
+      "step": 3766
+    },
+    {
+      "epoch": 1.6256742179072277,
+      "grad_norm": 1.2001140117645264,
+      "learning_rate": 8.975438528050095e-05,
+      "loss": 0.6073925495147705,
+      "step": 3768
+    },
+    {
+      "epoch": 1.626537216828479,
+      "grad_norm": 0.8740705847740173,
+      "learning_rate": 8.973948905703535e-05,
+      "loss": 0.5861713290214539,
+      "step": 3770
+    },
+    {
+      "epoch": 1.6274002157497303,
+      "grad_norm": 1.0697330236434937,
+      "learning_rate": 8.972458325057111e-05,
+      "loss": 0.7104416489601135,
+      "step": 3772
+    },
+    {
+      "epoch": 1.6282632146709817,
+      "grad_norm": 1.2699353694915771,
+      "learning_rate": 8.970966786470271e-05,
+      "loss": 0.652777373790741,
+      "step": 3774
+    },
+    {
+      "epoch": 1.629126213592233,
+      "grad_norm": 1.2045143842697144,
+      "learning_rate": 8.969474290302691e-05,
+      "loss": 0.6303225159645081,
+      "step": 3776
+    },
+    {
+      "epoch": 1.6299892125134843,
+      "grad_norm": 1.2175726890563965,
+      "learning_rate": 8.96798083691428e-05,
+      "loss": 0.594299852848053,
+      "step": 3778
+    },
+    {
+      "epoch": 1.6308522114347357,
+      "grad_norm": 0.9873425960540771,
+      "learning_rate": 8.966486426665178e-05,
+      "loss": 0.566745400428772,
+      "step": 3780
+    },
+    {
+      "epoch": 1.631715210355987,
+      "grad_norm": 1.4304969310760498,
+      "learning_rate": 8.964991059915756e-05,
+      "loss": 0.76564621925354,
+      "step": 3782
+    },
+    {
+      "epoch": 1.6325782092772383,
+      "grad_norm": 1.1225755214691162,
+      "learning_rate": 8.963494737026611e-05,
+      "loss": 0.7026722431182861,
+      "step": 3784
+    },
+    {
+      "epoch": 1.6334412081984897,
+      "grad_norm": 1.251551628112793,
+      "learning_rate": 8.961997458358581e-05,
+      "loss": 0.6778496503829956,
+      "step": 3786
+    },
+    {
+      "epoch": 1.6343042071197411,
+      "grad_norm": 0.9573827981948853,
+      "learning_rate": 8.960499224272725e-05,
+      "loss": 0.5866316556930542,
+      "step": 3788
+    },
+    {
+      "epoch": 1.6351672060409923,
+      "grad_norm": 1.0466986894607544,
+      "learning_rate": 8.959000035130334e-05,
+      "loss": 0.6244695782661438,
+      "step": 3790
+    },
+    {
+      "epoch": 1.636030204962244,
+      "grad_norm": 1.0724520683288574,
+      "learning_rate": 8.957499891292931e-05,
+      "loss": 0.6380400061607361,
+      "step": 3792
+    },
+    {
+      "epoch": 1.6368932038834951,
+      "grad_norm": 1.1590946912765503,
+      "learning_rate": 8.955998793122271e-05,
+      "loss": 0.625150203704834,
+      "step": 3794
+    },
+    {
+      "epoch": 1.6377562028047465,
+      "grad_norm": 1.3405407667160034,
+      "learning_rate": 8.954496740980334e-05,
+      "loss": 0.7554174065589905,
+      "step": 3796
+    },
+    {
+      "epoch": 1.638619201725998,
+      "grad_norm": 1.2695956230163574,
+      "learning_rate": 8.952993735229338e-05,
+      "loss": 0.6710763573646545,
+      "step": 3798
+    },
+    {
+      "epoch": 1.6394822006472491,
+      "grad_norm": 1.4330781698226929,
+      "learning_rate": 8.95148977623172e-05,
+      "loss": 0.7102031111717224,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6394822006472491,
+      "eval_loss": 0.6974969506263733,
+      "eval_runtime": 823.7017,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6403451995685006,
+      "grad_norm": 0.9818692803382874,
+      "learning_rate": 8.949984864350157e-05,
+      "loss": 0.6028781533241272,
+      "step": 3802
+    },
+    {
+      "epoch": 1.641208198489752,
+      "grad_norm": 1.0779424905776978,
+      "learning_rate": 8.94847899994755e-05,
+      "loss": 0.649263322353363,
+      "step": 3804
+    },
+    {
+      "epoch": 1.6420711974110032,
+      "grad_norm": 1.2096810340881348,
+      "learning_rate": 8.946972183387033e-05,
+      "loss": 0.6424242258071899,
+      "step": 3806
+    },
+    {
+      "epoch": 1.6429341963322546,
+      "grad_norm": 1.1697514057159424,
+      "learning_rate": 8.945464415031966e-05,
+      "loss": 0.6751719117164612,
+      "step": 3808
+    },
+    {
+      "epoch": 1.643797195253506,
+      "grad_norm": 1.6524088382720947,
+      "learning_rate": 8.943955695245942e-05,
+      "loss": 0.616973340511322,
+      "step": 3810
+    },
+    {
+      "epoch": 1.6446601941747572,
+      "grad_norm": 1.1154218912124634,
+      "learning_rate": 8.942446024392782e-05,
+      "loss": 0.6577548980712891,
+      "step": 3812
+    },
+    {
+      "epoch": 1.6455231930960086,
+      "grad_norm": 1.1747835874557495,
+      "learning_rate": 8.940935402836535e-05,
+      "loss": 0.6534224152565002,
+      "step": 3814
+    },
+    {
+      "epoch": 1.64638619201726,
+      "grad_norm": 1.189562439918518,
+      "learning_rate": 8.939423830941482e-05,
+      "loss": 0.6356330513954163,
+      "step": 3816
+    },
+    {
+      "epoch": 1.6472491909385112,
+      "grad_norm": 1.0842949151992798,
+      "learning_rate": 8.937911309072132e-05,
+      "loss": 0.6097415685653687,
+      "step": 3818
+    },
+    {
+      "epoch": 1.6481121898597628,
+      "grad_norm": 1.1931639909744263,
+      "learning_rate": 8.936397837593221e-05,
+      "loss": 0.6339239478111267,
+      "step": 3820
+    },
+    {
+      "epoch": 1.648975188781014,
+      "grad_norm": 1.4980956315994263,
+      "learning_rate": 8.934883416869717e-05,
+      "loss": 0.6588500142097473,
+      "step": 3822
+    },
+    {
+      "epoch": 1.6498381877022654,
+      "grad_norm": 1.2506407499313354,
+      "learning_rate": 8.933368047266814e-05,
+      "loss": 0.6418336033821106,
+      "step": 3824
+    },
+    {
+      "epoch": 1.6507011866235168,
+      "grad_norm": 1.3003007173538208,
+      "learning_rate": 8.93185172914994e-05,
+      "loss": 0.6475774049758911,
+      "step": 3826
+    },
+    {
+      "epoch": 1.651564185544768,
+      "grad_norm": 1.2024277448654175,
+      "learning_rate": 8.930334462884747e-05,
+      "loss": 0.7352058291435242,
+      "step": 3828
+    },
+    {
+      "epoch": 1.6524271844660194,
+      "grad_norm": 1.0713937282562256,
+      "learning_rate": 8.928816248837115e-05,
+      "loss": 0.6267798542976379,
+      "step": 3830
+    },
+    {
+      "epoch": 1.6532901833872709,
+      "grad_norm": 1.3538615703582764,
+      "learning_rate": 8.927297087373157e-05,
+      "loss": 0.7097233533859253,
+      "step": 3832
+    },
+    {
+      "epoch": 1.654153182308522,
+      "grad_norm": 1.1811957359313965,
+      "learning_rate": 8.92577697885921e-05,
+      "loss": 0.6012749671936035,
+      "step": 3834
+    },
+    {
+      "epoch": 1.6550161812297735,
+      "grad_norm": 1.064659595489502,
+      "learning_rate": 8.924255923661841e-05,
+      "loss": 0.6257606744766235,
+      "step": 3836
+    },
+    {
+      "epoch": 1.6558791801510249,
+      "grad_norm": 1.1258097887039185,
+      "learning_rate": 8.922733922147848e-05,
+      "loss": 0.7108950018882751,
+      "step": 3838
+    },
+    {
+      "epoch": 1.656742179072276,
+      "grad_norm": 1.4515002965927124,
+      "learning_rate": 8.921210974684253e-05,
+      "loss": 0.6432359218597412,
+      "step": 3840
+    },
+    {
+      "epoch": 1.6576051779935275,
+      "grad_norm": 1.040401816368103,
+      "learning_rate": 8.919687081638308e-05,
+      "loss": 0.5962746739387512,
+      "step": 3842
+    },
+    {
+      "epoch": 1.6584681769147789,
+      "grad_norm": 1.0090585947036743,
+      "learning_rate": 8.918162243377493e-05,
+      "loss": 0.6396839022636414,
+      "step": 3844
+    },
+    {
+      "epoch": 1.65933117583603,
+      "grad_norm": 1.2226985692977905,
+      "learning_rate": 8.916636460269517e-05,
+      "loss": 0.7058185935020447,
+      "step": 3846
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 1.117598295211792,
+      "learning_rate": 8.915109732682316e-05,
+      "loss": 0.6138123273849487,
+      "step": 3848
+    },
+    {
+      "epoch": 1.661057173678533,
+      "grad_norm": 1.1237187385559082,
+      "learning_rate": 8.913582060984052e-05,
+      "loss": 0.6389812231063843,
+      "step": 3850
+    },
+    {
+      "epoch": 1.661920172599784,
+      "grad_norm": 1.148144245147705,
+      "learning_rate": 8.912053445543114e-05,
+      "loss": 0.7189341187477112,
+      "step": 3852
+    },
+    {
+      "epoch": 1.6627831715210357,
+      "grad_norm": 1.3115259408950806,
+      "learning_rate": 8.910523886728126e-05,
+      "loss": 0.671591579914093,
+      "step": 3854
+    },
+    {
+      "epoch": 1.663646170442287,
+      "grad_norm": 1.0031025409698486,
+      "learning_rate": 8.908993384907928e-05,
+      "loss": 0.6695008873939514,
+      "step": 3856
+    },
+    {
+      "epoch": 1.6645091693635383,
+      "grad_norm": 1.2426995038986206,
+      "learning_rate": 8.907461940451601e-05,
+      "loss": 0.6557230353355408,
+      "step": 3858
+    },
+    {
+      "epoch": 1.6653721682847897,
+      "grad_norm": 1.1471003293991089,
+      "learning_rate": 8.90592955372844e-05,
+      "loss": 0.6126149296760559,
+      "step": 3860
+    },
+    {
+      "epoch": 1.666235167206041,
+      "grad_norm": 1.2763721942901611,
+      "learning_rate": 8.904396225107976e-05,
+      "loss": 0.742831826210022,
+      "step": 3862
+    },
+    {
+      "epoch": 1.6670981661272923,
+      "grad_norm": 1.0612274408340454,
+      "learning_rate": 8.902861954959963e-05,
+      "loss": 0.6506689190864563,
+      "step": 3864
+    },
+    {
+      "epoch": 1.6679611650485437,
+      "grad_norm": 1.160914421081543,
+      "learning_rate": 8.901326743654384e-05,
+      "loss": 0.6939412355422974,
+      "step": 3866
+    },
+    {
+      "epoch": 1.668824163969795,
+      "grad_norm": 1.2517154216766357,
+      "learning_rate": 8.899790591561447e-05,
+      "loss": 0.6331133842468262,
+      "step": 3868
+    },
+    {
+      "epoch": 1.6696871628910464,
+      "grad_norm": 0.9491620063781738,
+      "learning_rate": 8.89825349905159e-05,
+      "loss": 0.6425275206565857,
+      "step": 3870
+    },
+    {
+      "epoch": 1.6705501618122978,
+      "grad_norm": 0.9615546464920044,
+      "learning_rate": 8.896715466495476e-05,
+      "loss": 0.5937016606330872,
+      "step": 3872
+    },
+    {
+      "epoch": 1.671413160733549,
+      "grad_norm": 1.1985108852386475,
+      "learning_rate": 8.895176494263992e-05,
+      "loss": 0.7523217797279358,
+      "step": 3874
+    },
+    {
+      "epoch": 1.6722761596548006,
+      "grad_norm": 1.1150988340377808,
+      "learning_rate": 8.893636582728257e-05,
+      "loss": 0.6035639047622681,
+      "step": 3876
+    },
+    {
+      "epoch": 1.6731391585760518,
+      "grad_norm": 0.9545241594314575,
+      "learning_rate": 8.892095732259612e-05,
+      "loss": 0.6141317486763,
+      "step": 3878
+    },
+    {
+      "epoch": 1.674002157497303,
+      "grad_norm": 1.2937493324279785,
+      "learning_rate": 8.890553943229627e-05,
+      "loss": 0.6015914678573608,
+      "step": 3880
+    },
+    {
+      "epoch": 1.6748651564185546,
+      "grad_norm": 1.0606268644332886,
+      "learning_rate": 8.889011216010099e-05,
+      "loss": 0.6010603904724121,
+      "step": 3882
+    },
+    {
+      "epoch": 1.6757281553398058,
+      "grad_norm": 1.0271397829055786,
+      "learning_rate": 8.887467550973046e-05,
+      "loss": 0.6256402134895325,
+      "step": 3884
+    },
+    {
+      "epoch": 1.6765911542610572,
+      "grad_norm": 1.096360683441162,
+      "learning_rate": 8.885922948490719e-05,
+      "loss": 0.6119387745857239,
+      "step": 3886
+    },
+    {
+      "epoch": 1.6774541531823086,
+      "grad_norm": 1.1860841512680054,
+      "learning_rate": 8.884377408935589e-05,
+      "loss": 0.6526403427124023,
+      "step": 3888
+    },
+    {
+      "epoch": 1.6783171521035598,
+      "grad_norm": 1.1262441873550415,
+      "learning_rate": 8.882830932680359e-05,
+      "loss": 0.5966535806655884,
+      "step": 3890
+    },
+    {
+      "epoch": 1.6791801510248112,
+      "grad_norm": 1.0326828956604004,
+      "learning_rate": 8.881283520097953e-05,
+      "loss": 0.6420881748199463,
+      "step": 3892
+    },
+    {
+      "epoch": 1.6800431499460626,
+      "grad_norm": 1.1928532123565674,
+      "learning_rate": 8.879735171561525e-05,
+      "loss": 0.6054600477218628,
+      "step": 3894
+    },
+    {
+      "epoch": 1.6809061488673138,
+      "grad_norm": 1.2887117862701416,
+      "learning_rate": 8.878185887444448e-05,
+      "loss": 0.7106166481971741,
+      "step": 3896
+    },
+    {
+      "epoch": 1.6817691477885652,
+      "grad_norm": 1.1483163833618164,
+      "learning_rate": 8.876635668120327e-05,
+      "loss": 0.7001441717147827,
+      "step": 3898
+    },
+    {
+      "epoch": 1.6826321467098166,
+      "grad_norm": 1.2908320426940918,
+      "learning_rate": 8.875084513962993e-05,
+      "loss": 0.6607111692428589,
+      "step": 3900
+    },
+    {
+      "epoch": 1.6826321467098166,
+      "eval_loss": 0.6949095726013184,
+      "eval_runtime": 817.4497,
+      "eval_samples_per_second": 2.52,
+      "eval_steps_per_second": 2.52,
+      "step": 3900
+    },
+    {
+      "epoch": 1.6834951456310678,
+      "grad_norm": 1.00212824344635,
+      "learning_rate": 8.873532425346496e-05,
+      "loss": 0.6463292837142944,
+      "step": 3902
+    },
+    {
+      "epoch": 1.6843581445523195,
+      "grad_norm": 1.2223109006881714,
+      "learning_rate": 8.871979402645116e-05,
+      "loss": 0.6627839207649231,
+      "step": 3904
+    },
+    {
+      "epoch": 1.6852211434735707,
+      "grad_norm": 1.047956109046936,
+      "learning_rate": 8.870425446233359e-05,
+      "loss": 0.6311076879501343,
+      "step": 3906
+    },
+    {
+      "epoch": 1.6860841423948218,
+      "grad_norm": 1.178468108177185,
+      "learning_rate": 8.868870556485954e-05,
+      "loss": 0.6916288137435913,
+      "step": 3908
+    },
+    {
+      "epoch": 1.6869471413160735,
+      "grad_norm": 6.569681644439697,
+      "learning_rate": 8.867314733777854e-05,
+      "loss": 0.696479856967926,
+      "step": 3910
+    },
+    {
+      "epoch": 1.6878101402373247,
+      "grad_norm": 1.100183129310608,
+      "learning_rate": 8.865757978484241e-05,
+      "loss": 0.6245654225349426,
+      "step": 3912
+    },
+    {
+      "epoch": 1.688673139158576,
+      "grad_norm": 1.222562313079834,
+      "learning_rate": 8.864200290980518e-05,
+      "loss": 0.6178512573242188,
+      "step": 3914
+    },
+    {
+      "epoch": 1.6895361380798275,
+      "grad_norm": 1.235471487045288,
+      "learning_rate": 8.862641671642314e-05,
+      "loss": 0.6598679423332214,
+      "step": 3916
+    },
+    {
+      "epoch": 1.6903991370010787,
+      "grad_norm": 1.2539454698562622,
+      "learning_rate": 8.861082120845486e-05,
+      "loss": 0.6503698825836182,
+      "step": 3918
+    },
+    {
+      "epoch": 1.69126213592233,
+      "grad_norm": 1.1766268014907837,
+      "learning_rate": 8.859521638966107e-05,
+      "loss": 0.6858329176902771,
+      "step": 3920
+    },
+    {
+      "epoch": 1.6921251348435815,
+      "grad_norm": 1.3463754653930664,
+      "learning_rate": 8.857960226380486e-05,
+      "loss": 0.679460883140564,
+      "step": 3922
+    },
+    {
+      "epoch": 1.6929881337648327,
+      "grad_norm": 1.3011246919631958,
+      "learning_rate": 8.856397883465149e-05,
+      "loss": 0.6901042461395264,
+      "step": 3924
+    },
+    {
+      "epoch": 1.6938511326860841,
+      "grad_norm": 0.9615590572357178,
+      "learning_rate": 8.854834610596845e-05,
+      "loss": 0.6756444573402405,
+      "step": 3926
+    },
+    {
+      "epoch": 1.6947141316073355,
+      "grad_norm": 0.9635165333747864,
+      "learning_rate": 8.853270408152552e-05,
+      "loss": 0.6690155863761902,
+      "step": 3928
+    },
+    {
+      "epoch": 1.6955771305285867,
+      "grad_norm": 1.2924212217330933,
+      "learning_rate": 8.851705276509472e-05,
+      "loss": 0.6412895321846008,
+      "step": 3930
+    },
+    {
+      "epoch": 1.6964401294498384,
+      "grad_norm": 1.0281823873519897,
+      "learning_rate": 8.850139216045028e-05,
+      "loss": 0.6250708699226379,
+      "step": 3932
+    },
+    {
+      "epoch": 1.6973031283710895,
+      "grad_norm": 1.321605920791626,
+      "learning_rate": 8.848572227136868e-05,
+      "loss": 0.6545571684837341,
+      "step": 3934
+    },
+    {
+      "epoch": 1.6981661272923407,
+      "grad_norm": 1.1515092849731445,
+      "learning_rate": 8.847004310162866e-05,
+      "loss": 0.6466772556304932,
+      "step": 3936
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 1.1318165063858032,
+      "learning_rate": 8.845435465501116e-05,
+      "loss": 0.679786741733551,
+      "step": 3938
+    },
+    {
+      "epoch": 1.6998921251348436,
+      "grad_norm": 0.973469078540802,
+      "learning_rate": 8.84386569352994e-05,
+      "loss": 0.6546611189842224,
+      "step": 3940
+    },
+    {
+      "epoch": 1.700755124056095,
+      "grad_norm": 1.0992401838302612,
+      "learning_rate": 8.842294994627879e-05,
+      "loss": 0.6207357048988342,
+      "step": 3942
+    },
+    {
+      "epoch": 1.7016181229773464,
+      "grad_norm": 1.1114588975906372,
+      "learning_rate": 8.840723369173702e-05,
+      "loss": 0.6113827228546143,
+      "step": 3944
+    },
+    {
+      "epoch": 1.7024811218985976,
+      "grad_norm": 1.171692132949829,
+      "learning_rate": 8.839150817546398e-05,
+      "loss": 0.6246283650398254,
+      "step": 3946
+    },
+    {
+      "epoch": 1.703344120819849,
+      "grad_norm": 1.249851942062378,
+      "learning_rate": 8.837577340125181e-05,
+      "loss": 0.7027832269668579,
+      "step": 3948
+    },
+    {
+      "epoch": 1.7042071197411004,
+      "grad_norm": 1.2475496530532837,
+      "learning_rate": 8.836002937289489e-05,
+      "loss": 0.7121281027793884,
+      "step": 3950
+    },
+    {
+      "epoch": 1.7050701186623516,
+      "grad_norm": 1.048726201057434,
+      "learning_rate": 8.834427609418981e-05,
+      "loss": 0.6521060466766357,
+      "step": 3952
+    },
+    {
+      "epoch": 1.705933117583603,
+      "grad_norm": 1.1689599752426147,
+      "learning_rate": 8.832851356893541e-05,
+      "loss": 0.6735731959342957,
+      "step": 3954
+    },
+    {
+      "epoch": 1.7067961165048544,
+      "grad_norm": 1.4230015277862549,
+      "learning_rate": 8.831274180093275e-05,
+      "loss": 0.6530178189277649,
+      "step": 3956
+    },
+    {
+      "epoch": 1.7076591154261056,
+      "grad_norm": 1.2237123250961304,
+      "learning_rate": 8.829696079398511e-05,
+      "loss": 0.6858299970626831,
+      "step": 3958
+    },
+    {
+      "epoch": 1.708522114347357,
+      "grad_norm": 1.1330037117004395,
+      "learning_rate": 8.828117055189803e-05,
+      "loss": 0.6567219495773315,
+      "step": 3960
+    },
+    {
+      "epoch": 1.7093851132686084,
+      "grad_norm": 1.2441636323928833,
+      "learning_rate": 8.826537107847925e-05,
+      "loss": 0.6207150816917419,
+      "step": 3962
+    },
+    {
+      "epoch": 1.7102481121898596,
+      "grad_norm": 0.9952784776687622,
+      "learning_rate": 8.824956237753872e-05,
+      "loss": 0.6192324161529541,
+      "step": 3964
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 1.0648247003555298,
+      "learning_rate": 8.823374445288865e-05,
+      "loss": 0.6485670208930969,
+      "step": 3966
+    },
+    {
+      "epoch": 1.7119741100323624,
+      "grad_norm": 1.2040860652923584,
+      "learning_rate": 8.821791730834347e-05,
+      "loss": 0.6591274738311768,
+      "step": 3968
+    },
+    {
+      "epoch": 1.7128371089536139,
+      "grad_norm": 1.2346793413162231,
+      "learning_rate": 8.820208094771982e-05,
+      "loss": 0.6427604556083679,
+      "step": 3970
+    },
+    {
+      "epoch": 1.7137001078748653,
+      "grad_norm": 1.0716660022735596,
+      "learning_rate": 8.818623537483658e-05,
+      "loss": 0.5952412486076355,
+      "step": 3972
+    },
+    {
+      "epoch": 1.7145631067961165,
+      "grad_norm": 1.1809104681015015,
+      "learning_rate": 8.817038059351481e-05,
+      "loss": 0.6808739900588989,
+      "step": 3974
+    },
+    {
+      "epoch": 1.7154261057173679,
+      "grad_norm": 1.224974513053894,
+      "learning_rate": 8.815451660757783e-05,
+      "loss": 0.6927714347839355,
+      "step": 3976
+    },
+    {
+      "epoch": 1.7162891046386193,
+      "grad_norm": 1.141288161277771,
+      "learning_rate": 8.813864342085117e-05,
+      "loss": 0.6450885534286499,
+      "step": 3978
+    },
+    {
+      "epoch": 1.7171521035598705,
+      "grad_norm": 1.0749722719192505,
+      "learning_rate": 8.812276103716259e-05,
+      "loss": 0.6775641441345215,
+      "step": 3980
+    },
+    {
+      "epoch": 1.7180151024811219,
+      "grad_norm": 1.13978111743927,
+      "learning_rate": 8.810686946034203e-05,
+      "loss": 0.6083413362503052,
+      "step": 3982
+    },
+    {
+      "epoch": 1.7188781014023733,
+      "grad_norm": 1.3158389329910278,
+      "learning_rate": 8.80909686942217e-05,
+      "loss": 0.6711311936378479,
+      "step": 3984
+    },
+    {
+      "epoch": 1.7197411003236245,
+      "grad_norm": 1.0368914604187012,
+      "learning_rate": 8.807505874263597e-05,
+      "loss": 0.6818822026252747,
+      "step": 3986
+    },
+    {
+      "epoch": 1.720604099244876,
+      "grad_norm": 1.249948263168335,
+      "learning_rate": 8.805913960942148e-05,
+      "loss": 0.6346645355224609,
+      "step": 3988
+    },
+    {
+      "epoch": 1.7214670981661273,
+      "grad_norm": 1.0894579887390137,
+      "learning_rate": 8.804321129841703e-05,
+      "loss": 0.6550438404083252,
+      "step": 3990
+    },
+    {
+      "epoch": 1.7223300970873785,
+      "grad_norm": 1.2956905364990234,
+      "learning_rate": 8.80272738134637e-05,
+      "loss": 0.6257092952728271,
+      "step": 3992
+    },
+    {
+      "epoch": 1.7231930960086301,
+      "grad_norm": 1.0089573860168457,
+      "learning_rate": 8.80113271584047e-05,
+      "loss": 0.5877772569656372,
+      "step": 3994
+    },
+    {
+      "epoch": 1.7240560949298813,
+      "grad_norm": 1.0698174238204956,
+      "learning_rate": 8.799537133708552e-05,
+      "loss": 0.6440973281860352,
+      "step": 3996
+    },
+    {
+      "epoch": 1.7249190938511327,
+      "grad_norm": 1.1145644187927246,
+      "learning_rate": 8.797940635335382e-05,
+      "loss": 0.6270571351051331,
+      "step": 3998
+    },
+    {
+      "epoch": 1.7257820927723841,
+      "grad_norm": 1.0673385858535767,
+      "learning_rate": 8.796343221105949e-05,
+      "loss": 0.6463025212287903,
+      "step": 4000
+    },
+    {
+      "epoch": 1.7257820927723841,
+      "eval_loss": 0.6941319704055786,
+      "eval_runtime": 498.1423,
+      "eval_samples_per_second": 4.135,
+      "eval_steps_per_second": 4.135,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 1
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.170258107248804e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4000/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-4000/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4000/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/README.md b/sft_devstral_24B/checkpoints/checkpoint-4500/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-4500/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-4500/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..0f65b8c1e5dc3454dbcf69498c2037d670c73fe7
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2525c1928ca9de033f220a5e840c681a9d5f8ed45f23f11db8475248b70b962f
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-4500/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..7ce85fbceb8e800b02c440c7f232442d5a1069ff
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82399ac703b1022eccdfb3728089df0517cf814258fbe611daef7803a4a744b9
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-4500/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2baec527fe94ff791b956e4442f2fcbf4669cd37
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c97cbe0aa8fb0b300a4e0e6859ccc69916ea2336815abf25f3f0b056a807e08e
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-4500/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..aee8e5ff3686c9fe926af02a83ee2fa40301fcd1
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00e405200a08b50899317a837c90c8d382b890f73f59f1b37fbc0c69c6c7ef01
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-4500/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..9032d844dcda9bf46cbbb4a627f38ce3923301cd
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/trainer_state.json
@@ -0,0 +1,16153 @@
+{
+  "best_global_step": 4500,
+  "best_metric": 0.6851214170455933,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-4500",
+  "epoch": 1.9415318230852212,
+  "eval_steps": 100,
+  "global_step": 4500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    },
+    {
+      "epoch": 0.21661272923408845,
+      "grad_norm": 1.8110703229904175,
+      "learning_rate": 4.501347708894879e-05,
+      "loss": 0.9099015593528748,
+      "step": 502
+    },
+    {
+      "epoch": 0.2174757281553398,
+      "grad_norm": 2.0977623462677,
+      "learning_rate": 4.519317160826595e-05,
+      "loss": 0.875080943107605,
+      "step": 504
+    },
+    {
+      "epoch": 0.21833872707659116,
+      "grad_norm": 1.8648077249526978,
+      "learning_rate": 4.537286612758311e-05,
+      "loss": 0.8511829376220703,
+      "step": 506
+    },
+    {
+      "epoch": 0.2192017259978425,
+      "grad_norm": 1.754744291305542,
+      "learning_rate": 4.555256064690027e-05,
+      "loss": 0.8408182859420776,
+      "step": 508
+    },
+    {
+      "epoch": 0.22006472491909385,
+      "grad_norm": 1.9869136810302734,
+      "learning_rate": 4.573225516621743e-05,
+      "loss": 0.8824991583824158,
+      "step": 510
+    },
+    {
+      "epoch": 0.2209277238403452,
+      "grad_norm": 2.4601082801818848,
+      "learning_rate": 4.5911949685534594e-05,
+      "loss": 0.9949447512626648,
+      "step": 512
+    },
+    {
+      "epoch": 0.22179072276159656,
+      "grad_norm": 2.1224238872528076,
+      "learning_rate": 4.609164420485175e-05,
+      "loss": 0.8979368805885315,
+      "step": 514
+    },
+    {
+      "epoch": 0.2226537216828479,
+      "grad_norm": 1.6812392473220825,
+      "learning_rate": 4.627133872416892e-05,
+      "loss": 0.8728249669075012,
+      "step": 516
+    },
+    {
+      "epoch": 0.22351672060409924,
+      "grad_norm": 1.6505497694015503,
+      "learning_rate": 4.6451033243486075e-05,
+      "loss": 0.8083430528640747,
+      "step": 518
+    },
+    {
+      "epoch": 0.2243797195253506,
+      "grad_norm": 1.6454377174377441,
+      "learning_rate": 4.663072776280324e-05,
+      "loss": 0.8042294383049011,
+      "step": 520
+    },
+    {
+      "epoch": 0.22524271844660193,
+      "grad_norm": 2.0974535942077637,
+      "learning_rate": 4.68104222821204e-05,
+      "loss": 0.9263730049133301,
+      "step": 522
+    },
+    {
+      "epoch": 0.2261057173678533,
+      "grad_norm": 1.90152108669281,
+      "learning_rate": 4.6990116801437556e-05,
+      "loss": 0.9090206623077393,
+      "step": 524
+    },
+    {
+      "epoch": 0.22696871628910464,
+      "grad_norm": 1.9843928813934326,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.8310918807983398,
+      "step": 526
+    },
+    {
+      "epoch": 0.227831715210356,
+      "grad_norm": 1.7880226373672485,
+      "learning_rate": 4.734950584007188e-05,
+      "loss": 0.8838925957679749,
+      "step": 528
+    },
+    {
+      "epoch": 0.22869471413160733,
+      "grad_norm": 1.8534979820251465,
+      "learning_rate": 4.7529200359389045e-05,
+      "loss": 0.9431065917015076,
+      "step": 530
+    },
+    {
+      "epoch": 0.2295577130528587,
+      "grad_norm": 1.8936293125152588,
+      "learning_rate": 4.77088948787062e-05,
+      "loss": 0.7988291382789612,
+      "step": 532
+    },
+    {
+      "epoch": 0.23042071197411004,
+      "grad_norm": 1.919683814048767,
+      "learning_rate": 4.788858939802337e-05,
+      "loss": 0.8477044105529785,
+      "step": 534
+    },
+    {
+      "epoch": 0.23128371089536137,
+      "grad_norm": 1.8022494316101074,
+      "learning_rate": 4.8068283917340526e-05,
+      "loss": 0.8482418060302734,
+      "step": 536
+    },
+    {
+      "epoch": 0.23214670981661273,
+      "grad_norm": 1.8628984689712524,
+      "learning_rate": 4.8247978436657684e-05,
+      "loss": 0.8553584814071655,
+      "step": 538
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.8644732236862183,
+      "learning_rate": 4.842767295597484e-05,
+      "loss": 0.8287713527679443,
+      "step": 540
+    },
+    {
+      "epoch": 0.23387270765911541,
+      "grad_norm": 2.162926435470581,
+      "learning_rate": 4.8607367475292e-05,
+      "loss": 0.8984123468399048,
+      "step": 542
+    },
+    {
+      "epoch": 0.23473570658036677,
+      "grad_norm": 1.453223466873169,
+      "learning_rate": 4.8787061994609166e-05,
+      "loss": 0.725272536277771,
+      "step": 544
+    },
+    {
+      "epoch": 0.23559870550161813,
+      "grad_norm": 2.2145543098449707,
+      "learning_rate": 4.8966756513926324e-05,
+      "loss": 0.9128344655036926,
+      "step": 546
+    },
+    {
+      "epoch": 0.23646170442286948,
+      "grad_norm": 1.7639187574386597,
+      "learning_rate": 4.914645103324349e-05,
+      "loss": 0.782098114490509,
+      "step": 548
+    },
+    {
+      "epoch": 0.2373247033441208,
+      "grad_norm": 1.6597305536270142,
+      "learning_rate": 4.932614555256065e-05,
+      "loss": 0.8212350010871887,
+      "step": 550
+    },
+    {
+      "epoch": 0.23818770226537217,
+      "grad_norm": 1.9441677331924438,
+      "learning_rate": 4.950584007187781e-05,
+      "loss": 0.9033122062683105,
+      "step": 552
+    },
+    {
+      "epoch": 0.23905070118662353,
+      "grad_norm": 2.0332281589508057,
+      "learning_rate": 4.968553459119497e-05,
+      "loss": 0.9195884466171265,
+      "step": 554
+    },
+    {
+      "epoch": 0.23991370010787486,
+      "grad_norm": 1.9222822189331055,
+      "learning_rate": 4.986522911051213e-05,
+      "loss": 0.8828456401824951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2407766990291262,
+      "grad_norm": 1.9830708503723145,
+      "learning_rate": 5.004492362982929e-05,
+      "loss": 0.9089829921722412,
+      "step": 558
+    },
+    {
+      "epoch": 0.24163969795037757,
+      "grad_norm": 1.99044930934906,
+      "learning_rate": 5.022461814914645e-05,
+      "loss": 0.8747812509536743,
+      "step": 560
+    },
+    {
+      "epoch": 0.2425026968716289,
+      "grad_norm": 1.7961562871932983,
+      "learning_rate": 5.040431266846362e-05,
+      "loss": 0.8421230912208557,
+      "step": 562
+    },
+    {
+      "epoch": 0.24336569579288025,
+      "grad_norm": 1.9798882007598877,
+      "learning_rate": 5.058400718778078e-05,
+      "loss": 0.8669922351837158,
+      "step": 564
+    },
+    {
+      "epoch": 0.2442286947141316,
+      "grad_norm": 2.0138750076293945,
+      "learning_rate": 5.076370170709793e-05,
+      "loss": 0.8596820831298828,
+      "step": 566
+    },
+    {
+      "epoch": 0.24509169363538297,
+      "grad_norm": 2.069021701812744,
+      "learning_rate": 5.09433962264151e-05,
+      "loss": 0.8288601636886597,
+      "step": 568
+    },
+    {
+      "epoch": 0.2459546925566343,
+      "grad_norm": 2.6913585662841797,
+      "learning_rate": 5.112309074573226e-05,
+      "loss": 0.8321394324302673,
+      "step": 570
+    },
+    {
+      "epoch": 0.24681769147788565,
+      "grad_norm": 1.8057955503463745,
+      "learning_rate": 5.1302785265049415e-05,
+      "loss": 0.842660129070282,
+      "step": 572
+    },
+    {
+      "epoch": 0.247680690399137,
+      "grad_norm": 1.5741548538208008,
+      "learning_rate": 5.148247978436658e-05,
+      "loss": 0.8431437015533447,
+      "step": 574
+    },
+    {
+      "epoch": 0.24854368932038834,
+      "grad_norm": 2.0329413414001465,
+      "learning_rate": 5.166217430368374e-05,
+      "loss": 0.9046191573143005,
+      "step": 576
+    },
+    {
+      "epoch": 0.2494066882416397,
+      "grad_norm": 1.7638250589370728,
+      "learning_rate": 5.18418688230009e-05,
+      "loss": 0.8108726739883423,
+      "step": 578
+    },
+    {
+      "epoch": 0.25026968716289105,
+      "grad_norm": 2.1088576316833496,
+      "learning_rate": 5.202156334231806e-05,
+      "loss": 0.8621282577514648,
+      "step": 580
+    },
+    {
+      "epoch": 0.2511326860841424,
+      "grad_norm": 2.072145462036133,
+      "learning_rate": 5.220125786163522e-05,
+      "loss": 0.8774182200431824,
+      "step": 582
+    },
+    {
+      "epoch": 0.25199568500539377,
+      "grad_norm": 1.8093693256378174,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.863322377204895,
+      "step": 584
+    },
+    {
+      "epoch": 0.25285868392664507,
+      "grad_norm": 1.7119505405426025,
+      "learning_rate": 5.2560646900269536e-05,
+      "loss": 0.8493514657020569,
+      "step": 586
+    },
+    {
+      "epoch": 0.2537216828478964,
+      "grad_norm": 2.042052745819092,
+      "learning_rate": 5.27403414195867e-05,
+      "loss": 0.8278731107711792,
+      "step": 588
+    },
+    {
+      "epoch": 0.2545846817691478,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 5.2920035938903866e-05,
+      "loss": 0.8388869166374207,
+      "step": 590
+    },
+    {
+      "epoch": 0.25544768069039914,
+      "grad_norm": 1.8072630167007446,
+      "learning_rate": 5.309973045822103e-05,
+      "loss": 0.8666917085647583,
+      "step": 592
+    },
+    {
+      "epoch": 0.2563106796116505,
+      "grad_norm": 1.7099417448043823,
+      "learning_rate": 5.327942497753818e-05,
+      "loss": 0.8114299774169922,
+      "step": 594
+    },
+    {
+      "epoch": 0.25717367853290185,
+      "grad_norm": 1.8132266998291016,
+      "learning_rate": 5.345911949685535e-05,
+      "loss": 0.8842664361000061,
+      "step": 596
+    },
+    {
+      "epoch": 0.2580366774541532,
+      "grad_norm": 1.9578651189804077,
+      "learning_rate": 5.363881401617251e-05,
+      "loss": 0.7954385876655579,
+      "step": 598
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 1.832077145576477,
+      "learning_rate": 5.381850853548968e-05,
+      "loss": 0.912541925907135,
+      "step": 600
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "eval_loss": 0.8770867586135864,
+      "eval_runtime": 827.8893,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 600
+    },
+    {
+      "epoch": 0.25976267529665586,
+      "grad_norm": 1.7565301656723022,
+      "learning_rate": 5.399820305480683e-05,
+      "loss": 0.8436558246612549,
+      "step": 602
+    },
+    {
+      "epoch": 0.2606256742179072,
+      "grad_norm": 1.9199552536010742,
+      "learning_rate": 5.4177897574123994e-05,
+      "loss": 0.9384433031082153,
+      "step": 604
+    },
+    {
+      "epoch": 0.2614886731391586,
+      "grad_norm": 1.7609341144561768,
+      "learning_rate": 5.435759209344116e-05,
+      "loss": 0.8695753812789917,
+      "step": 606
+    },
+    {
+      "epoch": 0.26235167206040994,
+      "grad_norm": 1.771958589553833,
+      "learning_rate": 5.453728661275831e-05,
+      "loss": 0.8462895154953003,
+      "step": 608
+    },
+    {
+      "epoch": 0.2632146709816613,
+      "grad_norm": 1.8958079814910889,
+      "learning_rate": 5.4716981132075475e-05,
+      "loss": 0.9074611663818359,
+      "step": 610
+    },
+    {
+      "epoch": 0.26407766990291265,
+      "grad_norm": 1.8668403625488281,
+      "learning_rate": 5.489667565139264e-05,
+      "loss": 0.8603307008743286,
+      "step": 612
+    },
+    {
+      "epoch": 0.26494066882416395,
+      "grad_norm": 1.8902733325958252,
+      "learning_rate": 5.50763701707098e-05,
+      "loss": 0.8925216794013977,
+      "step": 614
+    },
+    {
+      "epoch": 0.2658036677454153,
+      "grad_norm": 1.7172386646270752,
+      "learning_rate": 5.525606469002696e-05,
+      "loss": 0.7872517704963684,
+      "step": 616
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.7814518213272095,
+      "learning_rate": 5.5435759209344115e-05,
+      "loss": 0.7951212525367737,
+      "step": 618
+    },
+    {
+      "epoch": 0.267529665587918,
+      "grad_norm": 1.8714463710784912,
+      "learning_rate": 5.561545372866128e-05,
+      "loss": 0.859940230846405,
+      "step": 620
+    },
+    {
+      "epoch": 0.2683926645091694,
+      "grad_norm": 1.9614174365997314,
+      "learning_rate": 5.579514824797844e-05,
+      "loss": 0.8953307867050171,
+      "step": 622
+    },
+    {
+      "epoch": 0.26925566343042073,
+      "grad_norm": 1.7190991640090942,
+      "learning_rate": 5.5974842767295596e-05,
+      "loss": 0.8346402645111084,
+      "step": 624
+    },
+    {
+      "epoch": 0.27011866235167203,
+      "grad_norm": 1.981070876121521,
+      "learning_rate": 5.615453728661276e-05,
+      "loss": 0.8545536398887634,
+      "step": 626
+    },
+    {
+      "epoch": 0.2709816612729234,
+      "grad_norm": 1.9186289310455322,
+      "learning_rate": 5.6334231805929926e-05,
+      "loss": 0.9064128398895264,
+      "step": 628
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.9937374591827393,
+      "learning_rate": 5.651392632524708e-05,
+      "loss": 0.883073627948761,
+      "step": 630
+    },
+    {
+      "epoch": 0.2727076591154261,
+      "grad_norm": 1.800236701965332,
+      "learning_rate": 5.669362084456424e-05,
+      "loss": 0.8777942657470703,
+      "step": 632
+    },
+    {
+      "epoch": 0.27357065803667746,
+      "grad_norm": 1.9164050817489624,
+      "learning_rate": 5.687331536388141e-05,
+      "loss": 0.8527862429618835,
+      "step": 634
+    },
+    {
+      "epoch": 0.2744336569579288,
+      "grad_norm": 1.7494043111801147,
+      "learning_rate": 5.705300988319856e-05,
+      "loss": 0.881090521812439,
+      "step": 636
+    },
+    {
+      "epoch": 0.2752966558791802,
+      "grad_norm": 1.9433623552322388,
+      "learning_rate": 5.7232704402515724e-05,
+      "loss": 0.8547250628471375,
+      "step": 638
+    },
+    {
+      "epoch": 0.2761596548004315,
+      "grad_norm": 1.7539145946502686,
+      "learning_rate": 5.741239892183289e-05,
+      "loss": 0.8429408073425293,
+      "step": 640
+    },
+    {
+      "epoch": 0.27702265372168283,
+      "grad_norm": 1.904911756515503,
+      "learning_rate": 5.7592093441150054e-05,
+      "loss": 0.8587079644203186,
+      "step": 642
+    },
+    {
+      "epoch": 0.2778856526429342,
+      "grad_norm": 1.5534838438034058,
+      "learning_rate": 5.7771787960467206e-05,
+      "loss": 0.7795872688293457,
+      "step": 644
+    },
+    {
+      "epoch": 0.27874865156418555,
+      "grad_norm": 1.704798698425293,
+      "learning_rate": 5.795148247978437e-05,
+      "loss": 0.7657744884490967,
+      "step": 646
+    },
+    {
+      "epoch": 0.2796116504854369,
+      "grad_norm": 1.8432048559188843,
+      "learning_rate": 5.8131176999101536e-05,
+      "loss": 0.8119702935218811,
+      "step": 648
+    },
+    {
+      "epoch": 0.28047464940668826,
+      "grad_norm": 1.9495919942855835,
+      "learning_rate": 5.831087151841869e-05,
+      "loss": 0.8846080899238586,
+      "step": 650
+    },
+    {
+      "epoch": 0.2813376483279396,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 5.849056603773585e-05,
+      "loss": 0.854324460029602,
+      "step": 652
+    },
+    {
+      "epoch": 0.2822006472491909,
+      "grad_norm": 1.767221212387085,
+      "learning_rate": 5.867026055705302e-05,
+      "loss": 0.7908785939216614,
+      "step": 654
+    },
+    {
+      "epoch": 0.2830636461704423,
+      "grad_norm": 1.8866491317749023,
+      "learning_rate": 5.8849955076370175e-05,
+      "loss": 0.8155530691146851,
+      "step": 656
+    },
+    {
+      "epoch": 0.28392664509169363,
+      "grad_norm": 1.9543417692184448,
+      "learning_rate": 5.9029649595687334e-05,
+      "loss": 0.9833307266235352,
+      "step": 658
+    },
+    {
+      "epoch": 0.284789644012945,
+      "grad_norm": 1.867039680480957,
+      "learning_rate": 5.920934411500449e-05,
+      "loss": 0.7991418242454529,
+      "step": 660
+    },
+    {
+      "epoch": 0.28565264293419634,
+      "grad_norm": 1.8487379550933838,
+      "learning_rate": 5.938903863432166e-05,
+      "loss": 0.8232532143592834,
+      "step": 662
+    },
+    {
+      "epoch": 0.2865156418554477,
+      "grad_norm": 1.5015742778778076,
+      "learning_rate": 5.956873315363881e-05,
+      "loss": 0.76528000831604,
+      "step": 664
+    },
+    {
+      "epoch": 0.287378640776699,
+      "grad_norm": 2.726778268814087,
+      "learning_rate": 5.974842767295597e-05,
+      "loss": 0.8220192790031433,
+      "step": 666
+    },
+    {
+      "epoch": 0.28824163969795036,
+      "grad_norm": 1.6754003763198853,
+      "learning_rate": 5.992812219227314e-05,
+      "loss": 0.7534968256950378,
+      "step": 668
+    },
+    {
+      "epoch": 0.2891046386192017,
+      "grad_norm": 1.9948434829711914,
+      "learning_rate": 6.01078167115903e-05,
+      "loss": 0.8405768871307373,
+      "step": 670
+    },
+    {
+      "epoch": 0.28996763754045307,
+      "grad_norm": 1.7982401847839355,
+      "learning_rate": 6.0287511230907455e-05,
+      "loss": 0.8143276572227478,
+      "step": 672
+    },
+    {
+      "epoch": 0.29083063646170443,
+      "grad_norm": 1.7311779260635376,
+      "learning_rate": 6.046720575022462e-05,
+      "loss": 0.9602267742156982,
+      "step": 674
+    },
+    {
+      "epoch": 0.2916936353829558,
+      "grad_norm": 1.9906021356582642,
+      "learning_rate": 6.0646900269541785e-05,
+      "loss": 0.8628544807434082,
+      "step": 676
+    },
+    {
+      "epoch": 0.29255663430420714,
+      "grad_norm": 1.7477494478225708,
+      "learning_rate": 6.082659478885895e-05,
+      "loss": 0.7800343632698059,
+      "step": 678
+    },
+    {
+      "epoch": 0.29341963322545844,
+      "grad_norm": 1.7978737354278564,
+      "learning_rate": 6.10062893081761e-05,
+      "loss": 0.8722298741340637,
+      "step": 680
+    },
+    {
+      "epoch": 0.2942826321467098,
+      "grad_norm": 2.1463348865509033,
+      "learning_rate": 6.118598382749326e-05,
+      "loss": 0.8895675539970398,
+      "step": 682
+    },
+    {
+      "epoch": 0.29514563106796116,
+      "grad_norm": 1.9256783723831177,
+      "learning_rate": 6.136567834681042e-05,
+      "loss": 0.8579491972923279,
+      "step": 684
+    },
+    {
+      "epoch": 0.2960086299892125,
+      "grad_norm": 1.4459543228149414,
+      "learning_rate": 6.154537286612758e-05,
+      "loss": 0.7811232805252075,
+      "step": 686
+    },
+    {
+      "epoch": 0.29687162891046387,
+      "grad_norm": 1.6202397346496582,
+      "learning_rate": 6.172506738544474e-05,
+      "loss": 0.7827678322792053,
+      "step": 688
+    },
+    {
+      "epoch": 0.2977346278317152,
+      "grad_norm": 2.0524775981903076,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.8186459541320801,
+      "step": 690
+    },
+    {
+      "epoch": 0.2985976267529666,
+      "grad_norm": 1.7867428064346313,
+      "learning_rate": 6.208445642407907e-05,
+      "loss": 0.8691027760505676,
+      "step": 692
+    },
+    {
+      "epoch": 0.2994606256742179,
+      "grad_norm": 1.9605737924575806,
+      "learning_rate": 6.226415094339622e-05,
+      "loss": 0.8233416080474854,
+      "step": 694
+    },
+    {
+      "epoch": 0.30032362459546924,
+      "grad_norm": 1.7035984992980957,
+      "learning_rate": 6.244384546271339e-05,
+      "loss": 0.8493421077728271,
+      "step": 696
+    },
+    {
+      "epoch": 0.3011866235167206,
+      "grad_norm": 1.9402128458023071,
+      "learning_rate": 6.262353998203055e-05,
+      "loss": 0.8114250302314758,
+      "step": 698
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "grad_norm": 1.8533520698547363,
+      "learning_rate": 6.28032345013477e-05,
+      "loss": 0.8996106386184692,
+      "step": 700
+    },
+    {
+      "epoch": 0.30204962243797195,
+      "eval_loss": 0.8631294369697571,
+      "eval_runtime": 839.6269,
+      "eval_samples_per_second": 2.453,
+      "eval_steps_per_second": 2.453,
+      "step": 700
+    },
+    {
+      "epoch": 0.3029126213592233,
+      "grad_norm": 1.727793574333191,
+      "learning_rate": 6.298292902066487e-05,
+      "loss": 0.7978370189666748,
+      "step": 702
+    },
+    {
+      "epoch": 0.30377562028047467,
+      "grad_norm": 1.9601492881774902,
+      "learning_rate": 6.316262353998203e-05,
+      "loss": 0.8120381236076355,
+      "step": 704
+    },
+    {
+      "epoch": 0.304638619201726,
+      "grad_norm": 1.8380948305130005,
+      "learning_rate": 6.33423180592992e-05,
+      "loss": 0.8005605340003967,
+      "step": 706
+    },
+    {
+      "epoch": 0.3055016181229773,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 6.352201257861635e-05,
+      "loss": 0.8967245817184448,
+      "step": 708
+    },
+    {
+      "epoch": 0.3063646170442287,
+      "grad_norm": 3.5359878540039062,
+      "learning_rate": 6.370170709793352e-05,
+      "loss": 0.8271524906158447,
+      "step": 710
+    },
+    {
+      "epoch": 0.30722761596548004,
+      "grad_norm": 1.7943733930587769,
+      "learning_rate": 6.388140161725068e-05,
+      "loss": 0.8759040236473083,
+      "step": 712
+    },
+    {
+      "epoch": 0.3080906148867314,
+      "grad_norm": 1.5483570098876953,
+      "learning_rate": 6.406109613656783e-05,
+      "loss": 0.7872476577758789,
+      "step": 714
+    },
+    {
+      "epoch": 0.30895361380798275,
+      "grad_norm": 1.8044816255569458,
+      "learning_rate": 6.4240790655885e-05,
+      "loss": 0.849986732006073,
+      "step": 716
+    },
+    {
+      "epoch": 0.3098166127292341,
+      "grad_norm": 1.548281192779541,
+      "learning_rate": 6.442048517520216e-05,
+      "loss": 0.8292878866195679,
+      "step": 718
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.643020749092102,
+      "learning_rate": 6.460017969451933e-05,
+      "loss": 0.8796502351760864,
+      "step": 720
+    },
+    {
+      "epoch": 0.31154261057173677,
+      "grad_norm": 1.633881688117981,
+      "learning_rate": 6.477987421383648e-05,
+      "loss": 0.806233286857605,
+      "step": 722
+    },
+    {
+      "epoch": 0.3124056094929881,
+      "grad_norm": 1.7712864875793457,
+      "learning_rate": 6.495956873315364e-05,
+      "loss": 0.802722156047821,
+      "step": 724
+    },
+    {
+      "epoch": 0.3132686084142395,
+      "grad_norm": 1.9158337116241455,
+      "learning_rate": 6.513926325247081e-05,
+      "loss": 0.8735752105712891,
+      "step": 726
+    },
+    {
+      "epoch": 0.31413160733549084,
+      "grad_norm": 1.8648416996002197,
+      "learning_rate": 6.531895777178796e-05,
+      "loss": 0.793403148651123,
+      "step": 728
+    },
+    {
+      "epoch": 0.3149946062567422,
+      "grad_norm": 1.5917021036148071,
+      "learning_rate": 6.549865229110512e-05,
+      "loss": 0.7772542834281921,
+      "step": 730
+    },
+    {
+      "epoch": 0.31585760517799355,
+      "grad_norm": 1.8560330867767334,
+      "learning_rate": 6.567834681042229e-05,
+      "loss": 0.8731698989868164,
+      "step": 732
+    },
+    {
+      "epoch": 0.31672060409924485,
+      "grad_norm": 1.6038225889205933,
+      "learning_rate": 6.585804132973945e-05,
+      "loss": 0.8167505264282227,
+      "step": 734
+    },
+    {
+      "epoch": 0.3175836030204962,
+      "grad_norm": 1.6221873760223389,
+      "learning_rate": 6.60377358490566e-05,
+      "loss": 0.7810112237930298,
+      "step": 736
+    },
+    {
+      "epoch": 0.31844660194174756,
+      "grad_norm": 1.8528746366500854,
+      "learning_rate": 6.621743036837377e-05,
+      "loss": 0.8735228776931763,
+      "step": 738
+    },
+    {
+      "epoch": 0.3193096008629989,
+      "grad_norm": 1.473233938217163,
+      "learning_rate": 6.639712488769094e-05,
+      "loss": 0.8231552243232727,
+      "step": 740
+    },
+    {
+      "epoch": 0.3201725997842503,
+      "grad_norm": 1.7726222276687622,
+      "learning_rate": 6.657681940700809e-05,
+      "loss": 0.8400886058807373,
+      "step": 742
+    },
+    {
+      "epoch": 0.32103559870550163,
+      "grad_norm": 1.7245025634765625,
+      "learning_rate": 6.675651392632525e-05,
+      "loss": 0.7371112108230591,
+      "step": 744
+    },
+    {
+      "epoch": 0.321898597626753,
+      "grad_norm": 1.6225395202636719,
+      "learning_rate": 6.693620844564242e-05,
+      "loss": 0.8057155013084412,
+      "step": 746
+    },
+    {
+      "epoch": 0.3227615965480043,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 6.711590296495958e-05,
+      "loss": 0.8471131324768066,
+      "step": 748
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 1.760038137435913,
+      "learning_rate": 6.729559748427673e-05,
+      "loss": 0.8251315951347351,
+      "step": 750
+    },
+    {
+      "epoch": 0.324487594390507,
+      "grad_norm": 1.8012818098068237,
+      "learning_rate": 6.74752920035939e-05,
+      "loss": 0.8215667605400085,
+      "step": 752
+    },
+    {
+      "epoch": 0.32535059331175836,
+      "grad_norm": 1.769614815711975,
+      "learning_rate": 6.765498652291105e-05,
+      "loss": 0.8113123178482056,
+      "step": 754
+    },
+    {
+      "epoch": 0.3262135922330097,
+      "grad_norm": 1.7554916143417358,
+      "learning_rate": 6.783468104222822e-05,
+      "loss": 0.8893613219261169,
+      "step": 756
+    },
+    {
+      "epoch": 0.3270765911542611,
+      "grad_norm": 1.5720254182815552,
+      "learning_rate": 6.801437556154538e-05,
+      "loss": 0.8165010809898376,
+      "step": 758
+    },
+    {
+      "epoch": 0.32793959007551243,
+      "grad_norm": 2.1887543201446533,
+      "learning_rate": 6.819407008086253e-05,
+      "loss": 0.8800036311149597,
+      "step": 760
+    },
+    {
+      "epoch": 0.32880258899676373,
+      "grad_norm": 1.63350510597229,
+      "learning_rate": 6.83737646001797e-05,
+      "loss": 0.7268908619880676,
+      "step": 762
+    },
+    {
+      "epoch": 0.3296655879180151,
+      "grad_norm": 1.5838189125061035,
+      "learning_rate": 6.855345911949685e-05,
+      "loss": 0.7695913910865784,
+      "step": 764
+    },
+    {
+      "epoch": 0.33052858683926645,
+      "grad_norm": 1.8752859830856323,
+      "learning_rate": 6.873315363881401e-05,
+      "loss": 0.9005667567253113,
+      "step": 766
+    },
+    {
+      "epoch": 0.3313915857605178,
+      "grad_norm": 1.4201446771621704,
+      "learning_rate": 6.891284815813118e-05,
+      "loss": 0.7960113883018494,
+      "step": 768
+    },
+    {
+      "epoch": 0.33225458468176916,
+      "grad_norm": 1.6665233373641968,
+      "learning_rate": 6.909254267744834e-05,
+      "loss": 0.851472020149231,
+      "step": 770
+    },
+    {
+      "epoch": 0.3331175836030205,
+      "grad_norm": 1.72721266746521,
+      "learning_rate": 6.92722371967655e-05,
+      "loss": 0.8403398990631104,
+      "step": 772
+    },
+    {
+      "epoch": 0.3339805825242718,
+      "grad_norm": 1.648680567741394,
+      "learning_rate": 6.945193171608266e-05,
+      "loss": 0.7582393884658813,
+      "step": 774
+    },
+    {
+      "epoch": 0.3348435814455232,
+      "grad_norm": 1.89604651927948,
+      "learning_rate": 6.963162623539982e-05,
+      "loss": 0.8250385522842407,
+      "step": 776
+    },
+    {
+      "epoch": 0.33570658036677453,
+      "grad_norm": 1.4748259782791138,
+      "learning_rate": 6.981132075471698e-05,
+      "loss": 0.7916358709335327,
+      "step": 778
+    },
+    {
+      "epoch": 0.3365695792880259,
+      "grad_norm": 1.4885735511779785,
+      "learning_rate": 6.999101527403414e-05,
+      "loss": 0.7432135939598083,
+      "step": 780
+    },
+    {
+      "epoch": 0.33743257820927725,
+      "grad_norm": 1.5150227546691895,
+      "learning_rate": 7.01707097933513e-05,
+      "loss": 0.7369741201400757,
+      "step": 782
+    },
+    {
+      "epoch": 0.3382955771305286,
+      "grad_norm": 2.0702431201934814,
+      "learning_rate": 7.035040431266847e-05,
+      "loss": 0.8347548246383667,
+      "step": 784
+    },
+    {
+      "epoch": 0.33915857605177996,
+      "grad_norm": 1.8330671787261963,
+      "learning_rate": 7.053009883198562e-05,
+      "loss": 0.7757654190063477,
+      "step": 786
+    },
+    {
+      "epoch": 0.34002157497303126,
+      "grad_norm": 1.6809033155441284,
+      "learning_rate": 7.070979335130279e-05,
+      "loss": 0.7966224551200867,
+      "step": 788
+    },
+    {
+      "epoch": 0.3408845738942826,
+      "grad_norm": 1.5811399221420288,
+      "learning_rate": 7.088948787061995e-05,
+      "loss": 0.7977543473243713,
+      "step": 790
+    },
+    {
+      "epoch": 0.341747572815534,
+      "grad_norm": 1.6572248935699463,
+      "learning_rate": 7.10691823899371e-05,
+      "loss": 0.7842527627944946,
+      "step": 792
+    },
+    {
+      "epoch": 0.34261057173678533,
+      "grad_norm": 1.6138155460357666,
+      "learning_rate": 7.124887690925427e-05,
+      "loss": 0.7586495876312256,
+      "step": 794
+    },
+    {
+      "epoch": 0.3434735706580367,
+      "grad_norm": 1.9198144674301147,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.9010837078094482,
+      "step": 796
+    },
+    {
+      "epoch": 0.34433656957928804,
+      "grad_norm": 1.6212328672409058,
+      "learning_rate": 7.16082659478886e-05,
+      "loss": 0.7857459783554077,
+      "step": 798
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "grad_norm": 2.1088991165161133,
+      "learning_rate": 7.178796046720575e-05,
+      "loss": 0.9087778925895691,
+      "step": 800
+    },
+    {
+      "epoch": 0.3451995685005394,
+      "eval_loss": 0.8476243615150452,
+      "eval_runtime": 831.6735,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 800
+    },
+    {
+      "epoch": 0.3460625674217907,
+      "grad_norm": 1.6325052976608276,
+      "learning_rate": 7.196765498652292e-05,
+      "loss": 0.7561838626861572,
+      "step": 802
+    },
+    {
+      "epoch": 0.34692556634304206,
+      "grad_norm": 1.8044698238372803,
+      "learning_rate": 7.214734950584008e-05,
+      "loss": 0.7940152883529663,
+      "step": 804
+    },
+    {
+      "epoch": 0.3477885652642934,
+      "grad_norm": 1.4942690134048462,
+      "learning_rate": 7.232704402515723e-05,
+      "loss": 0.8002730011940002,
+      "step": 806
+    },
+    {
+      "epoch": 0.34865156418554477,
+      "grad_norm": 1.5209156274795532,
+      "learning_rate": 7.25067385444744e-05,
+      "loss": 0.8211291432380676,
+      "step": 808
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.662622094154358,
+      "learning_rate": 7.268643306379156e-05,
+      "loss": 0.8690080642700195,
+      "step": 810
+    },
+    {
+      "epoch": 0.3503775620280475,
+      "grad_norm": 1.4452823400497437,
+      "learning_rate": 7.286612758310873e-05,
+      "loss": 0.8259872794151306,
+      "step": 812
+    },
+    {
+      "epoch": 0.3512405609492988,
+      "grad_norm": 1.6030809879302979,
+      "learning_rate": 7.304582210242588e-05,
+      "loss": 0.7973082661628723,
+      "step": 814
+    },
+    {
+      "epoch": 0.35210355987055014,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 7.322551662174304e-05,
+      "loss": 0.8258423805236816,
+      "step": 816
+    },
+    {
+      "epoch": 0.3529665587918015,
+      "grad_norm": 1.6284228563308716,
+      "learning_rate": 7.340521114106021e-05,
+      "loss": 0.8570499420166016,
+      "step": 818
+    },
+    {
+      "epoch": 0.35382955771305286,
+      "grad_norm": 1.399397373199463,
+      "learning_rate": 7.358490566037736e-05,
+      "loss": 0.7808573842048645,
+      "step": 820
+    },
+    {
+      "epoch": 0.3546925566343042,
+      "grad_norm": 1.6233313083648682,
+      "learning_rate": 7.376460017969452e-05,
+      "loss": 0.7514002323150635,
+      "step": 822
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.1610822677612305,
+      "learning_rate": 7.394429469901169e-05,
+      "loss": 0.8392049670219421,
+      "step": 824
+    },
+    {
+      "epoch": 0.3564185544768069,
+      "grad_norm": 1.5170103311538696,
+      "learning_rate": 7.412398921832885e-05,
+      "loss": 0.7574125528335571,
+      "step": 826
+    },
+    {
+      "epoch": 0.3572815533980582,
+      "grad_norm": 1.687654733657837,
+      "learning_rate": 7.4303683737646e-05,
+      "loss": 0.8394677639007568,
+      "step": 828
+    },
+    {
+      "epoch": 0.3581445523193096,
+      "grad_norm": 1.6306912899017334,
+      "learning_rate": 7.448337825696317e-05,
+      "loss": 0.7760812044143677,
+      "step": 830
+    },
+    {
+      "epoch": 0.35900755124056094,
+      "grad_norm": 1.4824501276016235,
+      "learning_rate": 7.466307277628034e-05,
+      "loss": 0.8146010041236877,
+      "step": 832
+    },
+    {
+      "epoch": 0.3598705501618123,
+      "grad_norm": 1.962728500366211,
+      "learning_rate": 7.484276729559749e-05,
+      "loss": 0.7824002504348755,
+      "step": 834
+    },
+    {
+      "epoch": 0.36073354908306365,
+      "grad_norm": 1.805561900138855,
+      "learning_rate": 7.502246181491465e-05,
+      "loss": 0.8955144882202148,
+      "step": 836
+    },
+    {
+      "epoch": 0.361596548004315,
+      "grad_norm": 1.6058090925216675,
+      "learning_rate": 7.52021563342318e-05,
+      "loss": 0.7638000249862671,
+      "step": 838
+    },
+    {
+      "epoch": 0.36245954692556637,
+      "grad_norm": 1.5326330661773682,
+      "learning_rate": 7.538185085354897e-05,
+      "loss": 0.8465489745140076,
+      "step": 840
+    },
+    {
+      "epoch": 0.36332254584681767,
+      "grad_norm": 1.563953161239624,
+      "learning_rate": 7.556154537286612e-05,
+      "loss": 0.7809846997261047,
+      "step": 842
+    },
+    {
+      "epoch": 0.364185544768069,
+      "grad_norm": 1.3842798471450806,
+      "learning_rate": 7.574123989218329e-05,
+      "loss": 0.7431546449661255,
+      "step": 844
+    },
+    {
+      "epoch": 0.3650485436893204,
+      "grad_norm": 1.843582272529602,
+      "learning_rate": 7.592093441150045e-05,
+      "loss": 0.8984674215316772,
+      "step": 846
+    },
+    {
+      "epoch": 0.36591154261057174,
+      "grad_norm": 1.7077877521514893,
+      "learning_rate": 7.610062893081762e-05,
+      "loss": 0.7809584736824036,
+      "step": 848
+    },
+    {
+      "epoch": 0.3667745415318231,
+      "grad_norm": 1.3528809547424316,
+      "learning_rate": 7.628032345013477e-05,
+      "loss": 0.7576664090156555,
+      "step": 850
+    },
+    {
+      "epoch": 0.36763754045307445,
+      "grad_norm": 1.4391332864761353,
+      "learning_rate": 7.646001796945193e-05,
+      "loss": 0.7666721940040588,
+      "step": 852
+    },
+    {
+      "epoch": 0.3685005393743258,
+      "grad_norm": 1.626839518547058,
+      "learning_rate": 7.66397124887691e-05,
+      "loss": 0.8356342911720276,
+      "step": 854
+    },
+    {
+      "epoch": 0.3693635382955771,
+      "grad_norm": 1.5550813674926758,
+      "learning_rate": 7.681940700808625e-05,
+      "loss": 0.7766274809837341,
+      "step": 856
+    },
+    {
+      "epoch": 0.37022653721682847,
+      "grad_norm": 1.4684302806854248,
+      "learning_rate": 7.699910152740341e-05,
+      "loss": 0.7646660804748535,
+      "step": 858
+    },
+    {
+      "epoch": 0.3710895361380798,
+      "grad_norm": 1.6446272134780884,
+      "learning_rate": 7.717879604672058e-05,
+      "loss": 0.8716844916343689,
+      "step": 860
+    },
+    {
+      "epoch": 0.3719525350593312,
+      "grad_norm": 1.5995690822601318,
+      "learning_rate": 7.735849056603774e-05,
+      "loss": 0.8561879992485046,
+      "step": 862
+    },
+    {
+      "epoch": 0.37281553398058254,
+      "grad_norm": 1.632079839706421,
+      "learning_rate": 7.75381850853549e-05,
+      "loss": 0.8795325756072998,
+      "step": 864
+    },
+    {
+      "epoch": 0.3736785329018339,
+      "grad_norm": 1.7169907093048096,
+      "learning_rate": 7.771787960467206e-05,
+      "loss": 0.8779821991920471,
+      "step": 866
+    },
+    {
+      "epoch": 0.3745415318230852,
+      "grad_norm": 1.8590933084487915,
+      "learning_rate": 7.789757412398922e-05,
+      "loss": 0.8811644911766052,
+      "step": 868
+    },
+    {
+      "epoch": 0.37540453074433655,
+      "grad_norm": 1.804978370666504,
+      "learning_rate": 7.807726864330638e-05,
+      "loss": 0.8391726016998291,
+      "step": 870
+    },
+    {
+      "epoch": 0.3762675296655879,
+      "grad_norm": 2.2929575443267822,
+      "learning_rate": 7.825696316262354e-05,
+      "loss": 0.8425854444503784,
+      "step": 872
+    },
+    {
+      "epoch": 0.37713052858683926,
+      "grad_norm": 1.551804780960083,
+      "learning_rate": 7.84366576819407e-05,
+      "loss": 0.8409838080406189,
+      "step": 874
+    },
+    {
+      "epoch": 0.3779935275080906,
+      "grad_norm": 1.4655314683914185,
+      "learning_rate": 7.861635220125787e-05,
+      "loss": 0.7441553473472595,
+      "step": 876
+    },
+    {
+      "epoch": 0.378856526429342,
+      "grad_norm": 1.7886772155761719,
+      "learning_rate": 7.879604672057502e-05,
+      "loss": 0.8850297331809998,
+      "step": 878
+    },
+    {
+      "epoch": 0.37971952535059333,
+      "grad_norm": 1.4744676351547241,
+      "learning_rate": 7.897574123989219e-05,
+      "loss": 0.7991185188293457,
+      "step": 880
+    },
+    {
+      "epoch": 0.38058252427184464,
+      "grad_norm": 1.6844931840896606,
+      "learning_rate": 7.915543575920935e-05,
+      "loss": 0.7903867363929749,
+      "step": 882
+    },
+    {
+      "epoch": 0.381445523193096,
+      "grad_norm": 1.674704909324646,
+      "learning_rate": 7.93351302785265e-05,
+      "loss": 0.812190592288971,
+      "step": 884
+    },
+    {
+      "epoch": 0.38230852211434735,
+      "grad_norm": 1.6548097133636475,
+      "learning_rate": 7.951482479784367e-05,
+      "loss": 0.781878650188446,
+      "step": 886
+    },
+    {
+      "epoch": 0.3831715210355987,
+      "grad_norm": 1.3750375509262085,
+      "learning_rate": 7.969451931716083e-05,
+      "loss": 0.8223106861114502,
+      "step": 888
+    },
+    {
+      "epoch": 0.38403451995685006,
+      "grad_norm": 1.42178475856781,
+      "learning_rate": 7.9874213836478e-05,
+      "loss": 0.7170665860176086,
+      "step": 890
+    },
+    {
+      "epoch": 0.3848975188781014,
+      "grad_norm": 1.7126129865646362,
+      "learning_rate": 8.005390835579515e-05,
+      "loss": 0.8533666133880615,
+      "step": 892
+    },
+    {
+      "epoch": 0.3857605177993528,
+      "grad_norm": 1.552474856376648,
+      "learning_rate": 8.023360287511232e-05,
+      "loss": 0.8799889087677002,
+      "step": 894
+    },
+    {
+      "epoch": 0.3866235167206041,
+      "grad_norm": 1.2612507343292236,
+      "learning_rate": 8.041329739442948e-05,
+      "loss": 0.7968940734863281,
+      "step": 896
+    },
+    {
+      "epoch": 0.38748651564185543,
+      "grad_norm": 1.4770987033843994,
+      "learning_rate": 8.059299191374663e-05,
+      "loss": 0.7910935282707214,
+      "step": 898
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4640419483184814,
+      "learning_rate": 8.07726864330638e-05,
+      "loss": 0.7873474955558777,
+      "step": 900
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_loss": 0.8351938128471375,
+      "eval_runtime": 823.5649,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 900
+    },
+    {
+      "epoch": 0.38921251348435815,
+      "grad_norm": 1.3780814409255981,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.8183950185775757,
+      "step": 902
+    },
+    {
+      "epoch": 0.3900755124056095,
+      "grad_norm": 1.5895066261291504,
+      "learning_rate": 8.113207547169813e-05,
+      "loss": 0.786055326461792,
+      "step": 904
+    },
+    {
+      "epoch": 0.39093851132686086,
+      "grad_norm": 1.6340118646621704,
+      "learning_rate": 8.131176999101528e-05,
+      "loss": 0.8638336658477783,
+      "step": 906
+    },
+    {
+      "epoch": 0.3918015102481122,
+      "grad_norm": 1.3990486860275269,
+      "learning_rate": 8.149146451033244e-05,
+      "loss": 0.702790379524231,
+      "step": 908
+    },
+    {
+      "epoch": 0.3926645091693635,
+      "grad_norm": 1.3703010082244873,
+      "learning_rate": 8.167115902964961e-05,
+      "loss": 0.7845349907875061,
+      "step": 910
+    },
+    {
+      "epoch": 0.3935275080906149,
+      "grad_norm": 1.4194090366363525,
+      "learning_rate": 8.185085354896676e-05,
+      "loss": 0.8117155432701111,
+      "step": 912
+    },
+    {
+      "epoch": 0.39439050701186623,
+      "grad_norm": 1.4716848134994507,
+      "learning_rate": 8.203054806828392e-05,
+      "loss": 0.8229454755783081,
+      "step": 914
+    },
+    {
+      "epoch": 0.3952535059331176,
+      "grad_norm": 1.51328706741333,
+      "learning_rate": 8.221024258760108e-05,
+      "loss": 0.698765218257904,
+      "step": 916
+    },
+    {
+      "epoch": 0.39611650485436894,
+      "grad_norm": 1.4755173921585083,
+      "learning_rate": 8.238993710691824e-05,
+      "loss": 0.7694190144538879,
+      "step": 918
+    },
+    {
+      "epoch": 0.3969795037756203,
+      "grad_norm": 1.5302451848983765,
+      "learning_rate": 8.25696316262354e-05,
+      "loss": 0.779274046421051,
+      "step": 920
+    },
+    {
+      "epoch": 0.3978425026968716,
+      "grad_norm": 1.474968433380127,
+      "learning_rate": 8.274932614555256e-05,
+      "loss": 0.713699221611023,
+      "step": 922
+    },
+    {
+      "epoch": 0.39870550161812296,
+      "grad_norm": 1.4024375677108765,
+      "learning_rate": 8.292902066486972e-05,
+      "loss": 0.7869517803192139,
+      "step": 924
+    },
+    {
+      "epoch": 0.3995685005393743,
+      "grad_norm": 1.5729585886001587,
+      "learning_rate": 8.310871518418689e-05,
+      "loss": 0.9040259718894958,
+      "step": 926
+    },
+    {
+      "epoch": 0.4004314994606257,
+      "grad_norm": 1.609850525856018,
+      "learning_rate": 8.328840970350404e-05,
+      "loss": 0.7589567303657532,
+      "step": 928
+    },
+    {
+      "epoch": 0.40129449838187703,
+      "grad_norm": 1.778781533241272,
+      "learning_rate": 8.34681042228212e-05,
+      "loss": 0.8725809454917908,
+      "step": 930
+    },
+    {
+      "epoch": 0.4021574973031284,
+      "grad_norm": 1.6333881616592407,
+      "learning_rate": 8.364779874213837e-05,
+      "loss": 0.8280068039894104,
+      "step": 932
+    },
+    {
+      "epoch": 0.40302049622437974,
+      "grad_norm": 1.6444122791290283,
+      "learning_rate": 8.382749326145552e-05,
+      "loss": 0.8293300867080688,
+      "step": 934
+    },
+    {
+      "epoch": 0.40388349514563104,
+      "grad_norm": 1.2815860509872437,
+      "learning_rate": 8.400718778077269e-05,
+      "loss": 0.7628244161605835,
+      "step": 936
+    },
+    {
+      "epoch": 0.4047464940668824,
+      "grad_norm": 1.7849020957946777,
+      "learning_rate": 8.418688230008985e-05,
+      "loss": 0.8951035141944885,
+      "step": 938
+    },
+    {
+      "epoch": 0.40560949298813376,
+      "grad_norm": 1.800937294960022,
+      "learning_rate": 8.436657681940702e-05,
+      "loss": 0.7677619457244873,
+      "step": 940
+    },
+    {
+      "epoch": 0.4064724919093851,
+      "grad_norm": 1.5004104375839233,
+      "learning_rate": 8.454627133872417e-05,
+      "loss": 0.8553397059440613,
+      "step": 942
+    },
+    {
+      "epoch": 0.40733549083063647,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 8.472596585804133e-05,
+      "loss": 0.7110850811004639,
+      "step": 944
+    },
+    {
+      "epoch": 0.4081984897518878,
+      "grad_norm": 1.8936012983322144,
+      "learning_rate": 8.49056603773585e-05,
+      "loss": 0.9379971623420715,
+      "step": 946
+    },
+    {
+      "epoch": 0.4090614886731392,
+      "grad_norm": 1.2648893594741821,
+      "learning_rate": 8.508535489667565e-05,
+      "loss": 0.7492738366127014,
+      "step": 948
+    },
+    {
+      "epoch": 0.4099244875943905,
+      "grad_norm": 1.5587494373321533,
+      "learning_rate": 8.526504941599281e-05,
+      "loss": 0.7922301292419434,
+      "step": 950
+    },
+    {
+      "epoch": 0.41078748651564184,
+      "grad_norm": 1.5042990446090698,
+      "learning_rate": 8.544474393530998e-05,
+      "loss": 0.7494275569915771,
+      "step": 952
+    },
+    {
+      "epoch": 0.4116504854368932,
+      "grad_norm": 1.4338537454605103,
+      "learning_rate": 8.562443845462714e-05,
+      "loss": 0.7983633279800415,
+      "step": 954
+    },
+    {
+      "epoch": 0.41251348435814456,
+      "grad_norm": 1.6135673522949219,
+      "learning_rate": 8.58041329739443e-05,
+      "loss": 0.813288152217865,
+      "step": 956
+    },
+    {
+      "epoch": 0.4133764832793959,
+      "grad_norm": 1.3761460781097412,
+      "learning_rate": 8.598382749326146e-05,
+      "loss": 0.7794904708862305,
+      "step": 958
+    },
+    {
+      "epoch": 0.41423948220064727,
+      "grad_norm": 1.51504385471344,
+      "learning_rate": 8.616352201257863e-05,
+      "loss": 0.7544562816619873,
+      "step": 960
+    },
+    {
+      "epoch": 0.41510248112189857,
+      "grad_norm": 1.3698909282684326,
+      "learning_rate": 8.634321653189578e-05,
+      "loss": 0.8151395916938782,
+      "step": 962
+    },
+    {
+      "epoch": 0.4159654800431499,
+      "grad_norm": 1.5509650707244873,
+      "learning_rate": 8.652291105121294e-05,
+      "loss": 0.7997464537620544,
+      "step": 964
+    },
+    {
+      "epoch": 0.4168284789644013,
+      "grad_norm": 1.2967383861541748,
+      "learning_rate": 8.67026055705301e-05,
+      "loss": 0.767808735370636,
+      "step": 966
+    },
+    {
+      "epoch": 0.41769147788565264,
+      "grad_norm": 1.2444552183151245,
+      "learning_rate": 8.688230008984727e-05,
+      "loss": 0.7537208795547485,
+      "step": 968
+    },
+    {
+      "epoch": 0.418554476806904,
+      "grad_norm": 1.4390850067138672,
+      "learning_rate": 8.706199460916442e-05,
+      "loss": 0.7796021103858948,
+      "step": 970
+    },
+    {
+      "epoch": 0.41941747572815535,
+      "grad_norm": 1.4822322130203247,
+      "learning_rate": 8.724168912848159e-05,
+      "loss": 0.7812987565994263,
+      "step": 972
+    },
+    {
+      "epoch": 0.4202804746494067,
+      "grad_norm": 1.3770837783813477,
+      "learning_rate": 8.742138364779875e-05,
+      "loss": 0.7396363019943237,
+      "step": 974
+    },
+    {
+      "epoch": 0.421143473570658,
+      "grad_norm": 1.5534236431121826,
+      "learning_rate": 8.76010781671159e-05,
+      "loss": 0.8390982151031494,
+      "step": 976
+    },
+    {
+      "epoch": 0.42200647249190937,
+      "grad_norm": 1.5428354740142822,
+      "learning_rate": 8.778077268643307e-05,
+      "loss": 0.8368987441062927,
+      "step": 978
+    },
+    {
+      "epoch": 0.4228694714131607,
+      "grad_norm": 1.3579497337341309,
+      "learning_rate": 8.796046720575023e-05,
+      "loss": 0.8203011155128479,
+      "step": 980
+    },
+    {
+      "epoch": 0.4237324703344121,
+      "grad_norm": 1.5543184280395508,
+      "learning_rate": 8.81401617250674e-05,
+      "loss": 0.7622520923614502,
+      "step": 982
+    },
+    {
+      "epoch": 0.42459546925566344,
+      "grad_norm": 1.5728850364685059,
+      "learning_rate": 8.831985624438455e-05,
+      "loss": 0.7764845490455627,
+      "step": 984
+    },
+    {
+      "epoch": 0.4254584681769148,
+      "grad_norm": 1.5842933654785156,
+      "learning_rate": 8.849955076370172e-05,
+      "loss": 0.8136081099510193,
+      "step": 986
+    },
+    {
+      "epoch": 0.42632146709816615,
+      "grad_norm": 1.5721858739852905,
+      "learning_rate": 8.867924528301888e-05,
+      "loss": 0.8388259410858154,
+      "step": 988
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6682584285736084,
+      "learning_rate": 8.885893980233603e-05,
+      "loss": 0.8829372525215149,
+      "step": 990
+    },
+    {
+      "epoch": 0.4280474649406688,
+      "grad_norm": 1.7098870277404785,
+      "learning_rate": 8.90386343216532e-05,
+      "loss": 0.8545100688934326,
+      "step": 992
+    },
+    {
+      "epoch": 0.42891046386192017,
+      "grad_norm": 1.5161751508712769,
+      "learning_rate": 8.921832884097035e-05,
+      "loss": 0.8233623504638672,
+      "step": 994
+    },
+    {
+      "epoch": 0.4297734627831715,
+      "grad_norm": 1.457794189453125,
+      "learning_rate": 8.939802336028751e-05,
+      "loss": 0.7576998472213745,
+      "step": 996
+    },
+    {
+      "epoch": 0.4306364617044229,
+      "grad_norm": 1.6383825540542603,
+      "learning_rate": 8.957771787960468e-05,
+      "loss": 0.8089564442634583,
+      "step": 998
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "grad_norm": 1.4149754047393799,
+      "learning_rate": 8.975741239892183e-05,
+      "loss": 0.7791779637336731,
+      "step": 1000
+    },
+    {
+      "epoch": 0.43149946062567424,
+      "eval_loss": 0.8266019821166992,
+      "eval_runtime": 835.0044,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4323624595469256,
+      "grad_norm": 1.5475730895996094,
+      "learning_rate": 8.9937106918239e-05,
+      "loss": 0.7443642616271973,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4332254584681769,
+      "grad_norm": 1.4621447324752808,
+      "learning_rate": 9.011680143755616e-05,
+      "loss": 0.8210510015487671,
+      "step": 1004
+    },
+    {
+      "epoch": 0.43408845738942825,
+      "grad_norm": 1.3651922941207886,
+      "learning_rate": 9.029649595687331e-05,
+      "loss": 0.8130766153335571,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4349514563106796,
+      "grad_norm": 1.3587357997894287,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.79145747423172,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43581445523193096,
+      "grad_norm": 1.464652180671692,
+      "learning_rate": 9.065588499550764e-05,
+      "loss": 0.8286946415901184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4366774541531823,
+      "grad_norm": 1.3912625312805176,
+      "learning_rate": 9.083557951482479e-05,
+      "loss": 0.717319667339325,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4375404530744337,
+      "grad_norm": 1.2045491933822632,
+      "learning_rate": 9.101527403414196e-05,
+      "loss": 0.7294418215751648,
+      "step": 1014
+    },
+    {
+      "epoch": 0.438403451995685,
+      "grad_norm": 1.3449069261550903,
+      "learning_rate": 9.119496855345912e-05,
+      "loss": 0.7366700768470764,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43926645091693634,
+      "grad_norm": 2.901384115219116,
+      "learning_rate": 9.137466307277629e-05,
+      "loss": 0.8551293611526489,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4401294498381877,
+      "grad_norm": 1.6648688316345215,
+      "learning_rate": 9.155435759209344e-05,
+      "loss": 0.8944112658500671,
+      "step": 1020
+    },
+    {
+      "epoch": 0.44099244875943905,
+      "grad_norm": 1.3050304651260376,
+      "learning_rate": 9.17340521114106e-05,
+      "loss": 0.7644214630126953,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4418554476806904,
+      "grad_norm": 1.4643231630325317,
+      "learning_rate": 9.191374663072777e-05,
+      "loss": 0.8644799590110779,
+      "step": 1024
+    },
+    {
+      "epoch": 0.44271844660194176,
+      "grad_norm": 1.4690274000167847,
+      "learning_rate": 9.209344115004492e-05,
+      "loss": 0.8281022906303406,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4435814455231931,
+      "grad_norm": 1.3321986198425293,
+      "learning_rate": 9.227313566936209e-05,
+      "loss": 0.7432141900062561,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.4342927932739258,
+      "learning_rate": 9.245283018867925e-05,
+      "loss": 0.8143638372421265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4453074433656958,
+      "grad_norm": 1.747295618057251,
+      "learning_rate": 9.263252470799642e-05,
+      "loss": 0.8102239370346069,
+      "step": 1032
+    },
+    {
+      "epoch": 0.44617044228694713,
+      "grad_norm": 1.3273073434829712,
+      "learning_rate": 9.281221922731357e-05,
+      "loss": 0.7768088579177856,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4470334412081985,
+      "grad_norm": 1.5019903182983398,
+      "learning_rate": 9.299191374663073e-05,
+      "loss": 0.7886938452720642,
+      "step": 1036
+    },
+    {
+      "epoch": 0.44789644012944985,
+      "grad_norm": 1.6711527109146118,
+      "learning_rate": 9.31716082659479e-05,
+      "loss": 0.8152258396148682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4487594390507012,
+      "grad_norm": 1.8504793643951416,
+      "learning_rate": 9.335130278526505e-05,
+      "loss": 0.8339349627494812,
+      "step": 1040
+    },
+    {
+      "epoch": 0.44962243797195256,
+      "grad_norm": 1.4333337545394897,
+      "learning_rate": 9.353099730458221e-05,
+      "loss": 0.7798911929130554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.45048543689320386,
+      "grad_norm": 1.3361700773239136,
+      "learning_rate": 9.371069182389938e-05,
+      "loss": 0.7676728963851929,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4513484358144552,
+      "grad_norm": 1.338934063911438,
+      "learning_rate": 9.389038634321654e-05,
+      "loss": 0.7729206085205078,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4522114347357066,
+      "grad_norm": 1.4347442388534546,
+      "learning_rate": 9.40700808625337e-05,
+      "loss": 0.7923640608787537,
+      "step": 1048
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 1.4451628923416138,
+      "learning_rate": 9.424977538185086e-05,
+      "loss": 0.8068996071815491,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4539374325782093,
+      "grad_norm": 1.1904500722885132,
+      "learning_rate": 9.442946990116803e-05,
+      "loss": 0.7532910704612732,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45480043149946064,
+      "grad_norm": 1.4103976488113403,
+      "learning_rate": 9.460916442048518e-05,
+      "loss": 0.8236854672431946,
+      "step": 1054
+    },
+    {
+      "epoch": 0.455663430420712,
+      "grad_norm": 1.3762754201889038,
+      "learning_rate": 9.478885893980234e-05,
+      "loss": 0.7532716393470764,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4565264293419633,
+      "grad_norm": 1.461421012878418,
+      "learning_rate": 9.496855345911951e-05,
+      "loss": 0.8506031632423401,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45738942826321466,
+      "grad_norm": 1.7207789421081543,
+      "learning_rate": 9.514824797843667e-05,
+      "loss": 0.7824664115905762,
+      "step": 1060
+    },
+    {
+      "epoch": 0.458252427184466,
+      "grad_norm": 1.5946974754333496,
+      "learning_rate": 9.532794249775382e-05,
+      "loss": 0.8674250245094299,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4591154261057174,
+      "grad_norm": 1.3552565574645996,
+      "learning_rate": 9.550763701707099e-05,
+      "loss": 0.7851845622062683,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45997842502696873,
+      "grad_norm": 1.414617657661438,
+      "learning_rate": 9.568733153638815e-05,
+      "loss": 0.8890700936317444,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4608414239482201,
+      "grad_norm": 1.3322749137878418,
+      "learning_rate": 9.58670260557053e-05,
+      "loss": 0.8226593732833862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4617044228694714,
+      "grad_norm": 1.41420316696167,
+      "learning_rate": 9.604672057502247e-05,
+      "loss": 0.7663303017616272,
+      "step": 1070
+    },
+    {
+      "epoch": 0.46256742179072274,
+      "grad_norm": 1.324967384338379,
+      "learning_rate": 9.622641509433963e-05,
+      "loss": 0.7122516632080078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4634304207119741,
+      "grad_norm": 1.2046154737472534,
+      "learning_rate": 9.640610961365679e-05,
+      "loss": 0.7177618145942688,
+      "step": 1074
+    },
+    {
+      "epoch": 0.46429341963322546,
+      "grad_norm": 1.5221766233444214,
+      "learning_rate": 9.658580413297395e-05,
+      "loss": 0.7618930339813232,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4651564185544768,
+      "grad_norm": 1.3517179489135742,
+      "learning_rate": 9.67654986522911e-05,
+      "loss": 0.7483780384063721,
+      "step": 1078
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.747865080833435,
+      "learning_rate": 9.694519317160827e-05,
+      "loss": 0.8507670164108276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4668824163969795,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 9.712488769092543e-05,
+      "loss": 0.7893540859222412,
+      "step": 1082
+    },
+    {
+      "epoch": 0.46774541531823083,
+      "grad_norm": 1.6290462017059326,
+      "learning_rate": 9.730458221024258e-05,
+      "loss": 0.7891272306442261,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4686084142394822,
+      "grad_norm": 1.6605802774429321,
+      "learning_rate": 9.748427672955975e-05,
+      "loss": 0.841018795967102,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46947141316073354,
+      "grad_norm": 1.388964295387268,
+      "learning_rate": 9.766397124887691e-05,
+      "loss": 0.8099003434181213,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4703344120819849,
+      "grad_norm": 1.3901978731155396,
+      "learning_rate": 9.784366576819407e-05,
+      "loss": 0.7177658677101135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.47119741100323626,
+      "grad_norm": 1.3341423273086548,
+      "learning_rate": 9.802336028751123e-05,
+      "loss": 0.8285676836967468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4720604099244876,
+      "grad_norm": 1.5451914072036743,
+      "learning_rate": 9.82030548068284e-05,
+      "loss": 0.7582666873931885,
+      "step": 1094
+    },
+    {
+      "epoch": 0.47292340884573897,
+      "grad_norm": 1.5395640134811401,
+      "learning_rate": 9.838274932614556e-05,
+      "loss": 0.7293475270271301,
+      "step": 1096
+    },
+    {
+      "epoch": 0.47378640776699027,
+      "grad_norm": 1.5005141496658325,
+      "learning_rate": 9.856244384546271e-05,
+      "loss": 0.8858493566513062,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "grad_norm": 1.3562248945236206,
+      "learning_rate": 9.874213836477988e-05,
+      "loss": 0.7176282405853271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4746494066882416,
+      "eval_loss": 0.8152187466621399,
+      "eval_runtime": 835.0394,
+      "eval_samples_per_second": 2.467,
+      "eval_steps_per_second": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.475512405609493,
+      "grad_norm": 1.5226716995239258,
+      "learning_rate": 9.892183288409704e-05,
+      "loss": 0.8048524856567383,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47637540453074434,
+      "grad_norm": 1.3544842004776,
+      "learning_rate": 9.91015274034142e-05,
+      "loss": 0.7954329252243042,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4772384034519957,
+      "grad_norm": 1.5169771909713745,
+      "learning_rate": 9.928122192273136e-05,
+      "loss": 0.804015040397644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47810140237324705,
+      "grad_norm": 1.4011081457138062,
+      "learning_rate": 9.946091644204852e-05,
+      "loss": 0.7652344703674316,
+      "step": 1108
+    },
+    {
+      "epoch": 0.47896440129449835,
+      "grad_norm": 1.4163414239883423,
+      "learning_rate": 9.964061096136569e-05,
+      "loss": 0.8501335382461548,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4798274002157497,
+      "grad_norm": 1.7406005859375,
+      "learning_rate": 9.982030548068284e-05,
+      "loss": 0.9024044871330261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.48069039913700107,
+      "grad_norm": 1.1490557193756104,
+      "learning_rate": 0.0001,
+      "loss": 0.7308287024497986,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4815533980582524,
+      "grad_norm": 1.2512775659561157,
+      "learning_rate": 9.99999939713627e-05,
+      "loss": 0.7273974418640137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4824163969795038,
+      "grad_norm": 1.3944242000579834,
+      "learning_rate": 9.999997588545227e-05,
+      "loss": 0.8071790337562561,
+      "step": 1118
+    },
+    {
+      "epoch": 0.48327939590075514,
+      "grad_norm": 1.2209112644195557,
+      "learning_rate": 9.999994574227303e-05,
+      "loss": 0.6586678624153137,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4841423948220065,
+      "grad_norm": 1.7328095436096191,
+      "learning_rate": 9.999990354183228e-05,
+      "loss": 0.8529233336448669,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4850053937432578,
+      "grad_norm": 1.3309327363967896,
+      "learning_rate": 9.999984928414021e-05,
+      "loss": 0.7529556155204773,
+      "step": 1124
+    },
+    {
+      "epoch": 0.48586839266450915,
+      "grad_norm": 1.22034752368927,
+      "learning_rate": 9.999978296920987e-05,
+      "loss": 0.6780388355255127,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4867313915857605,
+      "grad_norm": 1.975162386894226,
+      "learning_rate": 9.999970459705727e-05,
+      "loss": 0.7941693067550659,
+      "step": 1128
+    },
+    {
+      "epoch": 0.48759439050701187,
+      "grad_norm": 1.4939875602722168,
+      "learning_rate": 9.999961416770132e-05,
+      "loss": 0.7895156145095825,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4884573894282632,
+      "grad_norm": 1.4237034320831299,
+      "learning_rate": 9.999951168116379e-05,
+      "loss": 0.7662149667739868,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4893203883495146,
+      "grad_norm": 1.24979567527771,
+      "learning_rate": 9.999939713746942e-05,
+      "loss": 0.8079231381416321,
+      "step": 1134
+    },
+    {
+      "epoch": 0.49018338727076594,
+      "grad_norm": 1.2252353429794312,
+      "learning_rate": 9.999927053664585e-05,
+      "loss": 0.7616434693336487,
+      "step": 1136
+    },
+    {
+      "epoch": 0.49104638619201724,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.999913187872357e-05,
+      "loss": 0.8576793074607849,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4919093851132686,
+      "grad_norm": 1.2108005285263062,
+      "learning_rate": 9.999898116373605e-05,
+      "loss": 0.8060292601585388,
+      "step": 1140
+    },
+    {
+      "epoch": 0.49277238403451995,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 9.99988183917196e-05,
+      "loss": 0.7532875537872314,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4936353829557713,
+      "grad_norm": 1.261606216430664,
+      "learning_rate": 9.99986435627135e-05,
+      "loss": 0.7241997718811035,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49449838187702266,
+      "grad_norm": 1.2678134441375732,
+      "learning_rate": 9.99984566767599e-05,
+      "loss": 0.78656005859375,
+      "step": 1146
+    },
+    {
+      "epoch": 0.495361380798274,
+      "grad_norm": 1.4185810089111328,
+      "learning_rate": 9.999825773390387e-05,
+      "loss": 0.7423107624053955,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4962243797195254,
+      "grad_norm": 1.3218237161636353,
+      "learning_rate": 9.999804673419338e-05,
+      "loss": 0.8178372383117676,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4970873786407767,
+      "grad_norm": 1.9698240756988525,
+      "learning_rate": 9.99978236776793e-05,
+      "loss": 0.8016347885131836,
+      "step": 1152
+    },
+    {
+      "epoch": 0.49795037756202803,
+      "grad_norm": 1.2228590250015259,
+      "learning_rate": 9.999758856441544e-05,
+      "loss": 0.782421886920929,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4988133764832794,
+      "grad_norm": 1.2753938436508179,
+      "learning_rate": 9.999734139445849e-05,
+      "loss": 0.7235615849494934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.49967637540453075,
+      "grad_norm": 1.353155493736267,
+      "learning_rate": 9.999708216786805e-05,
+      "loss": 0.7686187028884888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5005393743257821,
+      "grad_norm": 1.4242770671844482,
+      "learning_rate": 9.999681088470663e-05,
+      "loss": 0.7842718362808228,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5014023732470334,
+      "grad_norm": 1.501827359199524,
+      "learning_rate": 9.999652754503966e-05,
+      "loss": 0.8324156999588013,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5022653721682848,
+      "grad_norm": 1.5403931140899658,
+      "learning_rate": 9.999623214893545e-05,
+      "loss": 0.7797855138778687,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5031283710895361,
+      "grad_norm": 1.4075721502304077,
+      "learning_rate": 9.999592469646525e-05,
+      "loss": 0.7667348384857178,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5039913700107875,
+      "grad_norm": 1.4411870241165161,
+      "learning_rate": 9.99956051877032e-05,
+      "loss": 0.8153020739555359,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.4197871685028076,
+      "learning_rate": 9.999527362272632e-05,
+      "loss": 0.8548861742019653,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5057173678532901,
+      "grad_norm": 1.5030468702316284,
+      "learning_rate": 9.999493000161458e-05,
+      "loss": 0.7978097796440125,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5065803667745415,
+      "grad_norm": 1.483534574508667,
+      "learning_rate": 9.999457432445088e-05,
+      "loss": 0.7990087866783142,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5074433656957928,
+      "grad_norm": 1.5424920320510864,
+      "learning_rate": 9.999420659132093e-05,
+      "loss": 0.8072807788848877,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5083063646170443,
+      "grad_norm": 1.129503846168518,
+      "learning_rate": 9.999382680231344e-05,
+      "loss": 0.7216097712516785,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5091693635382956,
+      "grad_norm": 1.3208131790161133,
+      "learning_rate": 9.999343495752001e-05,
+      "loss": 0.8637414574623108,
+      "step": 1180
+    },
+    {
+      "epoch": 0.510032362459547,
+      "grad_norm": 1.5239129066467285,
+      "learning_rate": 9.999303105703508e-05,
+      "loss": 0.8561362028121948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5108953613807983,
+      "grad_norm": 1.3193504810333252,
+      "learning_rate": 9.99926151009561e-05,
+      "loss": 0.7514780759811401,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5117583603020496,
+      "grad_norm": 1.312649130821228,
+      "learning_rate": 9.999218708938336e-05,
+      "loss": 0.8634516596794128,
+      "step": 1186
+    },
+    {
+      "epoch": 0.512621359223301,
+      "grad_norm": 1.2677429914474487,
+      "learning_rate": 9.999174702242006e-05,
+      "loss": 0.7493832111358643,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5134843581445523,
+      "grad_norm": 1.3901816606521606,
+      "learning_rate": 9.999129490017234e-05,
+      "loss": 0.8981389999389648,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5143473570658037,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 9.999083072274921e-05,
+      "loss": 0.6771019697189331,
+      "step": 1192
+    },
+    {
+      "epoch": 0.515210355987055,
+      "grad_norm": 1.1299076080322266,
+      "learning_rate": 9.999035449026261e-05,
+      "loss": 0.789888858795166,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5160733549083064,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 9.998986620282738e-05,
+      "loss": 0.6656325459480286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5169363538295577,
+      "grad_norm": 1.3438260555267334,
+      "learning_rate": 9.998936586056127e-05,
+      "loss": 0.795595645904541,
+      "step": 1198
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 1.45097017288208,
+      "learning_rate": 9.998885346358495e-05,
+      "loss": 0.7769836187362671,
+      "step": 1200
+    },
+    {
+      "epoch": 0.517799352750809,
+      "eval_loss": 0.8070361018180847,
+      "eval_runtime": 823.9296,
+      "eval_samples_per_second": 2.5,
+      "eval_steps_per_second": 2.5,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5186623516720604,
+      "grad_norm": 1.3791370391845703,
+      "learning_rate": 9.998832901202195e-05,
+      "loss": 0.7527318000793457,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5195253505933117,
+      "grad_norm": 1.2574659585952759,
+      "learning_rate": 9.998779250599876e-05,
+      "loss": 0.8021673560142517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5203883495145631,
+      "grad_norm": 1.3305809497833252,
+      "learning_rate": 9.998724394564476e-05,
+      "loss": 0.776639461517334,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5212513484358144,
+      "grad_norm": 1.2330507040023804,
+      "learning_rate": 9.998668333109223e-05,
+      "loss": 0.7381181120872498,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5221143473570659,
+      "grad_norm": 1.0482324361801147,
+      "learning_rate": 9.998611066247635e-05,
+      "loss": 0.6595721244812012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5229773462783172,
+      "grad_norm": 1.5456492900848389,
+      "learning_rate": 9.998552593993522e-05,
+      "loss": 0.8388534784317017,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5238403451995685,
+      "grad_norm": 1.3730167150497437,
+      "learning_rate": 9.998492916360986e-05,
+      "loss": 0.7593658566474915,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5247033441208199,
+      "grad_norm": 1.2668508291244507,
+      "learning_rate": 9.998432033364416e-05,
+      "loss": 0.7796505093574524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5255663430420712,
+      "grad_norm": 1.1175068616867065,
+      "learning_rate": 9.998369945018493e-05,
+      "loss": 0.7493594288825989,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5264293419633226,
+      "grad_norm": 1.121848464012146,
+      "learning_rate": 9.99830665133819e-05,
+      "loss": 0.7791322469711304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5272923408845739,
+      "grad_norm": 1.49788236618042,
+      "learning_rate": 9.998242152338772e-05,
+      "loss": 0.8299469947814941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5281553398058253,
+      "grad_norm": 1.227125644683838,
+      "learning_rate": 9.998176448035791e-05,
+      "loss": 0.7911439538002014,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5290183387270766,
+      "grad_norm": 1.176081895828247,
+      "learning_rate": 9.998109538445091e-05,
+      "loss": 0.6487553119659424,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5298813376483279,
+      "grad_norm": 1.169835090637207,
+      "learning_rate": 9.99804142358281e-05,
+      "loss": 0.7331781983375549,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5307443365695793,
+      "grad_norm": 1.4085001945495605,
+      "learning_rate": 9.997972103465367e-05,
+      "loss": 0.8198121190071106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5316073354908306,
+      "grad_norm": 1.8634135723114014,
+      "learning_rate": 9.997901578109484e-05,
+      "loss": 0.7099791169166565,
+      "step": 1232
+    },
+    {
+      "epoch": 0.532470334412082,
+      "grad_norm": 1.4719511270523071,
+      "learning_rate": 9.997829847532166e-05,
+      "loss": 0.7416951656341553,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.2430769205093384,
+      "learning_rate": 9.99775691175071e-05,
+      "loss": 0.7583408355712891,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5341963322545846,
+      "grad_norm": 1.2502130270004272,
+      "learning_rate": 9.997682770782706e-05,
+      "loss": 0.7639391422271729,
+      "step": 1238
+    },
+    {
+      "epoch": 0.535059331175836,
+      "grad_norm": 1.3577396869659424,
+      "learning_rate": 9.997607424646032e-05,
+      "loss": 0.8304449319839478,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5359223300970873,
+      "grad_norm": 1.0986218452453613,
+      "learning_rate": 9.997530873358854e-05,
+      "loss": 0.7331952452659607,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5367853290183388,
+      "grad_norm": 1.4607514142990112,
+      "learning_rate": 9.997453116939637e-05,
+      "loss": 0.8173373341560364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.53764832793959,
+      "grad_norm": 1.2897552251815796,
+      "learning_rate": 9.997374155407128e-05,
+      "loss": 0.7677372694015503,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5385113268608415,
+      "grad_norm": 1.4581117630004883,
+      "learning_rate": 9.997293988780372e-05,
+      "loss": 0.8524296283721924,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5393743257820928,
+      "grad_norm": 1.2029366493225098,
+      "learning_rate": 9.997212617078698e-05,
+      "loss": 0.7920866012573242,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5402373247033441,
+      "grad_norm": 1.109483242034912,
+      "learning_rate": 9.997130040321727e-05,
+      "loss": 0.7889736294746399,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5411003236245955,
+      "grad_norm": 1.1930333375930786,
+      "learning_rate": 9.997046258529377e-05,
+      "loss": 0.6593763828277588,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5419633225458468,
+      "grad_norm": 1.3668162822723389,
+      "learning_rate": 9.996961271721845e-05,
+      "loss": 0.7768212556838989,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5428263214670982,
+      "grad_norm": 1.2901886701583862,
+      "learning_rate": 9.996875079919632e-05,
+      "loss": 0.7392467260360718,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.366720199584961,
+      "learning_rate": 9.996787683143518e-05,
+      "loss": 0.6961479187011719,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5445523193096009,
+      "grad_norm": 1.2925949096679688,
+      "learning_rate": 9.99669908141458e-05,
+      "loss": 0.801517128944397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5454153182308522,
+      "grad_norm": 1.1967244148254395,
+      "learning_rate": 9.996609274754182e-05,
+      "loss": 0.7287030816078186,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5462783171521035,
+      "grad_norm": 1.419996738433838,
+      "learning_rate": 9.996518263183985e-05,
+      "loss": 0.7791255116462708,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5471413160733549,
+      "grad_norm": 1.3544845581054688,
+      "learning_rate": 9.996426046725933e-05,
+      "loss": 0.6665297150611877,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5480043149946062,
+      "grad_norm": 1.6272345781326294,
+      "learning_rate": 9.996332625402262e-05,
+      "loss": 0.7902728915214539,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5488673139158576,
+      "grad_norm": 1.4264246225357056,
+      "learning_rate": 9.996237999235502e-05,
+      "loss": 0.8483795523643494,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5497303128371089,
+      "grad_norm": 1.3180512189865112,
+      "learning_rate": 9.996142168248473e-05,
+      "loss": 0.7432113289833069,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5505933117583603,
+      "grad_norm": 0.9958710074424744,
+      "learning_rate": 9.996045132464283e-05,
+      "loss": 0.6721545457839966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5514563106796116,
+      "grad_norm": 1.371719241142273,
+      "learning_rate": 9.995946891906331e-05,
+      "loss": 0.7770468592643738,
+      "step": 1278
+    },
+    {
+      "epoch": 0.552319309600863,
+      "grad_norm": 1.1906969547271729,
+      "learning_rate": 9.995847446598307e-05,
+      "loss": 0.816495418548584,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5531823085221144,
+      "grad_norm": 1.262184977531433,
+      "learning_rate": 9.995746796564195e-05,
+      "loss": 0.7122345566749573,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5540453074433657,
+      "grad_norm": 1.454692006111145,
+      "learning_rate": 9.995644941828262e-05,
+      "loss": 0.8476479649543762,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5549083063646171,
+      "grad_norm": 1.44277024269104,
+      "learning_rate": 9.995541882415072e-05,
+      "loss": 0.7595248818397522,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5557713052858684,
+      "grad_norm": 1.151465654373169,
+      "learning_rate": 9.995437618349478e-05,
+      "loss": 0.7413201928138733,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5566343042071198,
+      "grad_norm": 1.3160182237625122,
+      "learning_rate": 9.995332149656621e-05,
+      "loss": 0.7242997884750366,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5574973031283711,
+      "grad_norm": 1.2748576402664185,
+      "learning_rate": 9.995225476361936e-05,
+      "loss": 0.7779543995857239,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5583603020496224,
+      "grad_norm": 1.3639990091323853,
+      "learning_rate": 9.995117598491145e-05,
+      "loss": 0.8075780868530273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5592233009708738,
+      "grad_norm": 1.0247999429702759,
+      "learning_rate": 9.995008516070263e-05,
+      "loss": 0.6940524578094482,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5600862998921251,
+      "grad_norm": 1.3383198976516724,
+      "learning_rate": 9.994898229125597e-05,
+      "loss": 0.799100399017334,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "grad_norm": 1.37660551071167,
+      "learning_rate": 9.994786737683738e-05,
+      "loss": 0.8038942217826843,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5609492988133765,
+      "eval_loss": 0.7962777018547058,
+      "eval_runtime": 827.8798,
+      "eval_samples_per_second": 2.488,
+      "eval_steps_per_second": 2.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5618122977346278,
+      "grad_norm": 1.5424575805664062,
+      "learning_rate": 9.994674041771574e-05,
+      "loss": 0.8553245067596436,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5626752966558792,
+      "grad_norm": 1.1047813892364502,
+      "learning_rate": 9.994560141416282e-05,
+      "loss": 0.6650274991989136,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5635382955771305,
+      "grad_norm": 1.4532393217086792,
+      "learning_rate": 9.994445036645328e-05,
+      "loss": 0.794471025466919,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5644012944983818,
+      "grad_norm": 1.2588927745819092,
+      "learning_rate": 9.994328727486468e-05,
+      "loss": 0.7241487503051758,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5652642934196332,
+      "grad_norm": 1.4981781244277954,
+      "learning_rate": 9.994211213967748e-05,
+      "loss": 0.7781044244766235,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5661272923408845,
+      "grad_norm": 1.426277995109558,
+      "learning_rate": 9.99409249611751e-05,
+      "loss": 0.7772427201271057,
+      "step": 1312
+    },
+    {
+      "epoch": 0.566990291262136,
+      "grad_norm": 1.1987197399139404,
+      "learning_rate": 9.993972573964379e-05,
+      "loss": 0.7443774342536926,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5678532901833873,
+      "grad_norm": 1.1490470170974731,
+      "learning_rate": 9.993851447537275e-05,
+      "loss": 0.7203393578529358,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5687162891046387,
+      "grad_norm": 1.3586851358413696,
+      "learning_rate": 9.993729116865406e-05,
+      "loss": 0.8038793206214905,
+      "step": 1318
+    },
+    {
+      "epoch": 0.56957928802589,
+      "grad_norm": 1.4160653352737427,
+      "learning_rate": 9.993605581978274e-05,
+      "loss": 0.827308714389801,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5704422869471413,
+      "grad_norm": 1.2725251913070679,
+      "learning_rate": 9.993480842905667e-05,
+      "loss": 0.8183196783065796,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5713052858683927,
+      "grad_norm": 1.3789756298065186,
+      "learning_rate": 9.993354899677665e-05,
+      "loss": 0.8268020153045654,
+      "step": 1324
+    },
+    {
+      "epoch": 0.572168284789644,
+      "grad_norm": 1.2641910314559937,
+      "learning_rate": 9.993227752324639e-05,
+      "loss": 0.7434715628623962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5730312837108954,
+      "grad_norm": 1.2429786920547485,
+      "learning_rate": 9.99309940087725e-05,
+      "loss": 0.7636955976486206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5738942826321467,
+      "grad_norm": 1.3281534910202026,
+      "learning_rate": 9.992969845366449e-05,
+      "loss": 0.794617772102356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.574757281553398,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 9.992839085823477e-05,
+      "loss": 0.7658806443214417,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5756202804746494,
+      "grad_norm": 1.4143259525299072,
+      "learning_rate": 9.99270712227987e-05,
+      "loss": 0.8483052849769592,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5764832793959007,
+      "grad_norm": 1.2387818098068237,
+      "learning_rate": 9.992573954767449e-05,
+      "loss": 0.8610434532165527,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5773462783171521,
+      "grad_norm": 1.3307154178619385,
+      "learning_rate": 9.992439583318322e-05,
+      "loss": 0.8021210432052612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5782092772384034,
+      "grad_norm": 1.293459415435791,
+      "learning_rate": 9.992304007964896e-05,
+      "loss": 0.7536460757255554,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5790722761596548,
+      "grad_norm": 1.1508948802947998,
+      "learning_rate": 9.992167228739865e-05,
+      "loss": 0.7519120573997498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5799352750809061,
+      "grad_norm": 1.5906150341033936,
+      "learning_rate": 9.992029245676211e-05,
+      "loss": 0.784415066242218,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5807982740021574,
+      "grad_norm": 1.1421746015548706,
+      "learning_rate": 9.99189005880721e-05,
+      "loss": 0.8217549324035645,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5816612729234089,
+      "grad_norm": 1.0767277479171753,
+      "learning_rate": 9.991749668166422e-05,
+      "loss": 0.7446820735931396,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 1.147594928741455,
+      "learning_rate": 9.991608073787706e-05,
+      "loss": 0.8426011800765991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5833872707659116,
+      "grad_norm": 1.6618242263793945,
+      "learning_rate": 9.991465275705206e-05,
+      "loss": 0.8185623288154602,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5842502696871629,
+      "grad_norm": 1.414926290512085,
+      "learning_rate": 9.991321273953355e-05,
+      "loss": 0.7521756887435913,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5851132686084143,
+      "grad_norm": 1.298245906829834,
+      "learning_rate": 9.991176068566882e-05,
+      "loss": 0.7517471313476562,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5859762675296656,
+      "grad_norm": 1.3243110179901123,
+      "learning_rate": 9.9910296595808e-05,
+      "loss": 0.749305248260498,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5868392664509169,
+      "grad_norm": 1.1953799724578857,
+      "learning_rate": 9.990882047030413e-05,
+      "loss": 0.6883649230003357,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5877022653721683,
+      "grad_norm": 1.3934388160705566,
+      "learning_rate": 9.990733230951322e-05,
+      "loss": 0.7859632968902588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5885652642934196,
+      "grad_norm": 1.2212340831756592,
+      "learning_rate": 9.990583211379409e-05,
+      "loss": 0.7313099503517151,
+      "step": 1364
+    },
+    {
+      "epoch": 0.589428263214671,
+      "grad_norm": 1.5374075174331665,
+      "learning_rate": 9.990431988350853e-05,
+      "loss": 0.8007960319519043,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5902912621359223,
+      "grad_norm": 1.1826236248016357,
+      "learning_rate": 9.990279561902118e-05,
+      "loss": 0.773026168346405,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5911542610571737,
+      "grad_norm": 1.1181561946868896,
+      "learning_rate": 9.990125932069964e-05,
+      "loss": 0.7469213604927063,
+      "step": 1370
+    },
+    {
+      "epoch": 0.592017259978425,
+      "grad_norm": 1.3735432624816895,
+      "learning_rate": 9.989971098891439e-05,
+      "loss": 0.7788950204849243,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5928802588996763,
+      "grad_norm": 1.1906198263168335,
+      "learning_rate": 9.989815062403877e-05,
+      "loss": 0.8117855191230774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5937432578209277,
+      "grad_norm": 1.326940894126892,
+      "learning_rate": 9.989657822644907e-05,
+      "loss": 0.7536540627479553,
+      "step": 1376
+    },
+    {
+      "epoch": 0.594606256742179,
+      "grad_norm": 1.4508434534072876,
+      "learning_rate": 9.989499379652447e-05,
+      "loss": 0.860320508480072,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5954692556634305,
+      "grad_norm": 1.114499568939209,
+      "learning_rate": 9.989339733464704e-05,
+      "loss": 0.749508798122406,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5963322545846818,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 9.989178884120176e-05,
+      "loss": 0.7294815182685852,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5971952535059332,
+      "grad_norm": 1.291913628578186,
+      "learning_rate": 9.989016831657652e-05,
+      "loss": 0.7797688245773315,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5980582524271845,
+      "grad_norm": 1.1245452165603638,
+      "learning_rate": 9.988853576116208e-05,
+      "loss": 0.7824407815933228,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5989212513484358,
+      "grad_norm": 1.2384979724884033,
+      "learning_rate": 9.988689117535215e-05,
+      "loss": 0.714794933795929,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5997842502696872,
+      "grad_norm": 1.1945903301239014,
+      "learning_rate": 9.988523455954332e-05,
+      "loss": 0.7072597742080688,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6006472491909385,
+      "grad_norm": 1.3414266109466553,
+      "learning_rate": 9.988356591413504e-05,
+      "loss": 0.7075814008712769,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6015102481121899,
+      "grad_norm": 1.4575660228729248,
+      "learning_rate": 9.988188523952973e-05,
+      "loss": 0.8237126469612122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6023732470334412,
+      "grad_norm": 1.3836476802825928,
+      "learning_rate": 9.988019253613265e-05,
+      "loss": 0.7991626858711243,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6032362459546926,
+      "grad_norm": 1.2176212072372437,
+      "learning_rate": 9.987848780435201e-05,
+      "loss": 0.6936330199241638,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "grad_norm": 1.171294093132019,
+      "learning_rate": 9.98767710445989e-05,
+      "loss": 0.7834921479225159,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6040992448759439,
+      "eval_loss": 0.7866972088813782,
+      "eval_runtime": 836.0759,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6049622437971952,
+      "grad_norm": 1.5934456586837769,
+      "learning_rate": 9.987504225728727e-05,
+      "loss": 0.765119194984436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6058252427184466,
+      "grad_norm": 1.231372356414795,
+      "learning_rate": 9.987330144283407e-05,
+      "loss": 0.752061665058136,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6066882416396979,
+      "grad_norm": 1.315215826034546,
+      "learning_rate": 9.987154860165903e-05,
+      "loss": 0.7475255727767944,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6075512405609493,
+      "grad_norm": 1.2640653848648071,
+      "learning_rate": 9.986978373418488e-05,
+      "loss": 0.777019202709198,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6084142394822006,
+      "grad_norm": 1.085493803024292,
+      "learning_rate": 9.98680068408372e-05,
+      "loss": 0.6682190299034119,
+      "step": 1410
+    },
+    {
+      "epoch": 0.609277238403452,
+      "grad_norm": 1.4247251749038696,
+      "learning_rate": 9.986621792204447e-05,
+      "loss": 0.7765674591064453,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6101402373247033,
+      "grad_norm": 1.2993831634521484,
+      "learning_rate": 9.986441697823808e-05,
+      "loss": 0.7429938316345215,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6110032362459547,
+      "grad_norm": 2.614980936050415,
+      "learning_rate": 9.986260400985233e-05,
+      "loss": 0.7838386297225952,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6118662351672061,
+      "grad_norm": 1.4305615425109863,
+      "learning_rate": 9.98607790173244e-05,
+      "loss": 0.7679475545883179,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6127292340884574,
+      "grad_norm": 1.1412850618362427,
+      "learning_rate": 9.98589420010944e-05,
+      "loss": 0.6889669895172119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6135922330097088,
+      "grad_norm": 1.372313141822815,
+      "learning_rate": 9.98570929616053e-05,
+      "loss": 0.7659094333648682,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6144552319309601,
+      "grad_norm": 1.2629581689834595,
+      "learning_rate": 9.985523189930297e-05,
+      "loss": 0.7404800653457642,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6153182308522115,
+      "grad_norm": 1.5652254819869995,
+      "learning_rate": 9.985335881463623e-05,
+      "loss": 0.8082308769226074,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6161812297734628,
+      "grad_norm": 1.2268856763839722,
+      "learning_rate": 9.985147370805675e-05,
+      "loss": 0.7773997187614441,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6170442286947141,
+      "grad_norm": 1.2190957069396973,
+      "learning_rate": 9.984957658001912e-05,
+      "loss": 0.7412725687026978,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6179072276159655,
+      "grad_norm": 1.481990098953247,
+      "learning_rate": 9.984766743098081e-05,
+      "loss": 0.7257611751556396,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6187702265372168,
+      "grad_norm": 1.0297027826309204,
+      "learning_rate": 9.984574626140223e-05,
+      "loss": 0.7151464819908142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6196332254584682,
+      "grad_norm": 1.2404571771621704,
+      "learning_rate": 9.984381307174663e-05,
+      "loss": 0.7855883240699768,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6204962243797195,
+      "grad_norm": 1.2952802181243896,
+      "learning_rate": 9.984186786248023e-05,
+      "loss": 0.6360295414924622,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.5405724048614502,
+      "learning_rate": 9.983991063407206e-05,
+      "loss": 0.7280036807060242,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.303699254989624,
+      "learning_rate": 9.983794138699411e-05,
+      "loss": 0.7553068995475769,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6230852211434735,
+      "grad_norm": 1.1452997922897339,
+      "learning_rate": 9.983596012172127e-05,
+      "loss": 0.7329047918319702,
+      "step": 1444
+    },
+    {
+      "epoch": 0.623948220064725,
+      "grad_norm": 1.2648253440856934,
+      "learning_rate": 9.983396683873133e-05,
+      "loss": 0.753168523311615,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6248112189859762,
+      "grad_norm": 1.2095013856887817,
+      "learning_rate": 9.983196153850493e-05,
+      "loss": 0.7378705739974976,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6256742179072277,
+      "grad_norm": 1.2855647802352905,
+      "learning_rate": 9.982994422152564e-05,
+      "loss": 0.7910448908805847,
+      "step": 1450
+    },
+    {
+      "epoch": 0.626537216828479,
+      "grad_norm": 1.3895999193191528,
+      "learning_rate": 9.982791488827994e-05,
+      "loss": 0.785302996635437,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6274002157497303,
+      "grad_norm": 1.2398446798324585,
+      "learning_rate": 9.982587353925719e-05,
+      "loss": 0.7542029023170471,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6282632146709817,
+      "grad_norm": 1.2903637886047363,
+      "learning_rate": 9.982382017494965e-05,
+      "loss": 0.8081737756729126,
+      "step": 1456
+    },
+    {
+      "epoch": 0.629126213592233,
+      "grad_norm": 1.223611831665039,
+      "learning_rate": 9.98217547958525e-05,
+      "loss": 0.7886892557144165,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6299892125134844,
+      "grad_norm": 1.3036459684371948,
+      "learning_rate": 9.981967740246377e-05,
+      "loss": 0.844115138053894,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6308522114347357,
+      "grad_norm": 1.6421090364456177,
+      "learning_rate": 9.98175879952844e-05,
+      "loss": 0.7784068584442139,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6317152103559871,
+      "grad_norm": 1.244330644607544,
+      "learning_rate": 9.981548657481828e-05,
+      "loss": 0.746979832649231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6325782092772384,
+      "grad_norm": 1.0323492288589478,
+      "learning_rate": 9.981337314157215e-05,
+      "loss": 0.6680153608322144,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6334412081984897,
+      "grad_norm": 1.2238314151763916,
+      "learning_rate": 9.981124769605563e-05,
+      "loss": 0.7103193402290344,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6343042071197411,
+      "grad_norm": 1.2647801637649536,
+      "learning_rate": 9.98091102387813e-05,
+      "loss": 0.7426751255989075,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6351672060409924,
+      "grad_norm": 1.1923259496688843,
+      "learning_rate": 9.980696077026456e-05,
+      "loss": 0.7664520144462585,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6360302049622438,
+      "grad_norm": 1.0503894090652466,
+      "learning_rate": 9.980479929102377e-05,
+      "loss": 0.7348257899284363,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6368932038834951,
+      "grad_norm": 1.2223190069198608,
+      "learning_rate": 9.980262580158014e-05,
+      "loss": 0.687703013420105,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6377562028047465,
+      "grad_norm": 1.1469025611877441,
+      "learning_rate": 9.980044030245782e-05,
+      "loss": 0.7903051972389221,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6386192017259978,
+      "grad_norm": 1.2581169605255127,
+      "learning_rate": 9.979824279418381e-05,
+      "loss": 0.7920540571212769,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6394822006472491,
+      "grad_norm": 1.2616536617279053,
+      "learning_rate": 9.979603327728806e-05,
+      "loss": 0.7030006647109985,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6403451995685006,
+      "grad_norm": 1.3970606327056885,
+      "learning_rate": 9.979381175230335e-05,
+      "loss": 0.7489339709281921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6412081984897519,
+      "grad_norm": 1.0060384273529053,
+      "learning_rate": 9.979157821976542e-05,
+      "loss": 0.6733010411262512,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6420711974110033,
+      "grad_norm": 1.4987449645996094,
+      "learning_rate": 9.978933268021286e-05,
+      "loss": 0.8221960663795471,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6429341963322546,
+      "grad_norm": 1.2843966484069824,
+      "learning_rate": 9.978707513418719e-05,
+      "loss": 0.8829001188278198,
+      "step": 1490
+    },
+    {
+      "epoch": 0.643797195253506,
+      "grad_norm": 1.175879955291748,
+      "learning_rate": 9.978480558223278e-05,
+      "loss": 0.7605764269828796,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6446601941747573,
+      "grad_norm": 1.1991835832595825,
+      "learning_rate": 9.978252402489692e-05,
+      "loss": 0.7618141174316406,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6455231930960086,
+      "grad_norm": 1.167635202407837,
+      "learning_rate": 9.978023046272983e-05,
+      "loss": 0.7071146368980408,
+      "step": 1496
+    },
+    {
+      "epoch": 0.64638619201726,
+      "grad_norm": 1.263371229171753,
+      "learning_rate": 9.977792489628456e-05,
+      "loss": 0.7559816837310791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 1.2108519077301025,
+      "learning_rate": 9.977560732611713e-05,
+      "loss": 0.7253944873809814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "eval_loss": 0.778505802154541,
+      "eval_runtime": 836.2317,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6481121898597627,
+      "grad_norm": 1.4794813394546509,
+      "learning_rate": 9.977327775278636e-05,
+      "loss": 0.8212620615959167,
+      "step": 1502
+    },
+    {
+      "epoch": 0.648975188781014,
+      "grad_norm": 1.2120516300201416,
+      "learning_rate": 9.977093617685404e-05,
+      "loss": 0.7220476865768433,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6498381877022654,
+      "grad_norm": 1.3129044771194458,
+      "learning_rate": 9.976858259888484e-05,
+      "loss": 0.7791266441345215,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6507011866235167,
+      "grad_norm": 1.146490454673767,
+      "learning_rate": 9.976621701944628e-05,
+      "loss": 0.7052001357078552,
+      "step": 1508
+    },
+    {
+      "epoch": 0.651564185544768,
+      "grad_norm": 1.1903722286224365,
+      "learning_rate": 9.976383943910885e-05,
+      "loss": 0.7135450839996338,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6524271844660194,
+      "grad_norm": 1.4106701612472534,
+      "learning_rate": 9.976144985844589e-05,
+      "loss": 0.8114889860153198,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6532901833872707,
+      "grad_norm": 1.245661973953247,
+      "learning_rate": 9.97590482780336e-05,
+      "loss": 0.7501562237739563,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6541531823085222,
+      "grad_norm": 1.2918543815612793,
+      "learning_rate": 9.975663469845114e-05,
+      "loss": 0.7697334289550781,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6550161812297735,
+      "grad_norm": 1.2699558734893799,
+      "learning_rate": 9.975420912028053e-05,
+      "loss": 0.7457168102264404,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6558791801510249,
+      "grad_norm": 1.3968263864517212,
+      "learning_rate": 9.975177154410668e-05,
+      "loss": 0.8368204236030579,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6567421790722762,
+      "grad_norm": 1.2161848545074463,
+      "learning_rate": 9.974932197051741e-05,
+      "loss": 0.7355788350105286,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6576051779935275,
+      "grad_norm": 1.129886507987976,
+      "learning_rate": 9.974686040010341e-05,
+      "loss": 0.8023449182510376,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6584681769147789,
+      "grad_norm": 1.1114557981491089,
+      "learning_rate": 9.974438683345829e-05,
+      "loss": 0.7965474724769592,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6593311758360302,
+      "grad_norm": 1.5741950273513794,
+      "learning_rate": 9.974190127117853e-05,
+      "loss": 0.7176838517189026,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 1.1439597606658936,
+      "learning_rate": 9.973940371386352e-05,
+      "loss": 0.7114462852478027,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6610571736785329,
+      "grad_norm": 1.273659348487854,
+      "learning_rate": 9.973689416211553e-05,
+      "loss": 0.754712700843811,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6619201725997842,
+      "grad_norm": 3.9874372482299805,
+      "learning_rate": 9.973437261653973e-05,
+      "loss": 0.6720641255378723,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6627831715210356,
+      "grad_norm": 1.2506502866744995,
+      "learning_rate": 9.973183907774417e-05,
+      "loss": 0.793043851852417,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6636461704422869,
+      "grad_norm": 1.43257737159729,
+      "learning_rate": 9.972929354633982e-05,
+      "loss": 0.7806380987167358,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6645091693635383,
+      "grad_norm": 1.4794938564300537,
+      "learning_rate": 9.97267360229405e-05,
+      "loss": 0.7176088094711304,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6653721682847896,
+      "grad_norm": 1.2398169040679932,
+      "learning_rate": 9.972416650816297e-05,
+      "loss": 0.7145742774009705,
+      "step": 1542
+    },
+    {
+      "epoch": 0.666235167206041,
+      "grad_norm": 1.1309953927993774,
+      "learning_rate": 9.972158500262683e-05,
+      "loss": 0.8025617003440857,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6670981661272923,
+      "grad_norm": 1.2479451894760132,
+      "learning_rate": 9.971899150695463e-05,
+      "loss": 0.7808589339256287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6679611650485436,
+      "grad_norm": 1.2538350820541382,
+      "learning_rate": 9.971638602177175e-05,
+      "loss": 0.7096936106681824,
+      "step": 1548
+    },
+    {
+      "epoch": 0.668824163969795,
+      "grad_norm": 1.1431525945663452,
+      "learning_rate": 9.97137685477065e-05,
+      "loss": 0.6529367566108704,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6696871628910464,
+      "grad_norm": 1.5966559648513794,
+      "learning_rate": 9.971113908539008e-05,
+      "loss": 0.7780346870422363,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6705501618122978,
+      "grad_norm": 1.186475157737732,
+      "learning_rate": 9.970849763545658e-05,
+      "loss": 0.6889166831970215,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6714131607335491,
+      "grad_norm": 1.1922476291656494,
+      "learning_rate": 9.970584419854295e-05,
+      "loss": 0.7154867649078369,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6722761596548005,
+      "grad_norm": 1.30132257938385,
+      "learning_rate": 9.970317877528906e-05,
+      "loss": 0.7419366240501404,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6731391585760518,
+      "grad_norm": 1.1539888381958008,
+      "learning_rate": 9.970050136633769e-05,
+      "loss": 0.7856602072715759,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6740021574973031,
+      "grad_norm": 1.3056161403656006,
+      "learning_rate": 9.969781197233445e-05,
+      "loss": 0.6834115386009216,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6748651564185545,
+      "grad_norm": 1.2899249792099,
+      "learning_rate": 9.96951105939279e-05,
+      "loss": 0.8490727543830872,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6757281553398058,
+      "grad_norm": 1.2182374000549316,
+      "learning_rate": 9.969239723176946e-05,
+      "loss": 0.7462526559829712,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6765911542610572,
+      "grad_norm": 1.1231273412704468,
+      "learning_rate": 9.968967188651344e-05,
+      "loss": 0.7145209908485413,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6774541531823085,
+      "grad_norm": 1.1776760816574097,
+      "learning_rate": 9.968693455881703e-05,
+      "loss": 0.7911707162857056,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6783171521035599,
+      "grad_norm": 1.3542319536209106,
+      "learning_rate": 9.968418524934035e-05,
+      "loss": 0.6946635246276855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6791801510248112,
+      "grad_norm": 1.3073060512542725,
+      "learning_rate": 9.96814239587464e-05,
+      "loss": 0.7597488164901733,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6800431499460625,
+      "grad_norm": 1.3205567598342896,
+      "learning_rate": 9.967865068770099e-05,
+      "loss": 0.6763246655464172,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6809061488673139,
+      "grad_norm": 1.1734344959259033,
+      "learning_rate": 9.967586543687295e-05,
+      "loss": 0.7560655474662781,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6817691477885652,
+      "grad_norm": 1.445770025253296,
+      "learning_rate": 9.967306820693388e-05,
+      "loss": 0.8050205707550049,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6826321467098166,
+      "grad_norm": 1.1185686588287354,
+      "learning_rate": 9.967025899855836e-05,
+      "loss": 0.6628211140632629,
+      "step": 1582
+    },
+    {
+      "epoch": 0.683495145631068,
+      "grad_norm": 1.3831934928894043,
+      "learning_rate": 9.966743781242379e-05,
+      "loss": 0.7684114575386047,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6843581445523194,
+      "grad_norm": 1.2417019605636597,
+      "learning_rate": 9.966460464921047e-05,
+      "loss": 0.7084140777587891,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6852211434735707,
+      "grad_norm": 1.2138086557388306,
+      "learning_rate": 9.966175950960163e-05,
+      "loss": 0.6817399263381958,
+      "step": 1588
+    },
+    {
+      "epoch": 0.686084142394822,
+      "grad_norm": 1.399778962135315,
+      "learning_rate": 9.965890239428339e-05,
+      "loss": 0.8422354459762573,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6869471413160734,
+      "grad_norm": 1.4348714351654053,
+      "learning_rate": 9.965603330394468e-05,
+      "loss": 0.7432681322097778,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6878101402373247,
+      "grad_norm": 1.192425012588501,
+      "learning_rate": 9.965315223927741e-05,
+      "loss": 0.7571184635162354,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6886731391585761,
+      "grad_norm": 1.2834038734436035,
+      "learning_rate": 9.96502592009763e-05,
+      "loss": 0.802443265914917,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6895361380798274,
+      "grad_norm": 1.1650444269180298,
+      "learning_rate": 9.964735418973901e-05,
+      "loss": 0.8113405704498291,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "grad_norm": 1.1714438199996948,
+      "learning_rate": 9.964443720626606e-05,
+      "loss": 0.7507118582725525,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6903991370010788,
+      "eval_loss": 0.7706249952316284,
+      "eval_runtime": 821.4894,
+      "eval_samples_per_second": 2.508,
+      "eval_steps_per_second": 2.508,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6912621359223301,
+      "grad_norm": 1.175459861755371,
+      "learning_rate": 9.964150825126088e-05,
+      "loss": 0.7422796487808228,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6921251348435814,
+      "grad_norm": 1.6418036222457886,
+      "learning_rate": 9.963856732542978e-05,
+      "loss": 0.7125853300094604,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6929881337648328,
+      "grad_norm": 1.3256275653839111,
+      "learning_rate": 9.963561442948193e-05,
+      "loss": 0.757859468460083,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6938511326860841,
+      "grad_norm": 1.19981050491333,
+      "learning_rate": 9.963264956412944e-05,
+      "loss": 0.6849292516708374,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6947141316073355,
+      "grad_norm": 1.0392169952392578,
+      "learning_rate": 9.962967273008723e-05,
+      "loss": 0.7636215090751648,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6955771305285868,
+      "grad_norm": 1.1106646060943604,
+      "learning_rate": 9.962668392807318e-05,
+      "loss": 0.7208617329597473,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6964401294498382,
+      "grad_norm": 1.1569621562957764,
+      "learning_rate": 9.962368315880801e-05,
+      "loss": 0.671805739402771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6973031283710895,
+      "grad_norm": 1.2157769203186035,
+      "learning_rate": 9.962067042301537e-05,
+      "loss": 0.7648762464523315,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6981661272923408,
+      "grad_norm": 1.4804017543792725,
+      "learning_rate": 9.961764572142173e-05,
+      "loss": 0.8490509390830994,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 1.0344431400299072,
+      "learning_rate": 9.961460905475651e-05,
+      "loss": 0.5950891971588135,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6998921251348436,
+      "grad_norm": 1.1109060049057007,
+      "learning_rate": 9.9611560423752e-05,
+      "loss": 0.7288329005241394,
+      "step": 1622
+    },
+    {
+      "epoch": 0.700755124056095,
+      "grad_norm": 1.096996545791626,
+      "learning_rate": 9.960849982914331e-05,
+      "loss": 0.7074542045593262,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7016181229773463,
+      "grad_norm": 1.3109341859817505,
+      "learning_rate": 9.960542727166853e-05,
+      "loss": 0.7348865270614624,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7024811218985976,
+      "grad_norm": 1.3402032852172852,
+      "learning_rate": 9.96023427520686e-05,
+      "loss": 0.7658320665359497,
+      "step": 1628
+    },
+    {
+      "epoch": 0.703344120819849,
+      "grad_norm": 1.3712561130523682,
+      "learning_rate": 9.959924627108732e-05,
+      "loss": 0.809951901435852,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7042071197411003,
+      "grad_norm": 1.2278292179107666,
+      "learning_rate": 9.959613782947139e-05,
+      "loss": 0.6834598779678345,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7050701186623517,
+      "grad_norm": 1.0510832071304321,
+      "learning_rate": 9.95930174279704e-05,
+      "loss": 0.7416030764579773,
+      "step": 1634
+    },
+    {
+      "epoch": 0.705933117583603,
+      "grad_norm": 1.130991816520691,
+      "learning_rate": 9.958988506733684e-05,
+      "loss": 0.7304092645645142,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7067961165048544,
+      "grad_norm": 1.3328672647476196,
+      "learning_rate": 9.958674074832603e-05,
+      "loss": 0.8230003118515015,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7076591154261057,
+      "grad_norm": 1.3842110633850098,
+      "learning_rate": 9.958358447169625e-05,
+      "loss": 0.8413107991218567,
+      "step": 1640
+    },
+    {
+      "epoch": 0.708522114347357,
+      "grad_norm": 1.3143812417984009,
+      "learning_rate": 9.958041623820858e-05,
+      "loss": 0.77432781457901,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7093851132686084,
+      "grad_norm": 1.4727009534835815,
+      "learning_rate": 9.957723604862704e-05,
+      "loss": 0.7563844323158264,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7102481121898597,
+      "grad_norm": 1.1892327070236206,
+      "learning_rate": 9.957404390371854e-05,
+      "loss": 0.7085853219032288,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.124180555343628,
+      "learning_rate": 9.957083980425282e-05,
+      "loss": 0.6837730407714844,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 1.2224774360656738,
+      "learning_rate": 9.956762375100255e-05,
+      "loss": 0.7255740165710449,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7128371089536139,
+      "grad_norm": 1.360425591468811,
+      "learning_rate": 9.956439574474326e-05,
+      "loss": 0.7849710583686829,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7137001078748652,
+      "grad_norm": 1.2818151712417603,
+      "learning_rate": 9.956115578625338e-05,
+      "loss": 0.7669665813446045,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7145631067961165,
+      "grad_norm": 1.2415164709091187,
+      "learning_rate": 9.95579038763142e-05,
+      "loss": 0.7052890658378601,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7154261057173679,
+      "grad_norm": 1.3983269929885864,
+      "learning_rate": 9.955464001570992e-05,
+      "loss": 0.8300023674964905,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7162891046386192,
+      "grad_norm": 1.1591373682022095,
+      "learning_rate": 9.955136420522759e-05,
+      "loss": 0.7742894887924194,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7171521035598706,
+      "grad_norm": 1.2503128051757812,
+      "learning_rate": 9.954807644565715e-05,
+      "loss": 0.8458141684532166,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7180151024811219,
+      "grad_norm": 1.2954912185668945,
+      "learning_rate": 9.954477673779146e-05,
+      "loss": 0.7044742107391357,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7188781014023733,
+      "grad_norm": 1.1669788360595703,
+      "learning_rate": 9.954146508242619e-05,
+      "loss": 0.7917183041572571,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7197411003236246,
+      "grad_norm": 1.1466871500015259,
+      "learning_rate": 9.953814148035996e-05,
+      "loss": 0.7174562811851501,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7206040992448759,
+      "grad_norm": 1.367224097251892,
+      "learning_rate": 9.953480593239421e-05,
+      "loss": 0.8925880193710327,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7214670981661273,
+      "grad_norm": 1.049497365951538,
+      "learning_rate": 9.953145843933334e-05,
+      "loss": 0.70949387550354,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7223300970873786,
+      "grad_norm": 0.9995084404945374,
+      "learning_rate": 9.952809900198455e-05,
+      "loss": 0.6731935143470764,
+      "step": 1674
+    },
+    {
+      "epoch": 0.72319309600863,
+      "grad_norm": 1.100899577140808,
+      "learning_rate": 9.952472762115794e-05,
+      "loss": 0.7316275835037231,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7240560949298813,
+      "grad_norm": 1.2594135999679565,
+      "learning_rate": 9.952134429766655e-05,
+      "loss": 0.7379520535469055,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7249190938511327,
+      "grad_norm": 1.3124911785125732,
+      "learning_rate": 9.951794903232621e-05,
+      "loss": 0.7072247266769409,
+      "step": 1680
+    },
+    {
+      "epoch": 0.725782092772384,
+      "grad_norm": 1.212144374847412,
+      "learning_rate": 9.95145418259557e-05,
+      "loss": 0.745680034160614,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7266450916936353,
+      "grad_norm": 1.3913586139678955,
+      "learning_rate": 9.951112267937663e-05,
+      "loss": 0.796852707862854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7275080906148867,
+      "grad_norm": 1.3227289915084839,
+      "learning_rate": 9.950769159341353e-05,
+      "loss": 0.700892984867096,
+      "step": 1686
+    },
+    {
+      "epoch": 0.728371089536138,
+      "grad_norm": 1.4656381607055664,
+      "learning_rate": 9.950424856889378e-05,
+      "loss": 0.7973095178604126,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7292340884573895,
+      "grad_norm": 1.1698973178863525,
+      "learning_rate": 9.950079360664766e-05,
+      "loss": 0.7811378836631775,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7300970873786408,
+      "grad_norm": 1.201695203781128,
+      "learning_rate": 9.949732670750831e-05,
+      "loss": 0.753012478351593,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7309600862998922,
+      "grad_norm": 1.3465325832366943,
+      "learning_rate": 9.949384787231175e-05,
+      "loss": 0.7620642781257629,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7318230852211435,
+      "grad_norm": 1.2191137075424194,
+      "learning_rate": 9.94903571018969e-05,
+      "loss": 0.7479873299598694,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7326860841423948,
+      "grad_norm": 1.1412187814712524,
+      "learning_rate": 9.948685439710554e-05,
+      "loss": 0.7327720522880554,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "grad_norm": 1.310327410697937,
+      "learning_rate": 9.948333975878233e-05,
+      "loss": 0.761475682258606,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7335490830636462,
+      "eval_loss": 0.7649155259132385,
+      "eval_runtime": 828.3542,
+      "eval_samples_per_second": 2.487,
+      "eval_steps_per_second": 2.487,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7344120819848975,
+      "grad_norm": 1.3031405210494995,
+      "learning_rate": 9.947981318777482e-05,
+      "loss": 0.7564467191696167,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7352750809061489,
+      "grad_norm": 1.2672377824783325,
+      "learning_rate": 9.94762746849334e-05,
+      "loss": 0.7689576745033264,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7361380798274002,
+      "grad_norm": 1.1159595251083374,
+      "learning_rate": 9.947272425111138e-05,
+      "loss": 0.7923810482025146,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7370010787486516,
+      "grad_norm": 1.1190921068191528,
+      "learning_rate": 9.946916188716495e-05,
+      "loss": 0.6540790796279907,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 1.373832106590271,
+      "learning_rate": 9.946558759395312e-05,
+      "loss": 0.7267473936080933,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7387270765911542,
+      "grad_norm": 1.2346374988555908,
+      "learning_rate": 9.946200137233785e-05,
+      "loss": 0.7693060636520386,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7395900755124056,
+      "grad_norm": 1.1629668474197388,
+      "learning_rate": 9.945840322318391e-05,
+      "loss": 0.7751407027244568,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7404530744336569,
+      "grad_norm": 1.2268708944320679,
+      "learning_rate": 9.945479314735899e-05,
+      "loss": 0.7798545360565186,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7413160733549083,
+      "grad_norm": 0.9968738555908203,
+      "learning_rate": 9.945117114573366e-05,
+      "loss": 0.6128867268562317,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7421790722761596,
+      "grad_norm": 1.2135992050170898,
+      "learning_rate": 9.944753721918133e-05,
+      "loss": 0.7536954283714294,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7430420711974111,
+      "grad_norm": 1.1361218690872192,
+      "learning_rate": 9.94438913685783e-05,
+      "loss": 0.7034479379653931,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7439050701186624,
+      "grad_norm": 1.0728967189788818,
+      "learning_rate": 9.944023359480378e-05,
+      "loss": 0.7285382747650146,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7447680690399137,
+      "grad_norm": 1.3613945245742798,
+      "learning_rate": 9.94365638987398e-05,
+      "loss": 0.7865195274353027,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7456310679611651,
+      "grad_norm": 1.0985898971557617,
+      "learning_rate": 9.94328822812713e-05,
+      "loss": 0.7511224746704102,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7464940668824164,
+      "grad_norm": 1.1506208181381226,
+      "learning_rate": 9.942918874328606e-05,
+      "loss": 0.732453465461731,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7473570658036678,
+      "grad_norm": 1.0995100736618042,
+      "learning_rate": 9.94254832856748e-05,
+      "loss": 0.702409029006958,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7482200647249191,
+      "grad_norm": 1.0075604915618896,
+      "learning_rate": 9.942176590933106e-05,
+      "loss": 0.7515703439712524,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7490830636461704,
+      "grad_norm": 1.3360812664031982,
+      "learning_rate": 9.941803661515126e-05,
+      "loss": 0.6910237073898315,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7499460625674218,
+      "grad_norm": 1.0172319412231445,
+      "learning_rate": 9.941429540403471e-05,
+      "loss": 0.7185606956481934,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7508090614886731,
+      "grad_norm": 1.2721658945083618,
+      "learning_rate": 9.941054227688358e-05,
+      "loss": 0.7136536836624146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7516720604099245,
+      "grad_norm": 1.0775413513183594,
+      "learning_rate": 9.940677723460292e-05,
+      "loss": 0.7595619559288025,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7525350593311758,
+      "grad_norm": 1.3436479568481445,
+      "learning_rate": 9.940300027810067e-05,
+      "loss": 0.7602618336677551,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7533980582524272,
+      "grad_norm": 1.2611958980560303,
+      "learning_rate": 9.93992114082876e-05,
+      "loss": 0.7954604029655457,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7542610571736785,
+      "grad_norm": 1.2794841527938843,
+      "learning_rate": 9.93954106260774e-05,
+      "loss": 0.7661190629005432,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7551240560949298,
+      "grad_norm": 1.3042595386505127,
+      "learning_rate": 9.939159793238661e-05,
+      "loss": 0.6855312585830688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7559870550161812,
+      "grad_norm": 1.2799923419952393,
+      "learning_rate": 9.938777332813463e-05,
+      "loss": 0.7970905900001526,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7568500539374325,
+      "grad_norm": 1.169205904006958,
+      "learning_rate": 9.938393681424375e-05,
+      "loss": 0.7072843313217163,
+      "step": 1754
+    },
+    {
+      "epoch": 0.757713052858684,
+      "grad_norm": 1.3833354711532593,
+      "learning_rate": 9.938008839163913e-05,
+      "loss": 0.7134629487991333,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7585760517799353,
+      "grad_norm": 1.223452091217041,
+      "learning_rate": 9.93762280612488e-05,
+      "loss": 0.7873107194900513,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7594390507011867,
+      "grad_norm": 1.2740365266799927,
+      "learning_rate": 9.937235582400367e-05,
+      "loss": 0.7680654525756836,
+      "step": 1760
+    },
+    {
+      "epoch": 0.760302049622438,
+      "grad_norm": 1.2564697265625,
+      "learning_rate": 9.93684716808375e-05,
+      "loss": 0.734126627445221,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7611650485436893,
+      "grad_norm": 1.3858847618103027,
+      "learning_rate": 9.936457563268694e-05,
+      "loss": 0.7972320914268494,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7620280474649407,
+      "grad_norm": 1.2705003023147583,
+      "learning_rate": 9.93606676804915e-05,
+      "loss": 0.7904276847839355,
+      "step": 1766
+    },
+    {
+      "epoch": 0.762891046386192,
+      "grad_norm": 1.210766077041626,
+      "learning_rate": 9.935674782519358e-05,
+      "loss": 0.7542425990104675,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7637540453074434,
+      "grad_norm": 1.4367541074752808,
+      "learning_rate": 9.93528160677384e-05,
+      "loss": 0.7975004315376282,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7646170442286947,
+      "grad_norm": 1.0994079113006592,
+      "learning_rate": 9.934887240907414e-05,
+      "loss": 0.7537807822227478,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7654800431499461,
+      "grad_norm": 1.111660122871399,
+      "learning_rate": 9.934491685015174e-05,
+      "loss": 0.7333855628967285,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7663430420711974,
+      "grad_norm": 1.2879043817520142,
+      "learning_rate": 9.934094939192508e-05,
+      "loss": 0.7472840547561646,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7672060409924487,
+      "grad_norm": 1.3618478775024414,
+      "learning_rate": 9.933697003535094e-05,
+      "loss": 0.7644904255867004,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7680690399137001,
+      "grad_norm": 1.1235666275024414,
+      "learning_rate": 9.933297878138887e-05,
+      "loss": 0.7872052192687988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7689320388349514,
+      "grad_norm": 1.3741775751113892,
+      "learning_rate": 9.932897563100136e-05,
+      "loss": 0.7343124747276306,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7697950377562028,
+      "grad_norm": 1.2115039825439453,
+      "learning_rate": 9.932496058515375e-05,
+      "loss": 0.7486474514007568,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7706580366774541,
+      "grad_norm": 1.183457851409912,
+      "learning_rate": 9.932093364481426e-05,
+      "loss": 0.7333025932312012,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7715210355987056,
+      "grad_norm": 1.2841869592666626,
+      "learning_rate": 9.931689481095394e-05,
+      "loss": 0.7241485118865967,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7723840345199569,
+      "grad_norm": 1.2312977313995361,
+      "learning_rate": 9.931284408454678e-05,
+      "loss": 0.7510396242141724,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7732470334412082,
+      "grad_norm": 1.144895315170288,
+      "learning_rate": 9.930878146656956e-05,
+      "loss": 0.7108346819877625,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7741100323624596,
+      "grad_norm": 1.18603515625,
+      "learning_rate": 9.9304706958002e-05,
+      "loss": 0.7687607407569885,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7749730312837109,
+      "grad_norm": 1.2827935218811035,
+      "learning_rate": 9.93006205598266e-05,
+      "loss": 0.7826682329177856,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7758360302049623,
+      "grad_norm": 1.164852261543274,
+      "learning_rate": 9.929652227302881e-05,
+      "loss": 0.6970428824424744,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 1.0439956188201904,
+      "learning_rate": 9.92924120985969e-05,
+      "loss": 0.7238826751708984,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "eval_loss": 0.7598537802696228,
+      "eval_runtime": 838.9645,
+      "eval_samples_per_second": 2.455,
+      "eval_steps_per_second": 2.455,
+      "step": 1800
+    },
+    {
+      "epoch": 0.777562028047465,
+      "grad_norm": 1.2464938163757324,
+      "learning_rate": 9.928829003752202e-05,
+      "loss": 0.7723641395568848,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7784250269687163,
+      "grad_norm": 1.2445050477981567,
+      "learning_rate": 9.92841560907982e-05,
+      "loss": 0.7154937982559204,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7792880258899676,
+      "grad_norm": 1.2235661745071411,
+      "learning_rate": 9.928001025942231e-05,
+      "loss": 0.8325800895690918,
+      "step": 1806
+    },
+    {
+      "epoch": 0.780151024811219,
+      "grad_norm": 1.268066644668579,
+      "learning_rate": 9.927585254439409e-05,
+      "loss": 0.7093122601509094,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7810140237324703,
+      "grad_norm": 1.1731390953063965,
+      "learning_rate": 9.927168294671618e-05,
+      "loss": 0.7601775527000427,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7818770226537217,
+      "grad_norm": 1.0685608386993408,
+      "learning_rate": 9.926750146739403e-05,
+      "loss": 0.6889236569404602,
+      "step": 1812
+    },
+    {
+      "epoch": 0.782740021574973,
+      "grad_norm": 1.1098639965057373,
+      "learning_rate": 9.926330810743602e-05,
+      "loss": 0.7397196888923645,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7836030204962244,
+      "grad_norm": 1.3606436252593994,
+      "learning_rate": 9.925910286785332e-05,
+      "loss": 0.7180029153823853,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7844660194174757,
+      "grad_norm": 1.143418550491333,
+      "learning_rate": 9.925488574966004e-05,
+      "loss": 0.7039036750793457,
+      "step": 1818
+    },
+    {
+      "epoch": 0.785329018338727,
+      "grad_norm": 1.180968999862671,
+      "learning_rate": 9.92506567538731e-05,
+      "loss": 0.741872251033783,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7861920172599784,
+      "grad_norm": 1.1953548192977905,
+      "learning_rate": 9.924641588151231e-05,
+      "loss": 0.7580198049545288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7870550161812297,
+      "grad_norm": 1.3049067258834839,
+      "learning_rate": 9.924216313360033e-05,
+      "loss": 0.7479216456413269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7879180151024812,
+      "grad_norm": 1.3962669372558594,
+      "learning_rate": 9.92378985111627e-05,
+      "loss": 0.7382357716560364,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7887810140237325,
+      "grad_norm": 1.1519217491149902,
+      "learning_rate": 9.92336220152278e-05,
+      "loss": 0.6356467604637146,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7896440129449838,
+      "grad_norm": 1.2544089555740356,
+      "learning_rate": 9.922933364682689e-05,
+      "loss": 0.7674193978309631,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7905070118662352,
+      "grad_norm": 1.1358295679092407,
+      "learning_rate": 9.922503340699412e-05,
+      "loss": 0.7083898782730103,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7913700107874865,
+      "grad_norm": 1.1744481325149536,
+      "learning_rate": 9.922072129676644e-05,
+      "loss": 0.6273066401481628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7922330097087379,
+      "grad_norm": 1.20509672164917,
+      "learning_rate": 9.92163973171837e-05,
+      "loss": 0.808181643486023,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7930960086299892,
+      "grad_norm": 1.2035597562789917,
+      "learning_rate": 9.921206146928863e-05,
+      "loss": 0.6631553769111633,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7939590075512406,
+      "grad_norm": 1.058695912361145,
+      "learning_rate": 9.920771375412678e-05,
+      "loss": 0.72016841173172,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7948220064724919,
+      "grad_norm": 1.035791039466858,
+      "learning_rate": 9.92033541727466e-05,
+      "loss": 0.7240060567855835,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7956850053937432,
+      "grad_norm": 1.1845753192901611,
+      "learning_rate": 9.919898272619935e-05,
+      "loss": 0.743224561214447,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7965480043149946,
+      "grad_norm": 1.2942636013031006,
+      "learning_rate": 9.919459941553922e-05,
+      "loss": 0.7064419388771057,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7974110032362459,
+      "grad_norm": 1.2736715078353882,
+      "learning_rate": 9.91902042418232e-05,
+      "loss": 0.805932343006134,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7982740021574973,
+      "grad_norm": 1.2803614139556885,
+      "learning_rate": 9.918579720611119e-05,
+      "loss": 0.8050802946090698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7991370010787486,
+      "grad_norm": 1.2225072383880615,
+      "learning_rate": 9.91813783094659e-05,
+      "loss": 0.7013180255889893,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3375571966171265,
+      "learning_rate": 9.917694755295294e-05,
+      "loss": 0.7415283918380737,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8008629989212513,
+      "grad_norm": 1.2209177017211914,
+      "learning_rate": 9.917250493764079e-05,
+      "loss": 0.7264692783355713,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8017259978425026,
+      "grad_norm": 1.1158864498138428,
+      "learning_rate": 9.916805046460074e-05,
+      "loss": 0.7083719372749329,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8025889967637541,
+      "grad_norm": 1.171000361442566,
+      "learning_rate": 9.916358413490696e-05,
+      "loss": 0.820396900177002,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8034519956850054,
+      "grad_norm": 0.9942941069602966,
+      "learning_rate": 9.91591059496365e-05,
+      "loss": 0.6754014492034912,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8043149946062568,
+      "grad_norm": 1.239795207977295,
+      "learning_rate": 9.915461590986926e-05,
+      "loss": 0.813154399394989,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8051779935275081,
+      "grad_norm": 1.076019525527954,
+      "learning_rate": 9.915011401668796e-05,
+      "loss": 0.745580792427063,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8060409924487595,
+      "grad_norm": 1.2134130001068115,
+      "learning_rate": 9.914560027117826e-05,
+      "loss": 0.7564949989318848,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8069039913700108,
+      "grad_norm": 1.1774994134902954,
+      "learning_rate": 9.914107467442859e-05,
+      "loss": 0.665183424949646,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8077669902912621,
+      "grad_norm": 1.0606465339660645,
+      "learning_rate": 9.91365372275303e-05,
+      "loss": 0.7597744464874268,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8086299892125135,
+      "grad_norm": 1.397161841392517,
+      "learning_rate": 9.913198793157757e-05,
+      "loss": 0.7453591227531433,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8094929881337648,
+      "grad_norm": 1.250003695487976,
+      "learning_rate": 9.912742678766742e-05,
+      "loss": 0.715792179107666,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8103559870550162,
+      "grad_norm": 1.2343473434448242,
+      "learning_rate": 9.912285379689979e-05,
+      "loss": 0.7094268798828125,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8112189859762675,
+      "grad_norm": 1.453764796257019,
+      "learning_rate": 9.911826896037739e-05,
+      "loss": 0.6652816534042358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8120819848975189,
+      "grad_norm": 1.030903935432434,
+      "learning_rate": 9.911367227920587e-05,
+      "loss": 0.6764529347419739,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8129449838187702,
+      "grad_norm": 1.2725969552993774,
+      "learning_rate": 9.91090637544937e-05,
+      "loss": 0.7043389678001404,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8138079827400215,
+      "grad_norm": 1.1610769033432007,
+      "learning_rate": 9.910444338735216e-05,
+      "loss": 0.6960558295249939,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8146709816612729,
+      "grad_norm": 1.2049180269241333,
+      "learning_rate": 9.909981117889548e-05,
+      "loss": 0.8029319643974304,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 1.1671873331069946,
+      "learning_rate": 9.909516713024068e-05,
+      "loss": 0.683114767074585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8163969795037757,
+      "grad_norm": 1.247768759727478,
+      "learning_rate": 9.909051124250764e-05,
+      "loss": 0.7193703055381775,
+      "step": 1892
+    },
+    {
+      "epoch": 0.817259978425027,
+      "grad_norm": 1.1376357078552246,
+      "learning_rate": 9.908584351681911e-05,
+      "loss": 0.7957084774971008,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8181229773462784,
+      "grad_norm": 1.1918323040008545,
+      "learning_rate": 9.90811639543007e-05,
+      "loss": 0.7168112993240356,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8189859762675297,
+      "grad_norm": 1.1249650716781616,
+      "learning_rate": 9.907647255608087e-05,
+      "loss": 0.7519245147705078,
+      "step": 1898
+    },
+    {
+      "epoch": 0.819848975188781,
+      "grad_norm": 1.0688576698303223,
+      "learning_rate": 9.907176932329092e-05,
+      "loss": 0.7496176362037659,
+      "step": 1900
+    },
+    {
+      "epoch": 0.819848975188781,
+      "eval_loss": 0.7533769607543945,
+      "eval_runtime": 830.2894,
+      "eval_samples_per_second": 2.481,
+      "eval_steps_per_second": 2.481,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8207119741100324,
+      "grad_norm": 0.9264270663261414,
+      "learning_rate": 9.906705425706499e-05,
+      "loss": 0.639653205871582,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8215749730312837,
+      "grad_norm": 1.2231628894805908,
+      "learning_rate": 9.906232735854016e-05,
+      "loss": 0.7555551528930664,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8224379719525351,
+      "grad_norm": 1.3369156122207642,
+      "learning_rate": 9.905758862885624e-05,
+      "loss": 0.7113592028617859,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8233009708737864,
+      "grad_norm": 1.217238187789917,
+      "learning_rate": 9.905283806915597e-05,
+      "loss": 0.6823481321334839,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8241639697950378,
+      "grad_norm": 1.2462831735610962,
+      "learning_rate": 9.904807568058495e-05,
+      "loss": 0.7347100377082825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8250269687162891,
+      "grad_norm": 1.2507221698760986,
+      "learning_rate": 9.904330146429157e-05,
+      "loss": 0.7981513738632202,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8258899676375404,
+      "grad_norm": 1.2246814966201782,
+      "learning_rate": 9.903851542142715e-05,
+      "loss": 0.759779155254364,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8267529665587918,
+      "grad_norm": 1.1746087074279785,
+      "learning_rate": 9.903371755314579e-05,
+      "loss": 0.741083562374115,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8276159654800431,
+      "grad_norm": 1.1913539171218872,
+      "learning_rate": 9.902890786060448e-05,
+      "loss": 0.7491101622581482,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8284789644012945,
+      "grad_norm": 1.2342501878738403,
+      "learning_rate": 9.902408634496307e-05,
+      "loss": 0.7655978202819824,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8293419633225458,
+      "grad_norm": 1.075892686843872,
+      "learning_rate": 9.901925300738424e-05,
+      "loss": 0.7199848890304565,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8302049622437971,
+      "grad_norm": 1.1517611742019653,
+      "learning_rate": 9.901440784903353e-05,
+      "loss": 0.8325693011283875,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8310679611650486,
+      "grad_norm": 1.2084628343582153,
+      "learning_rate": 9.900955087107933e-05,
+      "loss": 0.7674157023429871,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8319309600862999,
+      "grad_norm": 1.334679365158081,
+      "learning_rate": 9.900468207469287e-05,
+      "loss": 0.7415181398391724,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8327939590075513,
+      "grad_norm": 1.0505887269973755,
+      "learning_rate": 9.899980146104824e-05,
+      "loss": 0.672336220741272,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8336569579288026,
+      "grad_norm": 1.415797233581543,
+      "learning_rate": 9.899490903132238e-05,
+      "loss": 0.8077969551086426,
+      "step": 1932
+    },
+    {
+      "epoch": 0.834519956850054,
+      "grad_norm": 1.1793439388275146,
+      "learning_rate": 9.899000478669507e-05,
+      "loss": 0.8053045272827148,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8353829557713053,
+      "grad_norm": 1.196115493774414,
+      "learning_rate": 9.898508872834896e-05,
+      "loss": 0.7585914731025696,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8362459546925566,
+      "grad_norm": 1.2181763648986816,
+      "learning_rate": 9.898016085746953e-05,
+      "loss": 0.691072940826416,
+      "step": 1938
+    },
+    {
+      "epoch": 0.837108953613808,
+      "grad_norm": 1.2929240465164185,
+      "learning_rate": 9.897522117524512e-05,
+      "loss": 0.7723121643066406,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8379719525350593,
+      "grad_norm": 1.2859764099121094,
+      "learning_rate": 9.89702696828669e-05,
+      "loss": 0.8236626386642456,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8388349514563107,
+      "grad_norm": 1.2043784856796265,
+      "learning_rate": 9.89653063815289e-05,
+      "loss": 0.7315511703491211,
+      "step": 1944
+    },
+    {
+      "epoch": 0.839697950377562,
+      "grad_norm": 1.2063398361206055,
+      "learning_rate": 9.896033127242802e-05,
+      "loss": 0.6707082986831665,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8405609492988134,
+      "grad_norm": 1.1906715631484985,
+      "learning_rate": 9.895534435676393e-05,
+      "loss": 0.6920633912086487,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 1.122567892074585,
+      "learning_rate": 9.895034563573926e-05,
+      "loss": 0.7134768962860107,
+      "step": 1950
+    },
+    {
+      "epoch": 0.842286947141316,
+      "grad_norm": 1.0681787729263306,
+      "learning_rate": 9.894533511055943e-05,
+      "loss": 0.6333967447280884,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8431499460625674,
+      "grad_norm": 1.5216220617294312,
+      "learning_rate": 9.894031278243266e-05,
+      "loss": 0.7744026780128479,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8440129449838187,
+      "grad_norm": 1.0934535264968872,
+      "learning_rate": 9.89352786525701e-05,
+      "loss": 0.7678086757659912,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8448759439050701,
+      "grad_norm": 1.223285436630249,
+      "learning_rate": 9.89302327221857e-05,
+      "loss": 0.7344630360603333,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8457389428263214,
+      "grad_norm": 1.1123322248458862,
+      "learning_rate": 9.892517499249623e-05,
+      "loss": 0.76033616065979,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8466019417475729,
+      "grad_norm": 1.0472193956375122,
+      "learning_rate": 9.89201054647214e-05,
+      "loss": 0.7176017761230469,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8474649406688242,
+      "grad_norm": 1.8538368940353394,
+      "learning_rate": 9.891502414008365e-05,
+      "loss": 0.7220107316970825,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8483279395900755,
+      "grad_norm": 1.2280668020248413,
+      "learning_rate": 9.890993101980834e-05,
+      "loss": 0.7399556636810303,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8491909385113269,
+      "grad_norm": 1.1482552289962769,
+      "learning_rate": 9.890482610512364e-05,
+      "loss": 0.7700466513633728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8500539374325782,
+      "grad_norm": 1.1877986192703247,
+      "learning_rate": 9.889970939726061e-05,
+      "loss": 0.6666419506072998,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8509169363538296,
+      "grad_norm": 1.407792329788208,
+      "learning_rate": 9.88945808974531e-05,
+      "loss": 0.7883278131484985,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8517799352750809,
+      "grad_norm": 1.1426323652267456,
+      "learning_rate": 9.888944060693781e-05,
+      "loss": 0.7441611289978027,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8526429341963323,
+      "grad_norm": 1.2660622596740723,
+      "learning_rate": 9.888428852695432e-05,
+      "loss": 0.6464171409606934,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8535059331175836,
+      "grad_norm": 1.1639022827148438,
+      "learning_rate": 9.887912465874502e-05,
+      "loss": 0.7242478728294373,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 1.2546422481536865,
+      "learning_rate": 9.887394900355516e-05,
+      "loss": 0.7431875467300415,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8552319309600863,
+      "grad_norm": 1.1057130098342896,
+      "learning_rate": 9.886876156263282e-05,
+      "loss": 0.7662370204925537,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8560949298813376,
+      "grad_norm": 1.1017125844955444,
+      "learning_rate": 9.886356233722894e-05,
+      "loss": 0.7120938897132874,
+      "step": 1984
+    },
+    {
+      "epoch": 0.856957928802589,
+      "grad_norm": 1.3548704385757446,
+      "learning_rate": 9.885835132859727e-05,
+      "loss": 0.789219856262207,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8578209277238403,
+      "grad_norm": 1.1920398473739624,
+      "learning_rate": 9.885312853799443e-05,
+      "loss": 0.7599530816078186,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8586839266450917,
+      "grad_norm": 1.170946478843689,
+      "learning_rate": 9.884789396667988e-05,
+      "loss": 0.8315809369087219,
+      "step": 1990
+    },
+    {
+      "epoch": 0.859546925566343,
+      "grad_norm": 1.2530180215835571,
+      "learning_rate": 9.88426476159159e-05,
+      "loss": 0.713994026184082,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8604099244875943,
+      "grad_norm": 1.1056885719299316,
+      "learning_rate": 9.883738948696765e-05,
+      "loss": 0.6656507849693298,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8612729234088458,
+      "grad_norm": 1.169382929801941,
+      "learning_rate": 9.883211958110308e-05,
+      "loss": 0.7164257764816284,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8621359223300971,
+      "grad_norm": 1.1970325708389282,
+      "learning_rate": 9.882683789959299e-05,
+      "loss": 0.6885558366775513,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "grad_norm": 1.1700761318206787,
+      "learning_rate": 9.882154444371107e-05,
+      "loss": 0.7265812754631042,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8629989212513485,
+      "eval_loss": 0.7487027645111084,
+      "eval_runtime": 825.141,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8638619201725998,
+      "grad_norm": 1.3120243549346924,
+      "learning_rate": 9.88162392147338e-05,
+      "loss": 0.7371920347213745,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8647249190938512,
+      "grad_norm": 1.098697304725647,
+      "learning_rate": 9.881092221394049e-05,
+      "loss": 0.7453275918960571,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8655879180151025,
+      "grad_norm": 1.1961677074432373,
+      "learning_rate": 9.880559344261334e-05,
+      "loss": 0.7562432289123535,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8664509169363538,
+      "grad_norm": 1.1655291318893433,
+      "learning_rate": 9.880025290203735e-05,
+      "loss": 0.7654905319213867,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8673139158576052,
+      "grad_norm": 1.0979351997375488,
+      "learning_rate": 9.879490059350036e-05,
+      "loss": 0.6807888746261597,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8681769147788565,
+      "grad_norm": 1.0378475189208984,
+      "learning_rate": 9.878953651829306e-05,
+      "loss": 0.6887364387512207,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8690399137001079,
+      "grad_norm": 2.0228428840637207,
+      "learning_rate": 9.878416067770898e-05,
+      "loss": 0.718578577041626,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8699029126213592,
+      "grad_norm": 1.1610580682754517,
+      "learning_rate": 9.877877307304446e-05,
+      "loss": 0.7424600720405579,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8707659115426106,
+      "grad_norm": 1.1415833234786987,
+      "learning_rate": 9.877337370559873e-05,
+      "loss": 0.7450630068778992,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8716289104638619,
+      "grad_norm": 1.1008672714233398,
+      "learning_rate": 9.876796257667379e-05,
+      "loss": 0.7398725748062134,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8724919093851132,
+      "grad_norm": 1.1662763357162476,
+      "learning_rate": 9.876253968757452e-05,
+      "loss": 0.7122271060943604,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8733549083063646,
+      "grad_norm": 1.1835166215896606,
+      "learning_rate": 9.875710503960862e-05,
+      "loss": 0.7026485800743103,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8742179072276159,
+      "grad_norm": 1.1602214574813843,
+      "learning_rate": 9.875165863408666e-05,
+      "loss": 0.7211011052131653,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8750809061488674,
+      "grad_norm": 1.222589373588562,
+      "learning_rate": 9.874620047232198e-05,
+      "loss": 0.7565322518348694,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8759439050701187,
+      "grad_norm": 1.203661561012268,
+      "learning_rate": 9.87407305556308e-05,
+      "loss": 0.7173075675964355,
+      "step": 2030
+    },
+    {
+      "epoch": 0.87680690399137,
+      "grad_norm": 1.3316307067871094,
+      "learning_rate": 9.87352488853322e-05,
+      "loss": 0.820809006690979,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8776699029126214,
+      "grad_norm": 1.3071057796478271,
+      "learning_rate": 9.872975546274799e-05,
+      "loss": 0.8689352869987488,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8785329018338727,
+      "grad_norm": 1.3557701110839844,
+      "learning_rate": 9.872425028920294e-05,
+      "loss": 0.7090175151824951,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8793959007551241,
+      "grad_norm": 1.2234264612197876,
+      "learning_rate": 9.871873336602459e-05,
+      "loss": 0.6883202195167542,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8802588996763754,
+      "grad_norm": 1.3064714670181274,
+      "learning_rate": 9.87132046945433e-05,
+      "loss": 0.7134175896644592,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8811218985976268,
+      "grad_norm": 1.2502511739730835,
+      "learning_rate": 9.870766427609232e-05,
+      "loss": 0.7127227783203125,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8819848975188781,
+      "grad_norm": 1.1354259252548218,
+      "learning_rate": 9.870211211200766e-05,
+      "loss": 0.719437301158905,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8828478964401294,
+      "grad_norm": 1.1706442832946777,
+      "learning_rate": 9.869654820362821e-05,
+      "loss": 0.6933091282844543,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8837108953613808,
+      "grad_norm": 1.0227864980697632,
+      "learning_rate": 9.86909725522957e-05,
+      "loss": 0.6825079321861267,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8845738942826321,
+      "grad_norm": 1.103418231010437,
+      "learning_rate": 9.868538515935465e-05,
+      "loss": 0.7306133508682251,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8854368932038835,
+      "grad_norm": 1.0938395261764526,
+      "learning_rate": 9.867978602615245e-05,
+      "loss": 0.6956495046615601,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8862998921251348,
+      "grad_norm": 1.115692377090454,
+      "learning_rate": 9.86741751540393e-05,
+      "loss": 0.7740456461906433,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8871628910463862,
+      "grad_norm": 1.0264382362365723,
+      "learning_rate": 9.866855254436824e-05,
+      "loss": 0.6996002197265625,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8880258899676375,
+      "grad_norm": 1.2523163557052612,
+      "learning_rate": 9.866291819849511e-05,
+      "loss": 0.760357677936554,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.0595446825027466,
+      "learning_rate": 9.865727211777864e-05,
+      "loss": 0.6389813423156738,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8897518878101403,
+      "grad_norm": 1.198530912399292,
+      "learning_rate": 9.865161430358036e-05,
+      "loss": 0.7859060168266296,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8906148867313916,
+      "grad_norm": 1.0271193981170654,
+      "learning_rate": 9.864594475726461e-05,
+      "loss": 0.7120063900947571,
+      "step": 2064
+    },
+    {
+      "epoch": 0.891477885652643,
+      "grad_norm": 1.1619046926498413,
+      "learning_rate": 9.864026348019857e-05,
+      "loss": 0.6751878261566162,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8923408845738943,
+      "grad_norm": 1.2812464237213135,
+      "learning_rate": 9.863457047375227e-05,
+      "loss": 0.610360860824585,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 1.112628698348999,
+      "learning_rate": 9.862886573929855e-05,
+      "loss": 0.6540104746818542,
+      "step": 2070
+    },
+    {
+      "epoch": 0.894066882416397,
+      "grad_norm": 1.6257048845291138,
+      "learning_rate": 9.862314927821306e-05,
+      "loss": 0.726277232170105,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8949298813376483,
+      "grad_norm": 1.227334976196289,
+      "learning_rate": 9.861742109187432e-05,
+      "loss": 0.7798765301704407,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8957928802588997,
+      "grad_norm": 1.0728148221969604,
+      "learning_rate": 9.861168118166366e-05,
+      "loss": 0.637127161026001,
+      "step": 2076
+    },
+    {
+      "epoch": 0.896655879180151,
+      "grad_norm": 1.220282793045044,
+      "learning_rate": 9.860592954896523e-05,
+      "loss": 0.7332831025123596,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8975188781014024,
+      "grad_norm": 1.2287596464157104,
+      "learning_rate": 9.8600166195166e-05,
+      "loss": 0.7337175607681274,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8983818770226537,
+      "grad_norm": 1.0701712369918823,
+      "learning_rate": 9.859439112165579e-05,
+      "loss": 0.7422571182250977,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8992448759439051,
+      "grad_norm": 1.1583118438720703,
+      "learning_rate": 9.858860432982721e-05,
+      "loss": 0.7010003328323364,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9001078748651564,
+      "grad_norm": 0.9644797444343567,
+      "learning_rate": 9.858280582107575e-05,
+      "loss": 0.6211312413215637,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9009708737864077,
+      "grad_norm": 1.2448554039001465,
+      "learning_rate": 9.857699559679968e-05,
+      "loss": 0.7299860715866089,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9018338727076591,
+      "grad_norm": 1.2735580205917358,
+      "learning_rate": 9.85711736584001e-05,
+      "loss": 0.7461655735969543,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9026968716289104,
+      "grad_norm": 1.36841881275177,
+      "learning_rate": 9.856534000728093e-05,
+      "loss": 0.7218145132064819,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9035598705501618,
+      "grad_norm": 1.1471478939056396,
+      "learning_rate": 9.855949464484898e-05,
+      "loss": 0.7045709490776062,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9044228694714131,
+      "grad_norm": 1.058254361152649,
+      "learning_rate": 9.85536375725138e-05,
+      "loss": 0.7189744710922241,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9052858683926646,
+      "grad_norm": 1.4070709943771362,
+      "learning_rate": 9.854776879168779e-05,
+      "loss": 0.740775465965271,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 1.2182649374008179,
+      "learning_rate": 9.854188830378622e-05,
+      "loss": 0.8191262483596802,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "eval_loss": 0.7449111342430115,
+      "eval_runtime": 832.5135,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9070118662351672,
+      "grad_norm": 1.139978289604187,
+      "learning_rate": 9.853599611022708e-05,
+      "loss": 0.7295873761177063,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9078748651564186,
+      "grad_norm": 1.1802685260772705,
+      "learning_rate": 9.853009221243127e-05,
+      "loss": 0.7141753435134888,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9087378640776699,
+      "grad_norm": 1.076759934425354,
+      "learning_rate": 9.852417661182253e-05,
+      "loss": 0.7557607293128967,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9096008629989213,
+      "grad_norm": 1.1000943183898926,
+      "learning_rate": 9.851824930982733e-05,
+      "loss": 0.7025325298309326,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9104638619201726,
+      "grad_norm": 1.0235389471054077,
+      "learning_rate": 9.851231030787504e-05,
+      "loss": 0.6967638731002808,
+      "step": 2110
+    },
+    {
+      "epoch": 0.911326860841424,
+      "grad_norm": 1.2311428785324097,
+      "learning_rate": 9.850635960739782e-05,
+      "loss": 0.7890311479568481,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9121898597626753,
+      "grad_norm": 1.173770785331726,
+      "learning_rate": 9.850039720983064e-05,
+      "loss": 0.6824729442596436,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9130528586839266,
+      "grad_norm": 1.2165251970291138,
+      "learning_rate": 9.849442311661131e-05,
+      "loss": 0.7157722115516663,
+      "step": 2116
+    },
+    {
+      "epoch": 0.913915857605178,
+      "grad_norm": 1.251520037651062,
+      "learning_rate": 9.848843732918047e-05,
+      "loss": 0.7164111733436584,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9147788565264293,
+      "grad_norm": 1.1330616474151611,
+      "learning_rate": 9.848243984898153e-05,
+      "loss": 0.6830796003341675,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9156418554476807,
+      "grad_norm": 1.1052451133728027,
+      "learning_rate": 9.847643067746081e-05,
+      "loss": 0.7149171829223633,
+      "step": 2122
+    },
+    {
+      "epoch": 0.916504854368932,
+      "grad_norm": 1.0389703512191772,
+      "learning_rate": 9.847040981606735e-05,
+      "loss": 0.7213287949562073,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9173678532901833,
+      "grad_norm": 1.151895523071289,
+      "learning_rate": 9.846437726625308e-05,
+      "loss": 0.703312873840332,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9182308522114347,
+      "grad_norm": 1.1624029874801636,
+      "learning_rate": 9.845833302947269e-05,
+      "loss": 0.8059951066970825,
+      "step": 2128
+    },
+    {
+      "epoch": 0.919093851132686,
+      "grad_norm": 1.2247511148452759,
+      "learning_rate": 9.845227710718377e-05,
+      "loss": 0.735476553440094,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9199568500539375,
+      "grad_norm": 1.1649967432022095,
+      "learning_rate": 9.844620950084661e-05,
+      "loss": 0.8130571842193604,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9208198489751888,
+      "grad_norm": 1.1915717124938965,
+      "learning_rate": 9.844013021192447e-05,
+      "loss": 0.6908681392669678,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9216828478964402,
+      "grad_norm": 1.211199164390564,
+      "learning_rate": 9.843403924188328e-05,
+      "loss": 0.652173638343811,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9225458468176915,
+      "grad_norm": 1.267242193222046,
+      "learning_rate": 9.842793659219186e-05,
+      "loss": 0.6672233939170837,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9234088457389428,
+      "grad_norm": 1.7038761377334595,
+      "learning_rate": 9.842182226432186e-05,
+      "loss": 0.7274143695831299,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9242718446601942,
+      "grad_norm": 1.151426076889038,
+      "learning_rate": 9.841569625974769e-05,
+      "loss": 0.707998514175415,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9251348435814455,
+      "grad_norm": 1.0414749383926392,
+      "learning_rate": 9.840955857994663e-05,
+      "loss": 0.6639942526817322,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9259978425026969,
+      "grad_norm": 1.1045551300048828,
+      "learning_rate": 9.840340922639875e-05,
+      "loss": 0.7096070647239685,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9268608414239482,
+      "grad_norm": 1.1193996667861938,
+      "learning_rate": 9.839724820058695e-05,
+      "loss": 0.71246737241745,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9277238403451996,
+      "grad_norm": 1.2609117031097412,
+      "learning_rate": 9.839107550399692e-05,
+      "loss": 0.7351052761077881,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9285868392664509,
+      "grad_norm": 1.1127749681472778,
+      "learning_rate": 9.838489113811717e-05,
+      "loss": 0.7125008702278137,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9294498381877022,
+      "grad_norm": 1.4771665334701538,
+      "learning_rate": 9.837869510443905e-05,
+      "loss": 0.7891085743904114,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9303128371089536,
+      "grad_norm": 1.0701470375061035,
+      "learning_rate": 9.83724874044567e-05,
+      "loss": 0.7370918393135071,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9311758360302049,
+      "grad_norm": 1.2045820951461792,
+      "learning_rate": 9.836626803966708e-05,
+      "loss": 0.7543193697929382,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.1944687366485596,
+      "learning_rate": 9.836003701156996e-05,
+      "loss": 0.6903626322746277,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9329018338727076,
+      "grad_norm": 1.1035394668579102,
+      "learning_rate": 9.83537943216679e-05,
+      "loss": 0.6923050880432129,
+      "step": 2162
+    },
+    {
+      "epoch": 0.933764832793959,
+      "grad_norm": 1.3766103982925415,
+      "learning_rate": 9.834753997146633e-05,
+      "loss": 0.8099993467330933,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9346278317152104,
+      "grad_norm": 1.0874576568603516,
+      "learning_rate": 9.834127396247346e-05,
+      "loss": 0.7896219491958618,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9354908306364617,
+      "grad_norm": 1.1633459329605103,
+      "learning_rate": 9.833499629620028e-05,
+      "loss": 0.7659367918968201,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9363538295577131,
+      "grad_norm": 1.405423641204834,
+      "learning_rate": 9.832870697416066e-05,
+      "loss": 0.7474715709686279,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9372168284789644,
+      "grad_norm": 1.1794270277023315,
+      "learning_rate": 9.83224059978712e-05,
+      "loss": 0.7018750905990601,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9380798274002158,
+      "grad_norm": 1.082466959953308,
+      "learning_rate": 9.831609336885139e-05,
+      "loss": 0.6477128267288208,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9389428263214671,
+      "grad_norm": 0.9830297827720642,
+      "learning_rate": 9.830976908862346e-05,
+      "loss": 0.6516029238700867,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9398058252427185,
+      "grad_norm": 1.112289309501648,
+      "learning_rate": 9.83034331587125e-05,
+      "loss": 0.6793080568313599,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9406688241639698,
+      "grad_norm": 1.1212154626846313,
+      "learning_rate": 9.829708558064638e-05,
+      "loss": 0.7220560312271118,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9415318230852211,
+      "grad_norm": 1.246471643447876,
+      "learning_rate": 9.829072635595582e-05,
+      "loss": 0.7621577978134155,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9423948220064725,
+      "grad_norm": 1.1732288599014282,
+      "learning_rate": 9.828435548617428e-05,
+      "loss": 0.6988050937652588,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9432578209277238,
+      "grad_norm": 1.1767359972000122,
+      "learning_rate": 9.827797297283808e-05,
+      "loss": 0.7909345626831055,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9441208198489752,
+      "grad_norm": 1.007215976715088,
+      "learning_rate": 9.827157881748633e-05,
+      "loss": 0.6651493310928345,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9449838187702265,
+      "grad_norm": 1.184118628501892,
+      "learning_rate": 9.826517302166096e-05,
+      "loss": 0.7742198705673218,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9458468176914779,
+      "grad_norm": 1.1227030754089355,
+      "learning_rate": 9.82587555869067e-05,
+      "loss": 0.7614272236824036,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9467098166127292,
+      "grad_norm": 1.1030738353729248,
+      "learning_rate": 9.825232651477109e-05,
+      "loss": 0.7302108407020569,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9475728155339805,
+      "grad_norm": 1.5209563970565796,
+      "learning_rate": 9.824588580680445e-05,
+      "loss": 0.7408936023712158,
+      "step": 2196
+    },
+    {
+      "epoch": 0.948435814455232,
+      "grad_norm": 1.1853431463241577,
+      "learning_rate": 9.823943346455995e-05,
+      "loss": 0.8050196766853333,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "grad_norm": 1.285551905632019,
+      "learning_rate": 9.823296948959354e-05,
+      "loss": 0.726731538772583,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9492988133764833,
+      "eval_loss": 0.7400083541870117,
+      "eval_runtime": 838.7633,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9501618122977347,
+      "grad_norm": 1.2159804105758667,
+      "learning_rate": 9.822649388346396e-05,
+      "loss": 0.7061861753463745,
+      "step": 2202
+    },
+    {
+      "epoch": 0.951024811218986,
+      "grad_norm": 1.1189815998077393,
+      "learning_rate": 9.822000664773278e-05,
+      "loss": 0.824293851852417,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9518878101402374,
+      "grad_norm": 0.9968506097793579,
+      "learning_rate": 9.821350778396439e-05,
+      "loss": 0.7230689525604248,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9527508090614887,
+      "grad_norm": 1.101393222808838,
+      "learning_rate": 9.820699729372593e-05,
+      "loss": 0.6633790135383606,
+      "step": 2208
+    },
+    {
+      "epoch": 0.95361380798274,
+      "grad_norm": 1.084307312965393,
+      "learning_rate": 9.820047517858741e-05,
+      "loss": 0.6548227071762085,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9544768069039914,
+      "grad_norm": 1.1291496753692627,
+      "learning_rate": 9.819394144012158e-05,
+      "loss": 0.6434295177459717,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9553398058252427,
+      "grad_norm": 2.303483724594116,
+      "learning_rate": 9.818739607990404e-05,
+      "loss": 0.8122013211250305,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9562028047464941,
+      "grad_norm": 1.244541883468628,
+      "learning_rate": 9.818083909951315e-05,
+      "loss": 0.6120955348014832,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9570658036677454,
+      "grad_norm": 1.1051502227783203,
+      "learning_rate": 9.817427050053012e-05,
+      "loss": 0.7569628953933716,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9579288025889967,
+      "grad_norm": 1.3161911964416504,
+      "learning_rate": 9.816769028453893e-05,
+      "loss": 0.7588980197906494,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9587918015102481,
+      "grad_norm": 1.2526928186416626,
+      "learning_rate": 9.816109845312636e-05,
+      "loss": 0.743370771408081,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9596548004314994,
+      "grad_norm": 1.196215033531189,
+      "learning_rate": 9.815449500788202e-05,
+      "loss": 0.80516517162323,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9605177993527508,
+      "grad_norm": 1.3204832077026367,
+      "learning_rate": 9.81478799503983e-05,
+      "loss": 0.7331632971763611,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9613807982740021,
+      "grad_norm": 1.2442790269851685,
+      "learning_rate": 9.814125328227035e-05,
+      "loss": 0.6810339093208313,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9622437971952535,
+      "grad_norm": 1.1739158630371094,
+      "learning_rate": 9.813461500509623e-05,
+      "loss": 0.7269608378410339,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9631067961165048,
+      "grad_norm": 1.1016241312026978,
+      "learning_rate": 9.812796512047666e-05,
+      "loss": 0.7159227132797241,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9639697950377561,
+      "grad_norm": 1.1654549837112427,
+      "learning_rate": 9.812130363001529e-05,
+      "loss": 0.7145758867263794,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9648327939590076,
+      "grad_norm": 1.0521938800811768,
+      "learning_rate": 9.811463053531846e-05,
+      "loss": 0.6873835325241089,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9656957928802589,
+      "grad_norm": 1.2285776138305664,
+      "learning_rate": 9.810794583799538e-05,
+      "loss": 0.6683027148246765,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9665587918015103,
+      "grad_norm": 0.9826368689537048,
+      "learning_rate": 9.810124953965805e-05,
+      "loss": 0.663141131401062,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9674217907227616,
+      "grad_norm": 1.0102336406707764,
+      "learning_rate": 9.809454164192123e-05,
+      "loss": 0.6688778400421143,
+      "step": 2242
+    },
+    {
+      "epoch": 0.968284789644013,
+      "grad_norm": 0.9781575202941895,
+      "learning_rate": 9.808782214640251e-05,
+      "loss": 0.671240508556366,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9691477885652643,
+      "grad_norm": 1.137385368347168,
+      "learning_rate": 9.808109105472225e-05,
+      "loss": 0.7673252820968628,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9700107874865156,
+      "grad_norm": 1.3359450101852417,
+      "learning_rate": 9.807434836850362e-05,
+      "loss": 0.8085554838180542,
+      "step": 2248
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 1.150694727897644,
+      "learning_rate": 9.806759408937262e-05,
+      "loss": 0.6750202178955078,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9717367853290183,
+      "grad_norm": 1.080316424369812,
+      "learning_rate": 9.8060828218958e-05,
+      "loss": 0.5928418040275574,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9725997842502697,
+      "grad_norm": 1.0489610433578491,
+      "learning_rate": 9.805405075889129e-05,
+      "loss": 0.6947471499443054,
+      "step": 2254
+    },
+    {
+      "epoch": 0.973462783171521,
+      "grad_norm": 1.199119210243225,
+      "learning_rate": 9.80472617108069e-05,
+      "loss": 0.7273924350738525,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9743257820927724,
+      "grad_norm": 0.9830896258354187,
+      "learning_rate": 9.804046107634193e-05,
+      "loss": 0.668613612651825,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9751887810140237,
+      "grad_norm": 1.1655877828598022,
+      "learning_rate": 9.803364885713633e-05,
+      "loss": 0.7255772352218628,
+      "step": 2260
+    },
+    {
+      "epoch": 0.976051779935275,
+      "grad_norm": 1.0289462804794312,
+      "learning_rate": 9.802682505483284e-05,
+      "loss": 0.6865858435630798,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9769147788565264,
+      "grad_norm": 1.1020126342773438,
+      "learning_rate": 9.801998967107702e-05,
+      "loss": 0.6680663228034973,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 1.0846027135849,
+      "learning_rate": 9.801314270751716e-05,
+      "loss": 0.7476468086242676,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9786407766990292,
+      "grad_norm": 1.1834278106689453,
+      "learning_rate": 9.800628416580437e-05,
+      "loss": 0.7173228859901428,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9795037756202805,
+      "grad_norm": 1.0082321166992188,
+      "learning_rate": 9.799941404759257e-05,
+      "loss": 0.6768618226051331,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9803667745415319,
+      "grad_norm": 1.1573102474212646,
+      "learning_rate": 9.799253235453846e-05,
+      "loss": 0.7877311110496521,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9812297734627832,
+      "grad_norm": 1.2030092477798462,
+      "learning_rate": 9.798563908830152e-05,
+      "loss": 0.6843966841697693,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9820927723840345,
+      "grad_norm": 1.080560564994812,
+      "learning_rate": 9.797873425054404e-05,
+      "loss": 0.6084349155426025,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9829557713052859,
+      "grad_norm": 1.1488033533096313,
+      "learning_rate": 9.797181784293107e-05,
+      "loss": 0.6784868836402893,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9838187702265372,
+      "grad_norm": 1.144773006439209,
+      "learning_rate": 9.796488986713049e-05,
+      "loss": 0.6111500263214111,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9846817691477886,
+      "grad_norm": 1.1356254816055298,
+      "learning_rate": 9.795795032481297e-05,
+      "loss": 0.7634603977203369,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9855447680690399,
+      "grad_norm": 1.139446496963501,
+      "learning_rate": 9.79509992176519e-05,
+      "loss": 0.6470903158187866,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9864077669902913,
+      "grad_norm": 1.1564133167266846,
+      "learning_rate": 9.794403654732354e-05,
+      "loss": 0.71474289894104,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9872707659115426,
+      "grad_norm": 1.0605655908584595,
+      "learning_rate": 9.793706231550689e-05,
+      "loss": 0.6413968801498413,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9881337648327939,
+      "grad_norm": 1.1660863161087036,
+      "learning_rate": 9.793007652388378e-05,
+      "loss": 0.6595705151557922,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9889967637540453,
+      "grad_norm": 1.1661914587020874,
+      "learning_rate": 9.792307917413877e-05,
+      "loss": 0.7144066095352173,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9898597626752966,
+      "grad_norm": 1.2250727415084839,
+      "learning_rate": 9.791607026795926e-05,
+      "loss": 0.7956018447875977,
+      "step": 2294
+    },
+    {
+      "epoch": 0.990722761596548,
+      "grad_norm": 0.9593650698661804,
+      "learning_rate": 9.790904980703541e-05,
+      "loss": 0.5786031484603882,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9915857605177993,
+      "grad_norm": 1.0697777271270752,
+      "learning_rate": 9.790201779306018e-05,
+      "loss": 0.6817460656166077,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "grad_norm": 1.2658408880233765,
+      "learning_rate": 9.789497422772931e-05,
+      "loss": 0.6845148801803589,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9924487594390508,
+      "eval_loss": 0.7360134720802307,
+      "eval_runtime": 825.8559,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2300
+    },
+    {
+      "epoch": 0.993311758360302,
+      "grad_norm": 1.0869544744491577,
+      "learning_rate": 9.788791911274131e-05,
+      "loss": 0.6689707040786743,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9941747572815534,
+      "grad_norm": 1.0049875974655151,
+      "learning_rate": 9.788085244979748e-05,
+      "loss": 0.6470839381217957,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9950377562028048,
+      "grad_norm": 1.4471046924591064,
+      "learning_rate": 9.787377424060196e-05,
+      "loss": 0.8018016815185547,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9959007551240561,
+      "grad_norm": 1.3750648498535156,
+      "learning_rate": 9.786668448686159e-05,
+      "loss": 0.8078322410583496,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9967637540453075,
+      "grad_norm": 1.098283290863037,
+      "learning_rate": 9.7859583190286e-05,
+      "loss": 0.6740908622741699,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9976267529665588,
+      "grad_norm": 1.0312142372131348,
+      "learning_rate": 9.785247035258773e-05,
+      "loss": 0.6844386458396912,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9984897518878102,
+      "grad_norm": 1.4482319355010986,
+      "learning_rate": 9.784534597548193e-05,
+      "loss": 0.7547479271888733,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9993527508090615,
+      "grad_norm": 1.0897735357284546,
+      "learning_rate": 9.783821006068663e-05,
+      "loss": 0.7713987231254578,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7887805700302124,
+      "learning_rate": 9.783106260992264e-05,
+      "loss": 0.6868399381637573,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0008629989212514,
+      "grad_norm": 1.0021663904190063,
+      "learning_rate": 9.782390362491353e-05,
+      "loss": 0.696355938911438,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0017259978425026,
+      "grad_norm": 1.3650091886520386,
+      "learning_rate": 9.781673310738564e-05,
+      "loss": 0.7315902709960938,
+      "step": 2322
+    },
+    {
+      "epoch": 1.002588996763754,
+      "grad_norm": 1.2247875928878784,
+      "learning_rate": 9.780955105906813e-05,
+      "loss": 0.7667547464370728,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0034519956850054,
+      "grad_norm": 1.1917641162872314,
+      "learning_rate": 9.780235748169291e-05,
+      "loss": 0.7380582094192505,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0043149946062568,
+      "grad_norm": 1.146159052848816,
+      "learning_rate": 9.779515237699466e-05,
+      "loss": 0.6780733466148376,
+      "step": 2328
+    },
+    {
+      "epoch": 1.005177993527508,
+      "grad_norm": 1.4877620935440063,
+      "learning_rate": 9.778793574671089e-05,
+      "loss": 0.8116409182548523,
+      "step": 2330
+    },
+    {
+      "epoch": 1.0060409924487594,
+      "grad_norm": 1.3488554954528809,
+      "learning_rate": 9.778070759258185e-05,
+      "loss": 0.65245521068573,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0069039913700109,
+      "grad_norm": 1.096686840057373,
+      "learning_rate": 9.777346791635056e-05,
+      "loss": 0.584334135055542,
+      "step": 2334
+    },
+    {
+      "epoch": 1.007766990291262,
+      "grad_norm": 1.063981056213379,
+      "learning_rate": 9.776621671976285e-05,
+      "loss": 0.6594714522361755,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0086299892125135,
+      "grad_norm": 1.2811797857284546,
+      "learning_rate": 9.775895400456729e-05,
+      "loss": 0.6963099837303162,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0094929881337649,
+      "grad_norm": 0.9491971731185913,
+      "learning_rate": 9.775167977251529e-05,
+      "loss": 0.6246375441551208,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0103559870550163,
+      "grad_norm": 1.198683500289917,
+      "learning_rate": 9.774439402536098e-05,
+      "loss": 0.6352161169052124,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0112189859762675,
+      "grad_norm": 1.2178813219070435,
+      "learning_rate": 9.773709676486129e-05,
+      "loss": 0.6884894967079163,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0120819848975189,
+      "grad_norm": 1.1905319690704346,
+      "learning_rate": 9.77297879927759e-05,
+      "loss": 0.5945048928260803,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0129449838187703,
+      "grad_norm": 1.210302710533142,
+      "learning_rate": 9.77224677108673e-05,
+      "loss": 0.6533923149108887,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0138079827400215,
+      "grad_norm": 1.15508234500885,
+      "learning_rate": 9.771513592090076e-05,
+      "loss": 0.6619405746459961,
+      "step": 2350
+    },
+    {
+      "epoch": 1.014670981661273,
+      "grad_norm": 1.0833762884140015,
+      "learning_rate": 9.77077926246443e-05,
+      "loss": 0.6581310033798218,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0155339805825243,
+      "grad_norm": 1.0330514907836914,
+      "learning_rate": 9.770043782386869e-05,
+      "loss": 0.6670945882797241,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0163969795037757,
+      "grad_norm": 1.1472195386886597,
+      "learning_rate": 9.769307152034755e-05,
+      "loss": 0.6801814436912537,
+      "step": 2356
+    },
+    {
+      "epoch": 1.017259978425027,
+      "grad_norm": 1.2204132080078125,
+      "learning_rate": 9.76856937158572e-05,
+      "loss": 0.7206465005874634,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0181229773462783,
+      "grad_norm": 1.2335962057113647,
+      "learning_rate": 9.76783044121768e-05,
+      "loss": 0.6884574294090271,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0189859762675297,
+      "grad_norm": 1.0858685970306396,
+      "learning_rate": 9.767090361108822e-05,
+      "loss": 0.655097484588623,
+      "step": 2362
+    },
+    {
+      "epoch": 1.019848975188781,
+      "grad_norm": 1.117907166481018,
+      "learning_rate": 9.766349131437614e-05,
+      "loss": 0.6478561758995056,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0207119741100323,
+      "grad_norm": 1.1995056867599487,
+      "learning_rate": 9.765606752382797e-05,
+      "loss": 0.6310951113700867,
+      "step": 2366
+    },
+    {
+      "epoch": 1.0215749730312838,
+      "grad_norm": 1.354819655418396,
+      "learning_rate": 9.764863224123397e-05,
+      "loss": 0.6373490691184998,
+      "step": 2368
+    },
+    {
+      "epoch": 1.0224379719525352,
+      "grad_norm": 1.2194819450378418,
+      "learning_rate": 9.764118546838712e-05,
+      "loss": 0.6595761179924011,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0233009708737864,
+      "grad_norm": 1.1485158205032349,
+      "learning_rate": 9.763372720708314e-05,
+      "loss": 0.6299480199813843,
+      "step": 2372
+    },
+    {
+      "epoch": 1.0241639697950378,
+      "grad_norm": 1.1153662204742432,
+      "learning_rate": 9.76262574591206e-05,
+      "loss": 0.6788742542266846,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0250269687162892,
+      "grad_norm": 1.0690817832946777,
+      "learning_rate": 9.761877622630075e-05,
+      "loss": 0.559056282043457,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0258899676375404,
+      "grad_norm": 1.1896858215332031,
+      "learning_rate": 9.76112835104277e-05,
+      "loss": 0.7164555191993713,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0267529665587918,
+      "grad_norm": 1.1792134046554565,
+      "learning_rate": 9.760377931330824e-05,
+      "loss": 0.6791456937789917,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0276159654800432,
+      "grad_norm": 1.361763834953308,
+      "learning_rate": 9.759626363675202e-05,
+      "loss": 0.7421526312828064,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0284789644012946,
+      "grad_norm": 1.1960662603378296,
+      "learning_rate": 9.758873648257138e-05,
+      "loss": 0.7032517194747925,
+      "step": 2384
+    },
+    {
+      "epoch": 1.0293419633225458,
+      "grad_norm": 1.323255181312561,
+      "learning_rate": 9.758119785258146e-05,
+      "loss": 0.6953344345092773,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0302049622437972,
+      "grad_norm": 1.2343937158584595,
+      "learning_rate": 9.75736477486002e-05,
+      "loss": 0.7287829518318176,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0310679611650486,
+      "grad_norm": 1.0712062120437622,
+      "learning_rate": 9.756608617244823e-05,
+      "loss": 0.6812928915023804,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0319309600862998,
+      "grad_norm": 1.264726996421814,
+      "learning_rate": 9.755851312594902e-05,
+      "loss": 0.7173963785171509,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0327939590075512,
+      "grad_norm": 1.1563389301300049,
+      "learning_rate": 9.755092861092874e-05,
+      "loss": 0.6509836316108704,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0336569579288026,
+      "grad_norm": 1.326188087463379,
+      "learning_rate": 9.75433326292164e-05,
+      "loss": 0.718437135219574,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0345199568500538,
+      "grad_norm": 1.2263392210006714,
+      "learning_rate": 9.753572518264373e-05,
+      "loss": 0.664516031742096,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "grad_norm": 1.2874478101730347,
+      "learning_rate": 9.752810627304523e-05,
+      "loss": 0.6754546165466309,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0353829557713052,
+      "eval_loss": 0.7346783876419067,
+      "eval_runtime": 825.8356,
+      "eval_samples_per_second": 2.494,
+      "eval_steps_per_second": 2.494,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0362459546925566,
+      "grad_norm": 1.304518461227417,
+      "learning_rate": 9.752047590225814e-05,
+      "loss": 0.708167314529419,
+      "step": 2402
+    },
+    {
+      "epoch": 1.037108953613808,
+      "grad_norm": 1.21835196018219,
+      "learning_rate": 9.751283407212252e-05,
+      "loss": 0.7213018536567688,
+      "step": 2404
+    },
+    {
+      "epoch": 1.0379719525350593,
+      "grad_norm": 1.2947579622268677,
+      "learning_rate": 9.750518078448116e-05,
+      "loss": 0.6739231944084167,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 1.138368010520935,
+      "learning_rate": 9.749751604117959e-05,
+      "loss": 0.7671371102333069,
+      "step": 2408
+    },
+    {
+      "epoch": 1.039697950377562,
+      "grad_norm": 1.1255316734313965,
+      "learning_rate": 9.748983984406616e-05,
+      "loss": 0.6593353748321533,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0405609492988135,
+      "grad_norm": 1.2362794876098633,
+      "learning_rate": 9.748215219499193e-05,
+      "loss": 0.713056743144989,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0414239482200647,
+      "grad_norm": 1.2021304368972778,
+      "learning_rate": 9.747445309581076e-05,
+      "loss": 0.6989394426345825,
+      "step": 2414
+    },
+    {
+      "epoch": 1.042286947141316,
+      "grad_norm": 1.127341389656067,
+      "learning_rate": 9.746674254837925e-05,
+      "loss": 0.623035192489624,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0431499460625675,
+      "grad_norm": 1.1254255771636963,
+      "learning_rate": 9.745902055455672e-05,
+      "loss": 0.6490310430526733,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0440129449838187,
+      "grad_norm": 1.2142512798309326,
+      "learning_rate": 9.745128711620535e-05,
+      "loss": 0.6210505366325378,
+      "step": 2420
+    },
+    {
+      "epoch": 1.04487594390507,
+      "grad_norm": 1.1306709051132202,
+      "learning_rate": 9.744354223519001e-05,
+      "loss": 0.6243326663970947,
+      "step": 2422
+    },
+    {
+      "epoch": 1.0457389428263215,
+      "grad_norm": 1.202621340751648,
+      "learning_rate": 9.743578591337833e-05,
+      "loss": 0.6312575340270996,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0466019417475727,
+      "grad_norm": 1.0556137561798096,
+      "learning_rate": 9.742801815264072e-05,
+      "loss": 0.6190714836120605,
+      "step": 2426
+    },
+    {
+      "epoch": 1.0474649406688241,
+      "grad_norm": 1.5071814060211182,
+      "learning_rate": 9.742023895485033e-05,
+      "loss": 0.7496671080589294,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0483279395900755,
+      "grad_norm": 1.1448185443878174,
+      "learning_rate": 9.74124483218831e-05,
+      "loss": 0.6288847327232361,
+      "step": 2430
+    },
+    {
+      "epoch": 1.049190938511327,
+      "grad_norm": 1.0072964429855347,
+      "learning_rate": 9.74046462556177e-05,
+      "loss": 0.6938424706459045,
+      "step": 2432
+    },
+    {
+      "epoch": 1.0500539374325781,
+      "grad_norm": 1.198836326599121,
+      "learning_rate": 9.739683275793554e-05,
+      "loss": 0.7529639601707458,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0509169363538295,
+      "grad_norm": 1.2170380353927612,
+      "learning_rate": 9.738900783072083e-05,
+      "loss": 0.7193960547447205,
+      "step": 2436
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 1.0384336709976196,
+      "learning_rate": 9.738117147586051e-05,
+      "loss": 0.6488150954246521,
+      "step": 2438
+    },
+    {
+      "epoch": 1.0526429341963321,
+      "grad_norm": 1.1562492847442627,
+      "learning_rate": 9.737332369524429e-05,
+      "loss": 0.6672247052192688,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0535059331175836,
+      "grad_norm": 1.1553802490234375,
+      "learning_rate": 9.736546449076462e-05,
+      "loss": 0.651012122631073,
+      "step": 2442
+    },
+    {
+      "epoch": 1.054368932038835,
+      "grad_norm": 1.136838674545288,
+      "learning_rate": 9.735759386431671e-05,
+      "loss": 0.681246817111969,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0552319309600864,
+      "grad_norm": 1.177869200706482,
+      "learning_rate": 9.734971181779854e-05,
+      "loss": 0.7331019639968872,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0560949298813376,
+      "grad_norm": 1.2983415126800537,
+      "learning_rate": 9.73418183531108e-05,
+      "loss": 0.7316935658454895,
+      "step": 2448
+    },
+    {
+      "epoch": 1.056957928802589,
+      "grad_norm": 1.1189959049224854,
+      "learning_rate": 9.733391347215699e-05,
+      "loss": 0.6513608694076538,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0578209277238404,
+      "grad_norm": 1.3256937265396118,
+      "learning_rate": 9.732599717684334e-05,
+      "loss": 0.6469123363494873,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0586839266450916,
+      "grad_norm": 0.9766954779624939,
+      "learning_rate": 9.731806946907881e-05,
+      "loss": 0.5860153436660767,
+      "step": 2454
+    },
+    {
+      "epoch": 1.059546925566343,
+      "grad_norm": 1.1360441446304321,
+      "learning_rate": 9.731013035077514e-05,
+      "loss": 0.6783872842788696,
+      "step": 2456
+    },
+    {
+      "epoch": 1.0604099244875944,
+      "grad_norm": 1.3483080863952637,
+      "learning_rate": 9.73021798238468e-05,
+      "loss": 0.7894490361213684,
+      "step": 2458
+    },
+    {
+      "epoch": 1.0612729234088458,
+      "grad_norm": 1.1412159204483032,
+      "learning_rate": 9.729421789021105e-05,
+      "loss": 0.7098361849784851,
+      "step": 2460
+    },
+    {
+      "epoch": 1.062135922330097,
+      "grad_norm": 1.2670512199401855,
+      "learning_rate": 9.728624455178785e-05,
+      "loss": 0.7138097286224365,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0629989212513484,
+      "grad_norm": 1.2577205896377563,
+      "learning_rate": 9.727825981049994e-05,
+      "loss": 0.7136340141296387,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0638619201725998,
+      "grad_norm": 1.098997712135315,
+      "learning_rate": 9.727026366827283e-05,
+      "loss": 0.6273087859153748,
+      "step": 2466
+    },
+    {
+      "epoch": 1.064724919093851,
+      "grad_norm": 1.1550003290176392,
+      "learning_rate": 9.72622561270347e-05,
+      "loss": 0.7076449394226074,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0655879180151024,
+      "grad_norm": 1.0861129760742188,
+      "learning_rate": 9.725423718871659e-05,
+      "loss": 0.6856985092163086,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0664509169363539,
+      "grad_norm": 1.2742570638656616,
+      "learning_rate": 9.724620685525219e-05,
+      "loss": 0.7311446070671082,
+      "step": 2472
+    },
+    {
+      "epoch": 1.0673139158576053,
+      "grad_norm": 1.504339337348938,
+      "learning_rate": 9.7238165128578e-05,
+      "loss": 0.7728472948074341,
+      "step": 2474
+    },
+    {
+      "epoch": 1.0681769147788565,
+      "grad_norm": 1.2232030630111694,
+      "learning_rate": 9.723011201063323e-05,
+      "loss": 0.6259000301361084,
+      "step": 2476
+    },
+    {
+      "epoch": 1.0690399137001079,
+      "grad_norm": 1.3238953351974487,
+      "learning_rate": 9.722204750335989e-05,
+      "loss": 0.7132798433303833,
+      "step": 2478
+    },
+    {
+      "epoch": 1.0699029126213593,
+      "grad_norm": 1.2286484241485596,
+      "learning_rate": 9.721397160870265e-05,
+      "loss": 0.6067986488342285,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0707659115426105,
+      "grad_norm": 1.3101580142974854,
+      "learning_rate": 9.7205884328609e-05,
+      "loss": 0.6884648203849792,
+      "step": 2482
+    },
+    {
+      "epoch": 1.0716289104638619,
+      "grad_norm": 1.437833309173584,
+      "learning_rate": 9.719778566502915e-05,
+      "loss": 0.7164831757545471,
+      "step": 2484
+    },
+    {
+      "epoch": 1.0724919093851133,
+      "grad_norm": 1.0980390310287476,
+      "learning_rate": 9.718967561991606e-05,
+      "loss": 0.6326627731323242,
+      "step": 2486
+    },
+    {
+      "epoch": 1.0733549083063647,
+      "grad_norm": 0.9919570088386536,
+      "learning_rate": 9.718155419522543e-05,
+      "loss": 0.5712647438049316,
+      "step": 2488
+    },
+    {
+      "epoch": 1.074217907227616,
+      "grad_norm": 1.0773793458938599,
+      "learning_rate": 9.717342139291569e-05,
+      "loss": 0.6923331022262573,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0750809061488673,
+      "grad_norm": 1.1553318500518799,
+      "learning_rate": 9.716527721494805e-05,
+      "loss": 0.6737485527992249,
+      "step": 2492
+    },
+    {
+      "epoch": 1.0759439050701187,
+      "grad_norm": 1.112865924835205,
+      "learning_rate": 9.715712166328644e-05,
+      "loss": 0.6422063708305359,
+      "step": 2494
+    },
+    {
+      "epoch": 1.07680690399137,
+      "grad_norm": 1.0651512145996094,
+      "learning_rate": 9.71489547398975e-05,
+      "loss": 0.6399799585342407,
+      "step": 2496
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 1.1394895315170288,
+      "learning_rate": 9.71407764467507e-05,
+      "loss": 0.6218312382698059,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "grad_norm": 1.3323278427124023,
+      "learning_rate": 9.713258678581815e-05,
+      "loss": 0.65947425365448,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0785329018338727,
+      "eval_loss": 0.7294915914535522,
+      "eval_runtime": 836.4643,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0793959007551241,
+      "grad_norm": 1.3122533559799194,
+      "learning_rate": 9.712438575907479e-05,
+      "loss": 0.66681307554245,
+      "step": 2502
+    },
+    {
+      "epoch": 1.0802588996763753,
+      "grad_norm": 1.3631974458694458,
+      "learning_rate": 9.711617336849824e-05,
+      "loss": 0.6926553249359131,
+      "step": 2504
+    },
+    {
+      "epoch": 1.0811218985976268,
+      "grad_norm": 1.2383692264556885,
+      "learning_rate": 9.710794961606886e-05,
+      "loss": 0.6706596612930298,
+      "step": 2506
+    },
+    {
+      "epoch": 1.0819848975188782,
+      "grad_norm": 1.1660504341125488,
+      "learning_rate": 9.709971450376982e-05,
+      "loss": 0.6011359095573425,
+      "step": 2508
+    },
+    {
+      "epoch": 1.0828478964401294,
+      "grad_norm": 1.3021246194839478,
+      "learning_rate": 9.709146803358693e-05,
+      "loss": 0.689009428024292,
+      "step": 2510
+    },
+    {
+      "epoch": 1.0837108953613808,
+      "grad_norm": 1.0553252696990967,
+      "learning_rate": 9.708321020750883e-05,
+      "loss": 0.6216740012168884,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0845738942826322,
+      "grad_norm": 1.1862177848815918,
+      "learning_rate": 9.707494102752683e-05,
+      "loss": 0.6439616680145264,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0854368932038836,
+      "grad_norm": 1.1166558265686035,
+      "learning_rate": 9.706666049563504e-05,
+      "loss": 0.6825361251831055,
+      "step": 2516
+    },
+    {
+      "epoch": 1.0862998921251348,
+      "grad_norm": 1.128812551498413,
+      "learning_rate": 9.705836861383022e-05,
+      "loss": 0.6471308469772339,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0871628910463862,
+      "grad_norm": 1.324802279472351,
+      "learning_rate": 9.705006538411197e-05,
+      "loss": 0.6535511016845703,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0880258899676376,
+      "grad_norm": 1.3586591482162476,
+      "learning_rate": 9.704175080848255e-05,
+      "loss": 0.678022563457489,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 1.1334232091903687,
+      "learning_rate": 9.703342488894697e-05,
+      "loss": 0.5904073119163513,
+      "step": 2524
+    },
+    {
+      "epoch": 1.0897518878101402,
+      "grad_norm": 1.182215929031372,
+      "learning_rate": 9.702508762751303e-05,
+      "loss": 0.6072601079940796,
+      "step": 2526
+    },
+    {
+      "epoch": 1.0906148867313916,
+      "grad_norm": 1.2525769472122192,
+      "learning_rate": 9.701673902619119e-05,
+      "loss": 0.6752429604530334,
+      "step": 2528
+    },
+    {
+      "epoch": 1.091477885652643,
+      "grad_norm": 1.2811026573181152,
+      "learning_rate": 9.70083790869947e-05,
+      "loss": 0.7192611694335938,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0923408845738942,
+      "grad_norm": 1.1434061527252197,
+      "learning_rate": 9.700000781193949e-05,
+      "loss": 0.6991751194000244,
+      "step": 2532
+    },
+    {
+      "epoch": 1.0932038834951456,
+      "grad_norm": 1.26658296585083,
+      "learning_rate": 9.699162520304428e-05,
+      "loss": 0.7010498046875,
+      "step": 2534
+    },
+    {
+      "epoch": 1.094066882416397,
+      "grad_norm": 1.2083529233932495,
+      "learning_rate": 9.698323126233048e-05,
+      "loss": 0.6945615410804749,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0949298813376482,
+      "grad_norm": 1.2064114809036255,
+      "learning_rate": 9.697482599182227e-05,
+      "loss": 0.689496636390686,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0957928802588996,
+      "grad_norm": 1.17113196849823,
+      "learning_rate": 9.696640939354655e-05,
+      "loss": 0.6966884136199951,
+      "step": 2540
+    },
+    {
+      "epoch": 1.096655879180151,
+      "grad_norm": 1.1551356315612793,
+      "learning_rate": 9.695798146953291e-05,
+      "loss": 0.6524666547775269,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0975188781014025,
+      "grad_norm": 1.0966521501541138,
+      "learning_rate": 9.694954222181374e-05,
+      "loss": 0.5769726037979126,
+      "step": 2544
+    },
+    {
+      "epoch": 1.0983818770226537,
+      "grad_norm": 1.320669174194336,
+      "learning_rate": 9.694109165242411e-05,
+      "loss": 0.6986333727836609,
+      "step": 2546
+    },
+    {
+      "epoch": 1.099244875943905,
+      "grad_norm": 1.1678471565246582,
+      "learning_rate": 9.693262976340183e-05,
+      "loss": 0.7014455199241638,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1001078748651565,
+      "grad_norm": 1.1527155637741089,
+      "learning_rate": 9.692415655678745e-05,
+      "loss": 0.5786730647087097,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1009708737864077,
+      "grad_norm": 1.2511217594146729,
+      "learning_rate": 9.691567203462428e-05,
+      "loss": 0.755945086479187,
+      "step": 2552
+    },
+    {
+      "epoch": 1.101833872707659,
+      "grad_norm": 1.0910221338272095,
+      "learning_rate": 9.690717619895828e-05,
+      "loss": 0.6487610936164856,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1026968716289105,
+      "grad_norm": 1.3682197332382202,
+      "learning_rate": 9.68986690518382e-05,
+      "loss": 0.6346665024757385,
+      "step": 2556
+    },
+    {
+      "epoch": 1.103559870550162,
+      "grad_norm": 1.2161564826965332,
+      "learning_rate": 9.68901505953155e-05,
+      "loss": 0.647144615650177,
+      "step": 2558
+    },
+    {
+      "epoch": 1.104422869471413,
+      "grad_norm": 1.2559088468551636,
+      "learning_rate": 9.688162083144436e-05,
+      "loss": 0.6461755633354187,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1052858683926645,
+      "grad_norm": 1.1260225772857666,
+      "learning_rate": 9.68730797622817e-05,
+      "loss": 0.6888763904571533,
+      "step": 2562
+    },
+    {
+      "epoch": 1.106148867313916,
+      "grad_norm": 1.1499868631362915,
+      "learning_rate": 9.686452738988716e-05,
+      "loss": 0.6318327784538269,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1070118662351671,
+      "grad_norm": 1.1775025129318237,
+      "learning_rate": 9.68559637163231e-05,
+      "loss": 0.6538340449333191,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1078748651564185,
+      "grad_norm": 1.264656901359558,
+      "learning_rate": 9.684738874365465e-05,
+      "loss": 0.6706726551055908,
+      "step": 2568
+    },
+    {
+      "epoch": 1.10873786407767,
+      "grad_norm": 1.1130270957946777,
+      "learning_rate": 9.683880247394956e-05,
+      "loss": 0.7573289275169373,
+      "step": 2570
+    },
+    {
+      "epoch": 1.1096008629989211,
+      "grad_norm": 1.2781217098236084,
+      "learning_rate": 9.683020490927841e-05,
+      "loss": 0.6603306531906128,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1104638619201725,
+      "grad_norm": 1.251062035560608,
+      "learning_rate": 9.682159605171446e-05,
+      "loss": 0.6523779630661011,
+      "step": 2574
+    },
+    {
+      "epoch": 1.111326860841424,
+      "grad_norm": 1.2631903886795044,
+      "learning_rate": 9.68129759033337e-05,
+      "loss": 0.677550733089447,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1121898597626754,
+      "grad_norm": 1.0212602615356445,
+      "learning_rate": 9.680434446621483e-05,
+      "loss": 0.666540801525116,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1130528586839266,
+      "grad_norm": 1.221718430519104,
+      "learning_rate": 9.679570174243929e-05,
+      "loss": 0.7467086315155029,
+      "step": 2580
+    },
+    {
+      "epoch": 1.113915857605178,
+      "grad_norm": 1.1175146102905273,
+      "learning_rate": 9.678704773409124e-05,
+      "loss": 0.673492968082428,
+      "step": 2582
+    },
+    {
+      "epoch": 1.1147788565264294,
+      "grad_norm": 1.1274988651275635,
+      "learning_rate": 9.677838244325754e-05,
+      "loss": 0.5724847912788391,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1156418554476808,
+      "grad_norm": 1.297654390335083,
+      "learning_rate": 9.676970587202779e-05,
+      "loss": 0.634985625743866,
+      "step": 2586
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 1.1908061504364014,
+      "learning_rate": 9.67610180224943e-05,
+      "loss": 0.6813890933990479,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1173678532901834,
+      "grad_norm": 1.1787118911743164,
+      "learning_rate": 9.675231889675212e-05,
+      "loss": 0.6738101243972778,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1182308522114348,
+      "grad_norm": 1.1327152252197266,
+      "learning_rate": 9.674360849689899e-05,
+      "loss": 0.6662938594818115,
+      "step": 2592
+    },
+    {
+      "epoch": 1.119093851132686,
+      "grad_norm": 1.1246299743652344,
+      "learning_rate": 9.67348868250354e-05,
+      "loss": 0.677423357963562,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1199568500539374,
+      "grad_norm": 1.1411702632904053,
+      "learning_rate": 9.672615388326454e-05,
+      "loss": 0.6257020235061646,
+      "step": 2596
+    },
+    {
+      "epoch": 1.1208198489751888,
+      "grad_norm": 1.1607346534729004,
+      "learning_rate": 9.671740967369231e-05,
+      "loss": 0.7029470205307007,
+      "step": 2598
+    },
+    {
+      "epoch": 1.12168284789644,
+      "grad_norm": 1.047934651374817,
+      "learning_rate": 9.670865419842735e-05,
+      "loss": 0.6040990352630615,
+      "step": 2600
+    },
+    {
+      "epoch": 1.12168284789644,
+      "eval_loss": 0.7287447452545166,
+      "eval_runtime": 831.8138,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2600
+    },
+    {
+      "epoch": 1.1225458468176914,
+      "grad_norm": 1.0571783781051636,
+      "learning_rate": 9.669988745958098e-05,
+      "loss": 0.6872879266738892,
+      "step": 2602
+    },
+    {
+      "epoch": 1.1234088457389428,
+      "grad_norm": 1.0940403938293457,
+      "learning_rate": 9.669110945926727e-05,
+      "loss": 0.6173275113105774,
+      "step": 2604
+    },
+    {
+      "epoch": 1.1242718446601943,
+      "grad_norm": 1.1180222034454346,
+      "learning_rate": 9.6682320199603e-05,
+      "loss": 0.6142664551734924,
+      "step": 2606
+    },
+    {
+      "epoch": 1.1251348435814454,
+      "grad_norm": 1.251098871231079,
+      "learning_rate": 9.667351968270768e-05,
+      "loss": 0.7225713133811951,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1259978425026969,
+      "grad_norm": 1.2333059310913086,
+      "learning_rate": 9.666470791070347e-05,
+      "loss": 0.6830481290817261,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1268608414239483,
+      "grad_norm": 1.2904123067855835,
+      "learning_rate": 9.665588488571531e-05,
+      "loss": 0.6835238337516785,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1277238403451997,
+      "grad_norm": 1.4779443740844727,
+      "learning_rate": 9.664705060987085e-05,
+      "loss": 0.6993544697761536,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1285868392664509,
+      "grad_norm": 1.267021656036377,
+      "learning_rate": 9.663820508530042e-05,
+      "loss": 0.6819394826889038,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1294498381877023,
+      "grad_norm": 1.0755046606063843,
+      "learning_rate": 9.662934831413708e-05,
+      "loss": 0.6200308799743652,
+      "step": 2618
+    },
+    {
+      "epoch": 1.1303128371089537,
+      "grad_norm": 1.2043033838272095,
+      "learning_rate": 9.66204802985166e-05,
+      "loss": 0.7138658165931702,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1311758360302049,
+      "grad_norm": 1.2641128301620483,
+      "learning_rate": 9.661160104057744e-05,
+      "loss": 0.7238332033157349,
+      "step": 2622
+    },
+    {
+      "epoch": 1.1320388349514563,
+      "grad_norm": 1.0422990322113037,
+      "learning_rate": 9.660271054246085e-05,
+      "loss": 0.6401007175445557,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1329018338727077,
+      "grad_norm": 1.2298130989074707,
+      "learning_rate": 9.659380880631068e-05,
+      "loss": 0.6305879950523376,
+      "step": 2626
+    },
+    {
+      "epoch": 1.133764832793959,
+      "grad_norm": 1.254573106765747,
+      "learning_rate": 9.658489583427357e-05,
+      "loss": 0.5751872062683105,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1346278317152103,
+      "grad_norm": 1.15679132938385,
+      "learning_rate": 9.657597162849883e-05,
+      "loss": 0.6629595756530762,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1354908306364617,
+      "grad_norm": 1.193629264831543,
+      "learning_rate": 9.656703619113849e-05,
+      "loss": 0.6832258105278015,
+      "step": 2632
+    },
+    {
+      "epoch": 1.1363538295577131,
+      "grad_norm": 1.2330398559570312,
+      "learning_rate": 9.655808952434732e-05,
+      "loss": 0.6823931932449341,
+      "step": 2634
+    },
+    {
+      "epoch": 1.1372168284789643,
+      "grad_norm": 1.3926069736480713,
+      "learning_rate": 9.654913163028272e-05,
+      "loss": 0.6671395897865295,
+      "step": 2636
+    },
+    {
+      "epoch": 1.1380798274002157,
+      "grad_norm": 1.2995383739471436,
+      "learning_rate": 9.65401625111049e-05,
+      "loss": 0.662219226360321,
+      "step": 2638
+    },
+    {
+      "epoch": 1.1389428263214671,
+      "grad_norm": 1.1584415435791016,
+      "learning_rate": 9.653118216897667e-05,
+      "loss": 0.6419461965560913,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1398058252427186,
+      "grad_norm": 1.3584444522857666,
+      "learning_rate": 9.652219060606364e-05,
+      "loss": 0.6844428181648254,
+      "step": 2642
+    },
+    {
+      "epoch": 1.1406688241639698,
+      "grad_norm": 1.2619858980178833,
+      "learning_rate": 9.651318782453407e-05,
+      "loss": 0.6761999130249023,
+      "step": 2644
+    },
+    {
+      "epoch": 1.1415318230852212,
+      "grad_norm": 1.2202177047729492,
+      "learning_rate": 9.650417382655892e-05,
+      "loss": 0.7258123159408569,
+      "step": 2646
+    },
+    {
+      "epoch": 1.1423948220064726,
+      "grad_norm": 1.0225578546524048,
+      "learning_rate": 9.64951486143119e-05,
+      "loss": 0.5709917545318604,
+      "step": 2648
+    },
+    {
+      "epoch": 1.1432578209277238,
+      "grad_norm": 1.0961005687713623,
+      "learning_rate": 9.64861121899694e-05,
+      "loss": 0.7353709936141968,
+      "step": 2650
+    },
+    {
+      "epoch": 1.1441208198489752,
+      "grad_norm": 0.9393200874328613,
+      "learning_rate": 9.64770645557105e-05,
+      "loss": 0.7083833813667297,
+      "step": 2652
+    },
+    {
+      "epoch": 1.1449838187702266,
+      "grad_norm": 1.089733362197876,
+      "learning_rate": 9.646800571371701e-05,
+      "loss": 0.6405303478240967,
+      "step": 2654
+    },
+    {
+      "epoch": 1.1458468176914778,
+      "grad_norm": 1.1581398248672485,
+      "learning_rate": 9.645893566617341e-05,
+      "loss": 0.617733359336853,
+      "step": 2656
+    },
+    {
+      "epoch": 1.1467098166127292,
+      "grad_norm": 1.1336220502853394,
+      "learning_rate": 9.644985441526693e-05,
+      "loss": 0.6948232650756836,
+      "step": 2658
+    },
+    {
+      "epoch": 1.1475728155339806,
+      "grad_norm": 1.028591513633728,
+      "learning_rate": 9.644076196318743e-05,
+      "loss": 0.6392919421195984,
+      "step": 2660
+    },
+    {
+      "epoch": 1.148435814455232,
+      "grad_norm": 1.2088173627853394,
+      "learning_rate": 9.643165831212756e-05,
+      "loss": 0.6384931206703186,
+      "step": 2662
+    },
+    {
+      "epoch": 1.1492988133764832,
+      "grad_norm": 0.9905838966369629,
+      "learning_rate": 9.642254346428259e-05,
+      "loss": 0.5909387469291687,
+      "step": 2664
+    },
+    {
+      "epoch": 1.1501618122977346,
+      "grad_norm": 1.1670424938201904,
+      "learning_rate": 9.641341742185054e-05,
+      "loss": 0.700904130935669,
+      "step": 2666
+    },
+    {
+      "epoch": 1.151024811218986,
+      "grad_norm": 1.164746880531311,
+      "learning_rate": 9.640428018703211e-05,
+      "loss": 0.6036953330039978,
+      "step": 2668
+    },
+    {
+      "epoch": 1.1518878101402374,
+      "grad_norm": 1.00895357131958,
+      "learning_rate": 9.639513176203072e-05,
+      "loss": 0.6158549785614014,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1527508090614886,
+      "grad_norm": 1.2236207723617554,
+      "learning_rate": 9.638597214905244e-05,
+      "loss": 0.7280431389808655,
+      "step": 2672
+    },
+    {
+      "epoch": 1.15361380798274,
+      "grad_norm": 1.0930732488632202,
+      "learning_rate": 9.637680135030609e-05,
+      "loss": 0.6434410810470581,
+      "step": 2674
+    },
+    {
+      "epoch": 1.1544768069039915,
+      "grad_norm": 1.366325855255127,
+      "learning_rate": 9.636761936800315e-05,
+      "loss": 0.6771191954612732,
+      "step": 2676
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 2.249835968017578,
+      "learning_rate": 9.635842620435785e-05,
+      "loss": 0.6382432579994202,
+      "step": 2678
+    },
+    {
+      "epoch": 1.156202804746494,
+      "grad_norm": 1.274483561515808,
+      "learning_rate": 9.634922186158704e-05,
+      "loss": 0.6733657121658325,
+      "step": 2680
+    },
+    {
+      "epoch": 1.1570658036677455,
+      "grad_norm": 1.3715782165527344,
+      "learning_rate": 9.634000634191033e-05,
+      "loss": 0.7601200938224792,
+      "step": 2682
+    },
+    {
+      "epoch": 1.1579288025889967,
+      "grad_norm": 1.1312460899353027,
+      "learning_rate": 9.633077964754998e-05,
+      "loss": 0.5959098935127258,
+      "step": 2684
+    },
+    {
+      "epoch": 1.158791801510248,
+      "grad_norm": 1.063233494758606,
+      "learning_rate": 9.632154178073098e-05,
+      "loss": 0.615096926689148,
+      "step": 2686
+    },
+    {
+      "epoch": 1.1596548004314995,
+      "grad_norm": 1.074887752532959,
+      "learning_rate": 9.6312292743681e-05,
+      "loss": 0.6553784608840942,
+      "step": 2688
+    },
+    {
+      "epoch": 1.160517799352751,
+      "grad_norm": 1.17954683303833,
+      "learning_rate": 9.630303253863042e-05,
+      "loss": 0.6963269114494324,
+      "step": 2690
+    },
+    {
+      "epoch": 1.161380798274002,
+      "grad_norm": 1.2549158334732056,
+      "learning_rate": 9.629376116781226e-05,
+      "loss": 0.6633999347686768,
+      "step": 2692
+    },
+    {
+      "epoch": 1.1622437971952535,
+      "grad_norm": 1.237196683883667,
+      "learning_rate": 9.628447863346229e-05,
+      "loss": 0.6727597117424011,
+      "step": 2694
+    },
+    {
+      "epoch": 1.163106796116505,
+      "grad_norm": 1.4047741889953613,
+      "learning_rate": 9.627518493781895e-05,
+      "loss": 0.7018169164657593,
+      "step": 2696
+    },
+    {
+      "epoch": 1.1639697950377563,
+      "grad_norm": 1.15412175655365,
+      "learning_rate": 9.626588008312339e-05,
+      "loss": 0.6566510796546936,
+      "step": 2698
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "grad_norm": 1.194110631942749,
+      "learning_rate": 9.625656407161938e-05,
+      "loss": 0.7017931342124939,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1648327939590075,
+      "eval_loss": 0.72404944896698,
+      "eval_runtime": 824.2877,
+      "eval_samples_per_second": 2.499,
+      "eval_steps_per_second": 2.499,
+      "step": 2700
+    },
+    {
+      "epoch": 1.165695792880259,
+      "grad_norm": 1.1512954235076904,
+      "learning_rate": 9.62472369055535e-05,
+      "loss": 0.7058532238006592,
+      "step": 2702
+    },
+    {
+      "epoch": 1.1665587918015103,
+      "grad_norm": 1.4743460416793823,
+      "learning_rate": 9.623789858717492e-05,
+      "loss": 0.7458376288414001,
+      "step": 2704
+    },
+    {
+      "epoch": 1.1674217907227615,
+      "grad_norm": 1.0526450872421265,
+      "learning_rate": 9.622854911873551e-05,
+      "loss": 0.6526424288749695,
+      "step": 2706
+    },
+    {
+      "epoch": 1.168284789644013,
+      "grad_norm": 1.0408949851989746,
+      "learning_rate": 9.62191885024899e-05,
+      "loss": 0.6055451035499573,
+      "step": 2708
+    },
+    {
+      "epoch": 1.1691477885652644,
+      "grad_norm": 1.214052677154541,
+      "learning_rate": 9.620981674069534e-05,
+      "loss": 0.666462779045105,
+      "step": 2710
+    },
+    {
+      "epoch": 1.1700107874865155,
+      "grad_norm": 1.2526963949203491,
+      "learning_rate": 9.62004338356118e-05,
+      "loss": 0.7208677530288696,
+      "step": 2712
+    },
+    {
+      "epoch": 1.170873786407767,
+      "grad_norm": 1.2097395658493042,
+      "learning_rate": 9.61910397895019e-05,
+      "loss": 0.6269269585609436,
+      "step": 2714
+    },
+    {
+      "epoch": 1.1717367853290184,
+      "grad_norm": 1.035733699798584,
+      "learning_rate": 9.618163460463096e-05,
+      "loss": 0.6063299179077148,
+      "step": 2716
+    },
+    {
+      "epoch": 1.1725997842502698,
+      "grad_norm": 1.2881336212158203,
+      "learning_rate": 9.617221828326706e-05,
+      "loss": 0.7084641456604004,
+      "step": 2718
+    },
+    {
+      "epoch": 1.173462783171521,
+      "grad_norm": 1.2808257341384888,
+      "learning_rate": 9.616279082768085e-05,
+      "loss": 0.6515718102455139,
+      "step": 2720
+    },
+    {
+      "epoch": 1.1743257820927724,
+      "grad_norm": 1.1174249649047852,
+      "learning_rate": 9.615335224014574e-05,
+      "loss": 0.6745902299880981,
+      "step": 2722
+    },
+    {
+      "epoch": 1.1751887810140238,
+      "grad_norm": 1.1222290992736816,
+      "learning_rate": 9.61439025229378e-05,
+      "loss": 0.6824861168861389,
+      "step": 2724
+    },
+    {
+      "epoch": 1.176051779935275,
+      "grad_norm": 1.156187891960144,
+      "learning_rate": 9.613444167833579e-05,
+      "loss": 0.6020817160606384,
+      "step": 2726
+    },
+    {
+      "epoch": 1.1769147788565264,
+      "grad_norm": 1.2502888441085815,
+      "learning_rate": 9.612496970862113e-05,
+      "loss": 0.6430572271347046,
+      "step": 2728
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 1.6157052516937256,
+      "learning_rate": 9.611548661607797e-05,
+      "loss": 0.7195103168487549,
+      "step": 2730
+    },
+    {
+      "epoch": 1.1786407766990292,
+      "grad_norm": 1.3196234703063965,
+      "learning_rate": 9.610599240299308e-05,
+      "loss": 0.710418164730072,
+      "step": 2732
+    },
+    {
+      "epoch": 1.1795037756202804,
+      "grad_norm": 1.2117719650268555,
+      "learning_rate": 9.6096487071656e-05,
+      "loss": 0.650095522403717,
+      "step": 2734
+    },
+    {
+      "epoch": 1.1803667745415318,
+      "grad_norm": 2.2723631858825684,
+      "learning_rate": 9.608697062435884e-05,
+      "loss": 0.7150285243988037,
+      "step": 2736
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 1.0523152351379395,
+      "learning_rate": 9.607744306339647e-05,
+      "loss": 0.6271997690200806,
+      "step": 2738
+    },
+    {
+      "epoch": 1.1820927723840344,
+      "grad_norm": 1.2567129135131836,
+      "learning_rate": 9.606790439106646e-05,
+      "loss": 0.659799337387085,
+      "step": 2740
+    },
+    {
+      "epoch": 1.1829557713052858,
+      "grad_norm": 1.1448849439620972,
+      "learning_rate": 9.605835460966896e-05,
+      "loss": 0.7097326517105103,
+      "step": 2742
+    },
+    {
+      "epoch": 1.1838187702265373,
+      "grad_norm": 1.3698351383209229,
+      "learning_rate": 9.604879372150688e-05,
+      "loss": 0.6692788004875183,
+      "step": 2744
+    },
+    {
+      "epoch": 1.1846817691477884,
+      "grad_norm": 1.1424204111099243,
+      "learning_rate": 9.603922172888577e-05,
+      "loss": 0.6453474760055542,
+      "step": 2746
+    },
+    {
+      "epoch": 1.1855447680690399,
+      "grad_norm": 1.2701301574707031,
+      "learning_rate": 9.60296386341139e-05,
+      "loss": 0.6197408437728882,
+      "step": 2748
+    },
+    {
+      "epoch": 1.1864077669902913,
+      "grad_norm": 1.1171884536743164,
+      "learning_rate": 9.602004443950217e-05,
+      "loss": 0.6525035500526428,
+      "step": 2750
+    },
+    {
+      "epoch": 1.1872707659115427,
+      "grad_norm": 1.132633090019226,
+      "learning_rate": 9.60104391473642e-05,
+      "loss": 0.6459797620773315,
+      "step": 2752
+    },
+    {
+      "epoch": 1.1881337648327939,
+      "grad_norm": 1.1362011432647705,
+      "learning_rate": 9.600082276001623e-05,
+      "loss": 0.6175895929336548,
+      "step": 2754
+    },
+    {
+      "epoch": 1.1889967637540453,
+      "grad_norm": 1.1216920614242554,
+      "learning_rate": 9.599119527977722e-05,
+      "loss": 0.6782001852989197,
+      "step": 2756
+    },
+    {
+      "epoch": 1.1898597626752967,
+      "grad_norm": 1.024833083152771,
+      "learning_rate": 9.59815567089688e-05,
+      "loss": 0.6387497782707214,
+      "step": 2758
+    },
+    {
+      "epoch": 1.190722761596548,
+      "grad_norm": 1.3539615869522095,
+      "learning_rate": 9.597190704991527e-05,
+      "loss": 0.6974280476570129,
+      "step": 2760
+    },
+    {
+      "epoch": 1.1915857605177993,
+      "grad_norm": 1.002167820930481,
+      "learning_rate": 9.59622463049436e-05,
+      "loss": 0.6039770245552063,
+      "step": 2762
+    },
+    {
+      "epoch": 1.1924487594390507,
+      "grad_norm": 1.150636076927185,
+      "learning_rate": 9.595257447638343e-05,
+      "loss": 0.6543651819229126,
+      "step": 2764
+    },
+    {
+      "epoch": 1.1933117583603021,
+      "grad_norm": 1.413001537322998,
+      "learning_rate": 9.594289156656709e-05,
+      "loss": 0.7388567328453064,
+      "step": 2766
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 1.2041958570480347,
+      "learning_rate": 9.593319757782955e-05,
+      "loss": 0.7128161191940308,
+      "step": 2768
+    },
+    {
+      "epoch": 1.1950377562028047,
+      "grad_norm": 1.2151374816894531,
+      "learning_rate": 9.592349251250847e-05,
+      "loss": 0.6593868136405945,
+      "step": 2770
+    },
+    {
+      "epoch": 1.1959007551240561,
+      "grad_norm": 1.6225353479385376,
+      "learning_rate": 9.59137763729442e-05,
+      "loss": 0.6128022074699402,
+      "step": 2772
+    },
+    {
+      "epoch": 1.1967637540453073,
+      "grad_norm": 1.1904171705245972,
+      "learning_rate": 9.590404916147976e-05,
+      "loss": 0.6537567377090454,
+      "step": 2774
+    },
+    {
+      "epoch": 1.1976267529665587,
+      "grad_norm": 1.159506916999817,
+      "learning_rate": 9.589431088046079e-05,
+      "loss": 0.6325924396514893,
+      "step": 2776
+    },
+    {
+      "epoch": 1.1984897518878102,
+      "grad_norm": 1.1314560174942017,
+      "learning_rate": 9.588456153223565e-05,
+      "loss": 0.6438963413238525,
+      "step": 2778
+    },
+    {
+      "epoch": 1.1993527508090616,
+      "grad_norm": 1.0182478427886963,
+      "learning_rate": 9.587480111915532e-05,
+      "loss": 0.6703475117683411,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2002157497303128,
+      "grad_norm": 1.5506517887115479,
+      "learning_rate": 9.586502964357352e-05,
+      "loss": 0.7694190144538879,
+      "step": 2782
+    },
+    {
+      "epoch": 1.2010787486515642,
+      "grad_norm": 1.0769826173782349,
+      "learning_rate": 9.585524710784659e-05,
+      "loss": 0.6117724180221558,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2019417475728156,
+      "grad_norm": 1.207175850868225,
+      "learning_rate": 9.584545351433352e-05,
+      "loss": 0.6375927329063416,
+      "step": 2786
+    },
+    {
+      "epoch": 1.202804746494067,
+      "grad_norm": 1.1291100978851318,
+      "learning_rate": 9.583564886539602e-05,
+      "loss": 0.6526537537574768,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2036677454153182,
+      "grad_norm": 1.3076926469802856,
+      "learning_rate": 9.582583316339843e-05,
+      "loss": 0.7080113887786865,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2045307443365696,
+      "grad_norm": 1.166823148727417,
+      "learning_rate": 9.581600641070773e-05,
+      "loss": 0.6415606141090393,
+      "step": 2792
+    },
+    {
+      "epoch": 1.205393743257821,
+      "grad_norm": 1.1248292922973633,
+      "learning_rate": 9.580616860969364e-05,
+      "loss": 0.6538968086242676,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2062567421790722,
+      "grad_norm": 1.406731128692627,
+      "learning_rate": 9.579631976272849e-05,
+      "loss": 0.694507896900177,
+      "step": 2796
+    },
+    {
+      "epoch": 1.2071197411003236,
+      "grad_norm": 1.185895562171936,
+      "learning_rate": 9.578645987218728e-05,
+      "loss": 0.7064871788024902,
+      "step": 2798
+    },
+    {
+      "epoch": 1.207982740021575,
+      "grad_norm": 1.0860708951950073,
+      "learning_rate": 9.577658894044768e-05,
+      "loss": 0.5790154933929443,
+      "step": 2800
+    },
+    {
+      "epoch": 1.207982740021575,
+      "eval_loss": 0.7220398187637329,
+      "eval_runtime": 831.5812,
+      "eval_samples_per_second": 2.477,
+      "eval_steps_per_second": 2.477,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2088457389428262,
+      "grad_norm": 1.2761744260787964,
+      "learning_rate": 9.576670696989001e-05,
+      "loss": 0.6645335555076599,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2097087378640776,
+      "grad_norm": 1.3502917289733887,
+      "learning_rate": 9.575681396289728e-05,
+      "loss": 0.6320346593856812,
+      "step": 2804
+    },
+    {
+      "epoch": 1.210571736785329,
+      "grad_norm": 1.2638593912124634,
+      "learning_rate": 9.574690992185514e-05,
+      "loss": 0.6297362446784973,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2114347357065804,
+      "grad_norm": 1.0787774324417114,
+      "learning_rate": 9.573699484915189e-05,
+      "loss": 0.7569923996925354,
+      "step": 2808
+    },
+    {
+      "epoch": 1.2122977346278316,
+      "grad_norm": 1.097254991531372,
+      "learning_rate": 9.572706874717851e-05,
+      "loss": 0.6712964773178101,
+      "step": 2810
+    },
+    {
+      "epoch": 1.213160733549083,
+      "grad_norm": 1.0995726585388184,
+      "learning_rate": 9.571713161832865e-05,
+      "loss": 0.6775450706481934,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2140237324703345,
+      "grad_norm": 1.7386797666549683,
+      "learning_rate": 9.570718346499859e-05,
+      "loss": 0.6691418290138245,
+      "step": 2814
+    },
+    {
+      "epoch": 1.2148867313915859,
+      "grad_norm": 1.2281049489974976,
+      "learning_rate": 9.569722428958728e-05,
+      "loss": 0.6406403183937073,
+      "step": 2816
+    },
+    {
+      "epoch": 1.215749730312837,
+      "grad_norm": 1.1326305866241455,
+      "learning_rate": 9.568725409449635e-05,
+      "loss": 0.7008584141731262,
+      "step": 2818
+    },
+    {
+      "epoch": 1.2166127292340885,
+      "grad_norm": 1.3039381504058838,
+      "learning_rate": 9.567727288213005e-05,
+      "loss": 0.7438415288925171,
+      "step": 2820
+    },
+    {
+      "epoch": 1.2174757281553399,
+      "grad_norm": 1.219136357307434,
+      "learning_rate": 9.566728065489531e-05,
+      "loss": 0.6937777996063232,
+      "step": 2822
+    },
+    {
+      "epoch": 1.218338727076591,
+      "grad_norm": 1.321285605430603,
+      "learning_rate": 9.56572774152017e-05,
+      "loss": 0.7586468458175659,
+      "step": 2824
+    },
+    {
+      "epoch": 1.2192017259978425,
+      "grad_norm": 1.3538484573364258,
+      "learning_rate": 9.564726316546147e-05,
+      "loss": 0.6356340050697327,
+      "step": 2826
+    },
+    {
+      "epoch": 1.220064724919094,
+      "grad_norm": 1.1015816926956177,
+      "learning_rate": 9.563723790808952e-05,
+      "loss": 0.6304616928100586,
+      "step": 2828
+    },
+    {
+      "epoch": 1.220927723840345,
+      "grad_norm": 0.9658390283584595,
+      "learning_rate": 9.562720164550337e-05,
+      "loss": 0.6459048986434937,
+      "step": 2830
+    },
+    {
+      "epoch": 1.2217907227615965,
+      "grad_norm": 1.3981443643569946,
+      "learning_rate": 9.561715438012323e-05,
+      "loss": 0.7034182548522949,
+      "step": 2832
+    },
+    {
+      "epoch": 1.222653721682848,
+      "grad_norm": 1.1737093925476074,
+      "learning_rate": 9.560709611437195e-05,
+      "loss": 0.6629860997200012,
+      "step": 2834
+    },
+    {
+      "epoch": 1.2235167206040993,
+      "grad_norm": 1.2071303129196167,
+      "learning_rate": 9.559702685067506e-05,
+      "loss": 0.7166537642478943,
+      "step": 2836
+    },
+    {
+      "epoch": 1.2243797195253505,
+      "grad_norm": 1.2520930767059326,
+      "learning_rate": 9.558694659146068e-05,
+      "loss": 0.6711867451667786,
+      "step": 2838
+    },
+    {
+      "epoch": 1.225242718446602,
+      "grad_norm": 1.2304543256759644,
+      "learning_rate": 9.557685533915964e-05,
+      "loss": 0.7135099768638611,
+      "step": 2840
+    },
+    {
+      "epoch": 1.2261057173678533,
+      "grad_norm": 1.1509613990783691,
+      "learning_rate": 9.556675309620539e-05,
+      "loss": 0.6279205083847046,
+      "step": 2842
+    },
+    {
+      "epoch": 1.2269687162891048,
+      "grad_norm": 1.2024176120758057,
+      "learning_rate": 9.555663986503406e-05,
+      "loss": 0.6848570704460144,
+      "step": 2844
+    },
+    {
+      "epoch": 1.227831715210356,
+      "grad_norm": 1.360257863998413,
+      "learning_rate": 9.55465156480844e-05,
+      "loss": 0.7109206914901733,
+      "step": 2846
+    },
+    {
+      "epoch": 1.2286947141316074,
+      "grad_norm": 1.1556848287582397,
+      "learning_rate": 9.55363804477978e-05,
+      "loss": 0.5983237028121948,
+      "step": 2848
+    },
+    {
+      "epoch": 1.2295577130528588,
+      "grad_norm": 1.131439447402954,
+      "learning_rate": 9.552623426661836e-05,
+      "loss": 0.6387401223182678,
+      "step": 2850
+    },
+    {
+      "epoch": 1.23042071197411,
+      "grad_norm": 1.1854766607284546,
+      "learning_rate": 9.551607710699273e-05,
+      "loss": 0.6584593057632446,
+      "step": 2852
+    },
+    {
+      "epoch": 1.2312837108953614,
+      "grad_norm": 1.1811782121658325,
+      "learning_rate": 9.55059089713703e-05,
+      "loss": 0.6284405589103699,
+      "step": 2854
+    },
+    {
+      "epoch": 1.2321467098166128,
+      "grad_norm": 1.1038708686828613,
+      "learning_rate": 9.549572986220307e-05,
+      "loss": 0.6183417439460754,
+      "step": 2856
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 1.3645228147506714,
+      "learning_rate": 9.548553978194567e-05,
+      "loss": 0.7266764640808105,
+      "step": 2858
+    },
+    {
+      "epoch": 1.2338727076591154,
+      "grad_norm": 1.1896798610687256,
+      "learning_rate": 9.547533873305542e-05,
+      "loss": 0.6580738425254822,
+      "step": 2860
+    },
+    {
+      "epoch": 1.2347357065803668,
+      "grad_norm": 1.1351168155670166,
+      "learning_rate": 9.546512671799224e-05,
+      "loss": 0.669810950756073,
+      "step": 2862
+    },
+    {
+      "epoch": 1.2355987055016182,
+      "grad_norm": 1.1937936544418335,
+      "learning_rate": 9.545490373921869e-05,
+      "loss": 0.730329155921936,
+      "step": 2864
+    },
+    {
+      "epoch": 1.2364617044228694,
+      "grad_norm": 1.094347357749939,
+      "learning_rate": 9.544466979920002e-05,
+      "loss": 0.6867446303367615,
+      "step": 2866
+    },
+    {
+      "epoch": 1.2373247033441208,
+      "grad_norm": 1.4509600400924683,
+      "learning_rate": 9.54344249004041e-05,
+      "loss": 0.735468327999115,
+      "step": 2868
+    },
+    {
+      "epoch": 1.2381877022653722,
+      "grad_norm": 1.1195857524871826,
+      "learning_rate": 9.542416904530142e-05,
+      "loss": 0.6316167116165161,
+      "step": 2870
+    },
+    {
+      "epoch": 1.2390507011866236,
+      "grad_norm": 1.0227746963500977,
+      "learning_rate": 9.541390223636517e-05,
+      "loss": 0.6637873649597168,
+      "step": 2872
+    },
+    {
+      "epoch": 1.2399137001078748,
+      "grad_norm": 1.1871414184570312,
+      "learning_rate": 9.540362447607111e-05,
+      "loss": 0.6579877734184265,
+      "step": 2874
+    },
+    {
+      "epoch": 1.2407766990291262,
+      "grad_norm": 1.3265694379806519,
+      "learning_rate": 9.539333576689768e-05,
+      "loss": 0.7563158273696899,
+      "step": 2876
+    },
+    {
+      "epoch": 1.2416396979503777,
+      "grad_norm": 1.0074095726013184,
+      "learning_rate": 9.538303611132598e-05,
+      "loss": 0.6373291015625,
+      "step": 2878
+    },
+    {
+      "epoch": 1.2425026968716288,
+      "grad_norm": 1.421204924583435,
+      "learning_rate": 9.537272551183971e-05,
+      "loss": 0.6989303827285767,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2433656957928803,
+      "grad_norm": 1.256486415863037,
+      "learning_rate": 9.536240397092522e-05,
+      "loss": 0.7590261697769165,
+      "step": 2882
+    },
+    {
+      "epoch": 1.2442286947141317,
+      "grad_norm": 1.1513274908065796,
+      "learning_rate": 9.53520714910715e-05,
+      "loss": 0.6652005314826965,
+      "step": 2884
+    },
+    {
+      "epoch": 1.2450916936353829,
+      "grad_norm": 1.2129496335983276,
+      "learning_rate": 9.53417280747702e-05,
+      "loss": 0.6510938405990601,
+      "step": 2886
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 1.1393296718597412,
+      "learning_rate": 9.533137372451557e-05,
+      "loss": 0.6405928730964661,
+      "step": 2888
+    },
+    {
+      "epoch": 1.2468176914778857,
+      "grad_norm": 1.1467162370681763,
+      "learning_rate": 9.532100844280451e-05,
+      "loss": 0.6819307208061218,
+      "step": 2890
+    },
+    {
+      "epoch": 1.247680690399137,
+      "grad_norm": 1.1516613960266113,
+      "learning_rate": 9.531063223213659e-05,
+      "loss": 0.6333560347557068,
+      "step": 2892
+    },
+    {
+      "epoch": 1.2485436893203883,
+      "grad_norm": 1.463189959526062,
+      "learning_rate": 9.530024509501396e-05,
+      "loss": 0.7259839177131653,
+      "step": 2894
+    },
+    {
+      "epoch": 1.2494066882416397,
+      "grad_norm": 1.0450547933578491,
+      "learning_rate": 9.528984703394144e-05,
+      "loss": 0.661693811416626,
+      "step": 2896
+    },
+    {
+      "epoch": 1.250269687162891,
+      "grad_norm": 1.0525774955749512,
+      "learning_rate": 9.527943805142648e-05,
+      "loss": 0.5889303684234619,
+      "step": 2898
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "grad_norm": 1.130048394203186,
+      "learning_rate": 9.526901814997914e-05,
+      "loss": 0.6380610466003418,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2511326860841425,
+      "eval_loss": 0.7182289958000183,
+      "eval_runtime": 840.8156,
+      "eval_samples_per_second": 2.45,
+      "eval_steps_per_second": 2.45,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2519956850053937,
+      "grad_norm": 1.0194777250289917,
+      "learning_rate": 9.525858733211217e-05,
+      "loss": 0.6397356986999512,
+      "step": 2902
+    },
+    {
+      "epoch": 1.2528586839266451,
+      "grad_norm": 1.3902885913848877,
+      "learning_rate": 9.524814560034087e-05,
+      "loss": 0.7931577563285828,
+      "step": 2904
+    },
+    {
+      "epoch": 1.2537216828478965,
+      "grad_norm": 1.258169412612915,
+      "learning_rate": 9.523769295718325e-05,
+      "loss": 0.6997087597846985,
+      "step": 2906
+    },
+    {
+      "epoch": 1.2545846817691477,
+      "grad_norm": 1.3403711318969727,
+      "learning_rate": 9.522722940515991e-05,
+      "loss": 0.6946825385093689,
+      "step": 2908
+    },
+    {
+      "epoch": 1.2554476806903991,
+      "grad_norm": 1.1724950075149536,
+      "learning_rate": 9.521675494679409e-05,
+      "loss": 0.6309347748756409,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2563106796116505,
+      "grad_norm": 1.210274338722229,
+      "learning_rate": 9.520626958461163e-05,
+      "loss": 0.6453542113304138,
+      "step": 2912
+    },
+    {
+      "epoch": 1.2571736785329017,
+      "grad_norm": 1.233112096786499,
+      "learning_rate": 9.519577332114106e-05,
+      "loss": 0.6747867465019226,
+      "step": 2914
+    },
+    {
+      "epoch": 1.2580366774541532,
+      "grad_norm": 1.1785951852798462,
+      "learning_rate": 9.51852661589135e-05,
+      "loss": 0.6729235053062439,
+      "step": 2916
+    },
+    {
+      "epoch": 1.2588996763754046,
+      "grad_norm": 1.0781456232070923,
+      "learning_rate": 9.51747481004627e-05,
+      "loss": 0.6581047177314758,
+      "step": 2918
+    },
+    {
+      "epoch": 1.2597626752966558,
+      "grad_norm": 1.1122184991836548,
+      "learning_rate": 9.516421914832506e-05,
+      "loss": 0.6576398015022278,
+      "step": 2920
+    },
+    {
+      "epoch": 1.2606256742179072,
+      "grad_norm": 1.3378703594207764,
+      "learning_rate": 9.515367930503956e-05,
+      "loss": 0.7026275396347046,
+      "step": 2922
+    },
+    {
+      "epoch": 1.2614886731391586,
+      "grad_norm": 1.127657175064087,
+      "learning_rate": 9.514312857314785e-05,
+      "loss": 0.6111703515052795,
+      "step": 2924
+    },
+    {
+      "epoch": 1.26235167206041,
+      "grad_norm": 1.2058607339859009,
+      "learning_rate": 9.513256695519419e-05,
+      "loss": 0.6698760986328125,
+      "step": 2926
+    },
+    {
+      "epoch": 1.2632146709816614,
+      "grad_norm": 1.2206004858016968,
+      "learning_rate": 9.512199445372546e-05,
+      "loss": 0.6793603897094727,
+      "step": 2928
+    },
+    {
+      "epoch": 1.2640776699029126,
+      "grad_norm": 1.167311668395996,
+      "learning_rate": 9.511141107129119e-05,
+      "loss": 0.709406316280365,
+      "step": 2930
+    },
+    {
+      "epoch": 1.264940668824164,
+      "grad_norm": 1.3347517251968384,
+      "learning_rate": 9.51008168104435e-05,
+      "loss": 0.685570240020752,
+      "step": 2932
+    },
+    {
+      "epoch": 1.2658036677454154,
+      "grad_norm": 1.186468482017517,
+      "learning_rate": 9.509021167373714e-05,
+      "loss": 0.6775280833244324,
+      "step": 2934
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 1.3934271335601807,
+      "learning_rate": 9.507959566372953e-05,
+      "loss": 0.6587507128715515,
+      "step": 2936
+    },
+    {
+      "epoch": 1.267529665587918,
+      "grad_norm": 1.3766281604766846,
+      "learning_rate": 9.506896878298061e-05,
+      "loss": 0.692707896232605,
+      "step": 2938
+    },
+    {
+      "epoch": 1.2683926645091694,
+      "grad_norm": 1.034883975982666,
+      "learning_rate": 9.505833103405307e-05,
+      "loss": 0.6481655836105347,
+      "step": 2940
+    },
+    {
+      "epoch": 1.2692556634304206,
+      "grad_norm": 1.113044261932373,
+      "learning_rate": 9.504768241951212e-05,
+      "loss": 0.5867606997489929,
+      "step": 2942
+    },
+    {
+      "epoch": 1.270118662351672,
+      "grad_norm": 1.0769455432891846,
+      "learning_rate": 9.503702294192563e-05,
+      "loss": 0.6147398948669434,
+      "step": 2944
+    },
+    {
+      "epoch": 1.2709816612729234,
+      "grad_norm": 1.100616216659546,
+      "learning_rate": 9.502635260386408e-05,
+      "loss": 0.650299608707428,
+      "step": 2946
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 1.1982520818710327,
+      "learning_rate": 9.501567140790057e-05,
+      "loss": 0.6896426677703857,
+      "step": 2948
+    },
+    {
+      "epoch": 1.272707659115426,
+      "grad_norm": 1.276911735534668,
+      "learning_rate": 9.500497935661085e-05,
+      "loss": 0.6780133247375488,
+      "step": 2950
+    },
+    {
+      "epoch": 1.2735706580366775,
+      "grad_norm": 1.131345272064209,
+      "learning_rate": 9.499427645257323e-05,
+      "loss": 0.6906163692474365,
+      "step": 2952
+    },
+    {
+      "epoch": 1.2744336569579289,
+      "grad_norm": 1.125251293182373,
+      "learning_rate": 9.498356269836868e-05,
+      "loss": 0.6418634057044983,
+      "step": 2954
+    },
+    {
+      "epoch": 1.2752966558791803,
+      "grad_norm": 1.2508443593978882,
+      "learning_rate": 9.497283809658078e-05,
+      "loss": 0.690779983997345,
+      "step": 2956
+    },
+    {
+      "epoch": 1.2761596548004315,
+      "grad_norm": 0.9558717608451843,
+      "learning_rate": 9.49621026497957e-05,
+      "loss": 0.645685613155365,
+      "step": 2958
+    },
+    {
+      "epoch": 1.2770226537216829,
+      "grad_norm": 1.1781591176986694,
+      "learning_rate": 9.495135636060226e-05,
+      "loss": 0.6313972473144531,
+      "step": 2960
+    },
+    {
+      "epoch": 1.2778856526429343,
+      "grad_norm": 0.9713736772537231,
+      "learning_rate": 9.494059923159189e-05,
+      "loss": 0.5853794813156128,
+      "step": 2962
+    },
+    {
+      "epoch": 1.2787486515641855,
+      "grad_norm": 1.3793153762817383,
+      "learning_rate": 9.492983126535859e-05,
+      "loss": 0.6966362595558167,
+      "step": 2964
+    },
+    {
+      "epoch": 1.279611650485437,
+      "grad_norm": 1.250563383102417,
+      "learning_rate": 9.491905246449903e-05,
+      "loss": 0.7281587719917297,
+      "step": 2966
+    },
+    {
+      "epoch": 1.2804746494066883,
+      "grad_norm": 1.3268930912017822,
+      "learning_rate": 9.490826283161247e-05,
+      "loss": 0.7322974801063538,
+      "step": 2968
+    },
+    {
+      "epoch": 1.2813376483279395,
+      "grad_norm": 1.3876935243606567,
+      "learning_rate": 9.489746236930077e-05,
+      "loss": 0.7875081896781921,
+      "step": 2970
+    },
+    {
+      "epoch": 1.282200647249191,
+      "grad_norm": 1.02049720287323,
+      "learning_rate": 9.488665108016842e-05,
+      "loss": 0.601652979850769,
+      "step": 2972
+    },
+    {
+      "epoch": 1.2830636461704423,
+      "grad_norm": 1.0868756771087646,
+      "learning_rate": 9.487582896682252e-05,
+      "loss": 0.6255780458450317,
+      "step": 2974
+    },
+    {
+      "epoch": 1.2839266450916935,
+      "grad_norm": 1.00597083568573,
+      "learning_rate": 9.486499603187275e-05,
+      "loss": 0.5967418551445007,
+      "step": 2976
+    },
+    {
+      "epoch": 1.284789644012945,
+      "grad_norm": 1.3068969249725342,
+      "learning_rate": 9.485415227793145e-05,
+      "loss": 0.719573438167572,
+      "step": 2978
+    },
+    {
+      "epoch": 1.2856526429341963,
+      "grad_norm": 1.2478386163711548,
+      "learning_rate": 9.484329770761353e-05,
+      "loss": 0.665522038936615,
+      "step": 2980
+    },
+    {
+      "epoch": 1.2865156418554478,
+      "grad_norm": 1.1220918893814087,
+      "learning_rate": 9.483243232353652e-05,
+      "loss": 0.6798994541168213,
+      "step": 2982
+    },
+    {
+      "epoch": 1.287378640776699,
+      "grad_norm": 1.1369609832763672,
+      "learning_rate": 9.482155612832057e-05,
+      "loss": 0.672002911567688,
+      "step": 2984
+    },
+    {
+      "epoch": 1.2882416396979504,
+      "grad_norm": 1.2077715396881104,
+      "learning_rate": 9.48106691245884e-05,
+      "loss": 0.6755662560462952,
+      "step": 2986
+    },
+    {
+      "epoch": 1.2891046386192018,
+      "grad_norm": 1.3299074172973633,
+      "learning_rate": 9.479977131496539e-05,
+      "loss": 0.7504231333732605,
+      "step": 2988
+    },
+    {
+      "epoch": 1.2899676375404532,
+      "grad_norm": 1.08982253074646,
+      "learning_rate": 9.478886270207948e-05,
+      "loss": 0.6837405562400818,
+      "step": 2990
+    },
+    {
+      "epoch": 1.2908306364617044,
+      "grad_norm": 1.1015464067459106,
+      "learning_rate": 9.477794328856124e-05,
+      "loss": 0.6552648544311523,
+      "step": 2992
+    },
+    {
+      "epoch": 1.2916936353829558,
+      "grad_norm": 1.1793229579925537,
+      "learning_rate": 9.476701307704384e-05,
+      "loss": 0.6160483360290527,
+      "step": 2994
+    },
+    {
+      "epoch": 1.2925566343042072,
+      "grad_norm": 1.4781638383865356,
+      "learning_rate": 9.475607207016306e-05,
+      "loss": 0.7406170964241028,
+      "step": 2996
+    },
+    {
+      "epoch": 1.2934196332254584,
+      "grad_norm": 1.2391449213027954,
+      "learning_rate": 9.474512027055724e-05,
+      "loss": 0.5950883626937866,
+      "step": 2998
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "grad_norm": 1.166535496711731,
+      "learning_rate": 9.473415768086738e-05,
+      "loss": 0.6587089896202087,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2942826321467098,
+      "eval_loss": 0.7160727977752686,
+      "eval_runtime": 830.8186,
+      "eval_samples_per_second": 2.479,
+      "eval_steps_per_second": 2.479,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2951456310679612,
+      "grad_norm": 0.9833650588989258,
+      "learning_rate": 9.472318430373706e-05,
+      "loss": 0.5802602767944336,
+      "step": 3002
+    },
+    {
+      "epoch": 1.2960086299892124,
+      "grad_norm": 1.047341227531433,
+      "learning_rate": 9.471220014181246e-05,
+      "loss": 0.6255064606666565,
+      "step": 3004
+    },
+    {
+      "epoch": 1.2968716289104638,
+      "grad_norm": 1.2761468887329102,
+      "learning_rate": 9.470120519774237e-05,
+      "loss": 0.7038446068763733,
+      "step": 3006
+    },
+    {
+      "epoch": 1.2977346278317152,
+      "grad_norm": 1.2568607330322266,
+      "learning_rate": 9.469019947417815e-05,
+      "loss": 0.7063520550727844,
+      "step": 3008
+    },
+    {
+      "epoch": 1.2985976267529666,
+      "grad_norm": 1.2337472438812256,
+      "learning_rate": 9.467918297377378e-05,
+      "loss": 0.6931142807006836,
+      "step": 3010
+    },
+    {
+      "epoch": 1.2994606256742178,
+      "grad_norm": 1.1557739973068237,
+      "learning_rate": 9.466815569918588e-05,
+      "loss": 0.6660345196723938,
+      "step": 3012
+    },
+    {
+      "epoch": 1.3003236245954692,
+      "grad_norm": 1.2341985702514648,
+      "learning_rate": 9.465711765307357e-05,
+      "loss": 0.6522690653800964,
+      "step": 3014
+    },
+    {
+      "epoch": 1.3011866235167207,
+      "grad_norm": 1.0495452880859375,
+      "learning_rate": 9.464606883809867e-05,
+      "loss": 0.6304616332054138,
+      "step": 3016
+    },
+    {
+      "epoch": 1.302049622437972,
+      "grad_norm": 1.1920377016067505,
+      "learning_rate": 9.463500925692554e-05,
+      "loss": 0.7271291613578796,
+      "step": 3018
+    },
+    {
+      "epoch": 1.3029126213592233,
+      "grad_norm": 1.0697529315948486,
+      "learning_rate": 9.462393891222112e-05,
+      "loss": 0.5725325345993042,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3037756202804747,
+      "grad_norm": 1.0559576749801636,
+      "learning_rate": 9.461285780665502e-05,
+      "loss": 0.5725870728492737,
+      "step": 3022
+    },
+    {
+      "epoch": 1.304638619201726,
+      "grad_norm": 1.1314871311187744,
+      "learning_rate": 9.460176594289937e-05,
+      "loss": 0.6498687863349915,
+      "step": 3024
+    },
+    {
+      "epoch": 1.3055016181229773,
+      "grad_norm": 1.0620028972625732,
+      "learning_rate": 9.459066332362892e-05,
+      "loss": 0.5473160147666931,
+      "step": 3026
+    },
+    {
+      "epoch": 1.3063646170442287,
+      "grad_norm": 1.245997667312622,
+      "learning_rate": 9.457954995152103e-05,
+      "loss": 0.6976684927940369,
+      "step": 3028
+    },
+    {
+      "epoch": 1.30722761596548,
+      "grad_norm": 1.0973089933395386,
+      "learning_rate": 9.456842582925565e-05,
+      "loss": 0.6469701528549194,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3080906148867313,
+      "grad_norm": 1.1558068990707397,
+      "learning_rate": 9.455729095951527e-05,
+      "loss": 0.6712765097618103,
+      "step": 3032
+    },
+    {
+      "epoch": 1.3089536138079827,
+      "grad_norm": 1.0367828607559204,
+      "learning_rate": 9.454614534498506e-05,
+      "loss": 0.6504483222961426,
+      "step": 3034
+    },
+    {
+      "epoch": 1.309816612729234,
+      "grad_norm": 1.3485349416732788,
+      "learning_rate": 9.453498898835272e-05,
+      "loss": 0.7000672221183777,
+      "step": 3036
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 1.2435184717178345,
+      "learning_rate": 9.452382189230853e-05,
+      "loss": 0.6946983337402344,
+      "step": 3038
+    },
+    {
+      "epoch": 1.3115426105717367,
+      "grad_norm": 1.0728851556777954,
+      "learning_rate": 9.451264405954541e-05,
+      "loss": 0.5928027033805847,
+      "step": 3040
+    },
+    {
+      "epoch": 1.3124056094929881,
+      "grad_norm": 1.2320033311843872,
+      "learning_rate": 9.450145549275884e-05,
+      "loss": 0.6589536070823669,
+      "step": 3042
+    },
+    {
+      "epoch": 1.3132686084142395,
+      "grad_norm": 1.143538236618042,
+      "learning_rate": 9.449025619464689e-05,
+      "loss": 0.6516439914703369,
+      "step": 3044
+    },
+    {
+      "epoch": 1.314131607335491,
+      "grad_norm": 1.3419140577316284,
+      "learning_rate": 9.447904616791023e-05,
+      "loss": 0.6889938116073608,
+      "step": 3046
+    },
+    {
+      "epoch": 1.3149946062567421,
+      "grad_norm": 1.4128848314285278,
+      "learning_rate": 9.44678254152521e-05,
+      "loss": 0.7203409075737,
+      "step": 3048
+    },
+    {
+      "epoch": 1.3158576051779936,
+      "grad_norm": 1.1409581899642944,
+      "learning_rate": 9.445659393937833e-05,
+      "loss": 0.656524658203125,
+      "step": 3050
+    },
+    {
+      "epoch": 1.316720604099245,
+      "grad_norm": 1.0904145240783691,
+      "learning_rate": 9.444535174299734e-05,
+      "loss": 0.6888052821159363,
+      "step": 3052
+    },
+    {
+      "epoch": 1.3175836030204962,
+      "grad_norm": 1.1533669233322144,
+      "learning_rate": 9.443409882882014e-05,
+      "loss": 0.6149572730064392,
+      "step": 3054
+    },
+    {
+      "epoch": 1.3184466019417476,
+      "grad_norm": 1.0775257349014282,
+      "learning_rate": 9.442283519956031e-05,
+      "loss": 0.6445736885070801,
+      "step": 3056
+    },
+    {
+      "epoch": 1.319309600862999,
+      "grad_norm": 1.1126797199249268,
+      "learning_rate": 9.441156085793403e-05,
+      "loss": 0.6723806858062744,
+      "step": 3058
+    },
+    {
+      "epoch": 1.3201725997842502,
+      "grad_norm": 1.1584032773971558,
+      "learning_rate": 9.440027580666008e-05,
+      "loss": 0.5901219248771667,
+      "step": 3060
+    },
+    {
+      "epoch": 1.3210355987055016,
+      "grad_norm": 1.0951672792434692,
+      "learning_rate": 9.438898004845975e-05,
+      "loss": 0.6816843748092651,
+      "step": 3062
+    },
+    {
+      "epoch": 1.321898597626753,
+      "grad_norm": 1.2165061235427856,
+      "learning_rate": 9.4377673586057e-05,
+      "loss": 0.6883957386016846,
+      "step": 3064
+    },
+    {
+      "epoch": 1.3227615965480042,
+      "grad_norm": 1.1537514925003052,
+      "learning_rate": 9.436635642217832e-05,
+      "loss": 0.6828808188438416,
+      "step": 3066
+    },
+    {
+      "epoch": 1.3236245954692556,
+      "grad_norm": 1.2186168432235718,
+      "learning_rate": 9.435502855955281e-05,
+      "loss": 0.6086717844009399,
+      "step": 3068
+    },
+    {
+      "epoch": 1.324487594390507,
+      "grad_norm": 1.0345975160598755,
+      "learning_rate": 9.43436900009121e-05,
+      "loss": 0.6000884771347046,
+      "step": 3070
+    },
+    {
+      "epoch": 1.3253505933117584,
+      "grad_norm": 1.1357595920562744,
+      "learning_rate": 9.433234074899044e-05,
+      "loss": 0.6812195181846619,
+      "step": 3072
+    },
+    {
+      "epoch": 1.3262135922330098,
+      "grad_norm": 1.1914640665054321,
+      "learning_rate": 9.432098080652467e-05,
+      "loss": 0.6666595935821533,
+      "step": 3074
+    },
+    {
+      "epoch": 1.327076591154261,
+      "grad_norm": 1.0881390571594238,
+      "learning_rate": 9.430961017625418e-05,
+      "loss": 0.6455721259117126,
+      "step": 3076
+    },
+    {
+      "epoch": 1.3279395900755124,
+      "grad_norm": 1.1859697103500366,
+      "learning_rate": 9.429822886092093e-05,
+      "loss": 0.5861079692840576,
+      "step": 3078
+    },
+    {
+      "epoch": 1.3288025889967638,
+      "grad_norm": 1.083703637123108,
+      "learning_rate": 9.42868368632695e-05,
+      "loss": 0.6173920035362244,
+      "step": 3080
+    },
+    {
+      "epoch": 1.329665587918015,
+      "grad_norm": 1.2820687294006348,
+      "learning_rate": 9.427543418604701e-05,
+      "loss": 0.817629873752594,
+      "step": 3082
+    },
+    {
+      "epoch": 1.3305285868392664,
+      "grad_norm": 1.1267426013946533,
+      "learning_rate": 9.426402083200315e-05,
+      "loss": 0.7454650402069092,
+      "step": 3084
+    },
+    {
+      "epoch": 1.3313915857605179,
+      "grad_norm": 1.0095305442810059,
+      "learning_rate": 9.425259680389022e-05,
+      "loss": 0.625328779220581,
+      "step": 3086
+    },
+    {
+      "epoch": 1.332254584681769,
+      "grad_norm": 1.183151364326477,
+      "learning_rate": 9.424116210446305e-05,
+      "loss": 0.5977230072021484,
+      "step": 3088
+    },
+    {
+      "epoch": 1.3331175836030205,
+      "grad_norm": 1.1312501430511475,
+      "learning_rate": 9.42297167364791e-05,
+      "loss": 0.6251013278961182,
+      "step": 3090
+    },
+    {
+      "epoch": 1.3339805825242719,
+      "grad_norm": 1.3226121664047241,
+      "learning_rate": 9.421826070269832e-05,
+      "loss": 0.7469011545181274,
+      "step": 3092
+    },
+    {
+      "epoch": 1.334843581445523,
+      "grad_norm": 1.1389979124069214,
+      "learning_rate": 9.420679400588334e-05,
+      "loss": 0.7012571096420288,
+      "step": 3094
+    },
+    {
+      "epoch": 1.3357065803667745,
+      "grad_norm": 1.2065027952194214,
+      "learning_rate": 9.419531664879925e-05,
+      "loss": 0.7112689018249512,
+      "step": 3096
+    },
+    {
+      "epoch": 1.3365695792880259,
+      "grad_norm": 1.2532200813293457,
+      "learning_rate": 9.418382863421379e-05,
+      "loss": 0.6219019293785095,
+      "step": 3098
+    },
+    {
+      "epoch": 1.3374325782092773,
+      "grad_norm": 1.1859256029129028,
+      "learning_rate": 9.417232996489724e-05,
+      "loss": 0.6905092597007751,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3374325782092773,
+      "eval_loss": 0.7127581834793091,
+      "eval_runtime": 821.1346,
+      "eval_samples_per_second": 2.509,
+      "eval_steps_per_second": 2.509,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3382955771305287,
+      "grad_norm": 1.172967791557312,
+      "learning_rate": 9.416082064362244e-05,
+      "loss": 0.6473057866096497,
+      "step": 3102
+    },
+    {
+      "epoch": 1.33915857605178,
+      "grad_norm": 1.2966243028640747,
+      "learning_rate": 9.414930067316482e-05,
+      "loss": 0.7377036213874817,
+      "step": 3104
+    },
+    {
+      "epoch": 1.3400215749730313,
+      "grad_norm": 1.0844253301620483,
+      "learning_rate": 9.413777005630237e-05,
+      "loss": 0.6842042207717896,
+      "step": 3106
+    },
+    {
+      "epoch": 1.3408845738942827,
+      "grad_norm": 0.930098831653595,
+      "learning_rate": 9.412622879581565e-05,
+      "loss": 0.662327229976654,
+      "step": 3108
+    },
+    {
+      "epoch": 1.341747572815534,
+      "grad_norm": 1.1382031440734863,
+      "learning_rate": 9.411467689448777e-05,
+      "loss": 0.6979284882545471,
+      "step": 3110
+    },
+    {
+      "epoch": 1.3426105717367853,
+      "grad_norm": 1.1555129289627075,
+      "learning_rate": 9.410311435510444e-05,
+      "loss": 0.7032427191734314,
+      "step": 3112
+    },
+    {
+      "epoch": 1.3434735706580367,
+      "grad_norm": 1.147678017616272,
+      "learning_rate": 9.409154118045389e-05,
+      "loss": 0.6282536387443542,
+      "step": 3114
+    },
+    {
+      "epoch": 1.344336569579288,
+      "grad_norm": 1.0925829410552979,
+      "learning_rate": 9.407995737332696e-05,
+      "loss": 0.6572934985160828,
+      "step": 3116
+    },
+    {
+      "epoch": 1.3451995685005393,
+      "grad_norm": 1.3632638454437256,
+      "learning_rate": 9.4068362936517e-05,
+      "loss": 0.7233583927154541,
+      "step": 3118
+    },
+    {
+      "epoch": 1.3460625674217908,
+      "grad_norm": 1.3994121551513672,
+      "learning_rate": 9.405675787282e-05,
+      "loss": 0.6465377807617188,
+      "step": 3120
+    },
+    {
+      "epoch": 1.346925566343042,
+      "grad_norm": 1.1862388849258423,
+      "learning_rate": 9.404514218503444e-05,
+      "loss": 0.6348391771316528,
+      "step": 3122
+    },
+    {
+      "epoch": 1.3477885652642934,
+      "grad_norm": 1.1724227666854858,
+      "learning_rate": 9.403351587596141e-05,
+      "loss": 0.6559093594551086,
+      "step": 3124
+    },
+    {
+      "epoch": 1.3486515641855448,
+      "grad_norm": 1.175567865371704,
+      "learning_rate": 9.402187894840452e-05,
+      "loss": 0.6693678498268127,
+      "step": 3126
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 1.1273375749588013,
+      "learning_rate": 9.401023140516999e-05,
+      "loss": 0.6766137480735779,
+      "step": 3128
+    },
+    {
+      "epoch": 1.3503775620280476,
+      "grad_norm": 1.2515393495559692,
+      "learning_rate": 9.399857324906652e-05,
+      "loss": 0.7133797407150269,
+      "step": 3130
+    },
+    {
+      "epoch": 1.3512405609492988,
+      "grad_norm": 1.1575038433074951,
+      "learning_rate": 9.398690448290547e-05,
+      "loss": 0.6572166681289673,
+      "step": 3132
+    },
+    {
+      "epoch": 1.3521035598705502,
+      "grad_norm": 1.2350742816925049,
+      "learning_rate": 9.397522510950072e-05,
+      "loss": 0.6389561891555786,
+      "step": 3134
+    },
+    {
+      "epoch": 1.3529665587918016,
+      "grad_norm": 0.9648056626319885,
+      "learning_rate": 9.396353513166867e-05,
+      "loss": 0.5810164213180542,
+      "step": 3136
+    },
+    {
+      "epoch": 1.3538295577130528,
+      "grad_norm": 1.147218108177185,
+      "learning_rate": 9.39518345522283e-05,
+      "loss": 0.7157397866249084,
+      "step": 3138
+    },
+    {
+      "epoch": 1.3546925566343042,
+      "grad_norm": 1.1948896646499634,
+      "learning_rate": 9.394012337400117e-05,
+      "loss": 0.6749072670936584,
+      "step": 3140
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 1.0945216417312622,
+      "learning_rate": 9.392840159981136e-05,
+      "loss": 0.63704514503479,
+      "step": 3142
+    },
+    {
+      "epoch": 1.3564185544768068,
+      "grad_norm": 1.0654480457305908,
+      "learning_rate": 9.391666923248555e-05,
+      "loss": 0.6804014444351196,
+      "step": 3144
+    },
+    {
+      "epoch": 1.3572815533980582,
+      "grad_norm": 1.3977962732315063,
+      "learning_rate": 9.390492627485292e-05,
+      "loss": 0.7055573463439941,
+      "step": 3146
+    },
+    {
+      "epoch": 1.3581445523193096,
+      "grad_norm": 1.1607128381729126,
+      "learning_rate": 9.389317272974526e-05,
+      "loss": 0.7383730411529541,
+      "step": 3148
+    },
+    {
+      "epoch": 1.3590075512405608,
+      "grad_norm": 1.192213773727417,
+      "learning_rate": 9.388140859999686e-05,
+      "loss": 0.7339736223220825,
+      "step": 3150
+    },
+    {
+      "epoch": 1.3598705501618122,
+      "grad_norm": 1.1597269773483276,
+      "learning_rate": 9.386963388844459e-05,
+      "loss": 0.6564369797706604,
+      "step": 3152
+    },
+    {
+      "epoch": 1.3607335490830637,
+      "grad_norm": 1.2374637126922607,
+      "learning_rate": 9.385784859792788e-05,
+      "loss": 0.7393131256103516,
+      "step": 3154
+    },
+    {
+      "epoch": 1.361596548004315,
+      "grad_norm": 1.1975921392440796,
+      "learning_rate": 9.384605273128868e-05,
+      "loss": 0.6794763207435608,
+      "step": 3156
+    },
+    {
+      "epoch": 1.3624595469255665,
+      "grad_norm": 1.158282995223999,
+      "learning_rate": 9.383424629137154e-05,
+      "loss": 0.5906237363815308,
+      "step": 3158
+    },
+    {
+      "epoch": 1.3633225458468177,
+      "grad_norm": 1.0605288743972778,
+      "learning_rate": 9.38224292810235e-05,
+      "loss": 0.6813581585884094,
+      "step": 3160
+    },
+    {
+      "epoch": 1.364185544768069,
+      "grad_norm": 0.9797501564025879,
+      "learning_rate": 9.38106017030942e-05,
+      "loss": 0.639259934425354,
+      "step": 3162
+    },
+    {
+      "epoch": 1.3650485436893205,
+      "grad_norm": 1.1023533344268799,
+      "learning_rate": 9.379876356043578e-05,
+      "loss": 0.667378306388855,
+      "step": 3164
+    },
+    {
+      "epoch": 1.3659115426105717,
+      "grad_norm": 1.1105194091796875,
+      "learning_rate": 9.378691485590299e-05,
+      "loss": 0.671601414680481,
+      "step": 3166
+    },
+    {
+      "epoch": 1.366774541531823,
+      "grad_norm": 0.9441736936569214,
+      "learning_rate": 9.377505559235308e-05,
+      "loss": 0.623394787311554,
+      "step": 3168
+    },
+    {
+      "epoch": 1.3676375404530745,
+      "grad_norm": 1.308560848236084,
+      "learning_rate": 9.376318577264585e-05,
+      "loss": 0.6558915972709656,
+      "step": 3170
+    },
+    {
+      "epoch": 1.3685005393743257,
+      "grad_norm": 1.065693974494934,
+      "learning_rate": 9.375130539964363e-05,
+      "loss": 0.6431660652160645,
+      "step": 3172
+    },
+    {
+      "epoch": 1.369363538295577,
+      "grad_norm": 1.027817726135254,
+      "learning_rate": 9.373941447621137e-05,
+      "loss": 0.6541599631309509,
+      "step": 3174
+    },
+    {
+      "epoch": 1.3702265372168285,
+      "grad_norm": 1.1651864051818848,
+      "learning_rate": 9.372751300521649e-05,
+      "loss": 0.6710350513458252,
+      "step": 3176
+    },
+    {
+      "epoch": 1.3710895361380797,
+      "grad_norm": 1.1066170930862427,
+      "learning_rate": 9.371560098952896e-05,
+      "loss": 0.6525241732597351,
+      "step": 3178
+    },
+    {
+      "epoch": 1.3719525350593311,
+      "grad_norm": 1.0942546129226685,
+      "learning_rate": 9.370367843202133e-05,
+      "loss": 0.6362038850784302,
+      "step": 3180
+    },
+    {
+      "epoch": 1.3728155339805825,
+      "grad_norm": 1.175432562828064,
+      "learning_rate": 9.369174533556868e-05,
+      "loss": 0.6017009615898132,
+      "step": 3182
+    },
+    {
+      "epoch": 1.373678532901834,
+      "grad_norm": 1.3605095148086548,
+      "learning_rate": 9.367980170304857e-05,
+      "loss": 0.6853877305984497,
+      "step": 3184
+    },
+    {
+      "epoch": 1.3745415318230851,
+      "grad_norm": 1.224565029144287,
+      "learning_rate": 9.366784753734121e-05,
+      "loss": 0.6434065699577332,
+      "step": 3186
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 1.1151230335235596,
+      "learning_rate": 9.365588284132927e-05,
+      "loss": 0.596630334854126,
+      "step": 3188
+    },
+    {
+      "epoch": 1.376267529665588,
+      "grad_norm": 1.2674700021743774,
+      "learning_rate": 9.364390761789799e-05,
+      "loss": 0.6716568470001221,
+      "step": 3190
+    },
+    {
+      "epoch": 1.3771305285868394,
+      "grad_norm": 1.3221330642700195,
+      "learning_rate": 9.363192186993512e-05,
+      "loss": 0.6230685710906982,
+      "step": 3192
+    },
+    {
+      "epoch": 1.3779935275080906,
+      "grad_norm": 1.2978439331054688,
+      "learning_rate": 9.3619925600331e-05,
+      "loss": 0.6495190858840942,
+      "step": 3194
+    },
+    {
+      "epoch": 1.378856526429342,
+      "grad_norm": 1.1542012691497803,
+      "learning_rate": 9.360791881197846e-05,
+      "loss": 0.6534912586212158,
+      "step": 3196
+    },
+    {
+      "epoch": 1.3797195253505934,
+      "grad_norm": 1.1294280290603638,
+      "learning_rate": 9.359590150777288e-05,
+      "loss": 0.5706549286842346,
+      "step": 3198
+    },
+    {
+      "epoch": 1.3805825242718446,
+      "grad_norm": 1.0622307062149048,
+      "learning_rate": 9.358387369061218e-05,
+      "loss": 0.6193811893463135,
+      "step": 3200
+    },
+    {
+      "epoch": 1.3805825242718446,
+      "eval_loss": 0.7103806734085083,
+      "eval_runtime": 831.4047,
+      "eval_samples_per_second": 2.478,
+      "eval_steps_per_second": 2.478,
+      "step": 3200
+    },
+    {
+      "epoch": 1.381445523193096,
+      "grad_norm": 1.1722023487091064,
+      "learning_rate": 9.357183536339683e-05,
+      "loss": 0.6021195650100708,
+      "step": 3202
+    },
+    {
+      "epoch": 1.3823085221143474,
+      "grad_norm": 1.2733373641967773,
+      "learning_rate": 9.355978652902978e-05,
+      "loss": 0.7195428013801575,
+      "step": 3204
+    },
+    {
+      "epoch": 1.3831715210355986,
+      "grad_norm": 1.019609808921814,
+      "learning_rate": 9.354772719041658e-05,
+      "loss": 0.6314851641654968,
+      "step": 3206
+    },
+    {
+      "epoch": 1.38403451995685,
+      "grad_norm": 1.36198890209198,
+      "learning_rate": 9.35356573504653e-05,
+      "loss": 0.722943902015686,
+      "step": 3208
+    },
+    {
+      "epoch": 1.3848975188781014,
+      "grad_norm": 1.1291067600250244,
+      "learning_rate": 9.35235770120865e-05,
+      "loss": 0.7058439254760742,
+      "step": 3210
+    },
+    {
+      "epoch": 1.3857605177993528,
+      "grad_norm": 1.2160366773605347,
+      "learning_rate": 9.351148617819332e-05,
+      "loss": 0.6186655759811401,
+      "step": 3212
+    },
+    {
+      "epoch": 1.386623516720604,
+      "grad_norm": 1.1663079261779785,
+      "learning_rate": 9.349938485170139e-05,
+      "loss": 0.5912084579467773,
+      "step": 3214
+    },
+    {
+      "epoch": 1.3874865156418554,
+      "grad_norm": 1.2145130634307861,
+      "learning_rate": 9.34872730355289e-05,
+      "loss": 0.6494051218032837,
+      "step": 3216
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 1.018477439880371,
+      "learning_rate": 9.347515073259657e-05,
+      "loss": 0.6258159279823303,
+      "step": 3218
+    },
+    {
+      "epoch": 1.3892125134843583,
+      "grad_norm": 1.2019606828689575,
+      "learning_rate": 9.346301794582761e-05,
+      "loss": 0.7076614499092102,
+      "step": 3220
+    },
+    {
+      "epoch": 1.3900755124056094,
+      "grad_norm": 1.072648048400879,
+      "learning_rate": 9.345087467814782e-05,
+      "loss": 0.5958849191665649,
+      "step": 3222
+    },
+    {
+      "epoch": 1.3909385113268609,
+      "grad_norm": 1.087908387184143,
+      "learning_rate": 9.343872093248546e-05,
+      "loss": 0.5659268498420715,
+      "step": 3224
+    },
+    {
+      "epoch": 1.3918015102481123,
+      "grad_norm": 1.2628878355026245,
+      "learning_rate": 9.342655671177138e-05,
+      "loss": 0.6369889974594116,
+      "step": 3226
+    },
+    {
+      "epoch": 1.3926645091693635,
+      "grad_norm": 1.1580097675323486,
+      "learning_rate": 9.341438201893891e-05,
+      "loss": 0.64048171043396,
+      "step": 3228
+    },
+    {
+      "epoch": 1.3935275080906149,
+      "grad_norm": 1.2230379581451416,
+      "learning_rate": 9.340219685692395e-05,
+      "loss": 0.7039018273353577,
+      "step": 3230
+    },
+    {
+      "epoch": 1.3943905070118663,
+      "grad_norm": 1.1932404041290283,
+      "learning_rate": 9.339000122866486e-05,
+      "loss": 0.6461225152015686,
+      "step": 3232
+    },
+    {
+      "epoch": 1.3952535059331175,
+      "grad_norm": 1.4337340593338013,
+      "learning_rate": 9.337779513710257e-05,
+      "loss": 0.6538844108581543,
+      "step": 3234
+    },
+    {
+      "epoch": 1.396116504854369,
+      "grad_norm": 1.2300320863723755,
+      "learning_rate": 9.336557858518055e-05,
+      "loss": 0.6788939237594604,
+      "step": 3236
+    },
+    {
+      "epoch": 1.3969795037756203,
+      "grad_norm": 1.1031018495559692,
+      "learning_rate": 9.335335157584473e-05,
+      "loss": 0.6180470585823059,
+      "step": 3238
+    },
+    {
+      "epoch": 1.3978425026968715,
+      "grad_norm": 1.1876481771469116,
+      "learning_rate": 9.334111411204363e-05,
+      "loss": 0.7097930908203125,
+      "step": 3240
+    },
+    {
+      "epoch": 1.398705501618123,
+      "grad_norm": 1.1510672569274902,
+      "learning_rate": 9.332886619672823e-05,
+      "loss": 0.6521070003509521,
+      "step": 3242
+    },
+    {
+      "epoch": 1.3995685005393743,
+      "grad_norm": 1.2715339660644531,
+      "learning_rate": 9.331660783285208e-05,
+      "loss": 0.6738741397857666,
+      "step": 3244
+    },
+    {
+      "epoch": 1.4004314994606257,
+      "grad_norm": 1.1528351306915283,
+      "learning_rate": 9.330433902337124e-05,
+      "loss": 0.6261277794837952,
+      "step": 3246
+    },
+    {
+      "epoch": 1.4012944983818771,
+      "grad_norm": 1.1715319156646729,
+      "learning_rate": 9.329205977124424e-05,
+      "loss": 0.7071060538291931,
+      "step": 3248
+    },
+    {
+      "epoch": 1.4021574973031283,
+      "grad_norm": 1.1527135372161865,
+      "learning_rate": 9.32797700794322e-05,
+      "loss": 0.6339676380157471,
+      "step": 3250
+    },
+    {
+      "epoch": 1.4030204962243797,
+      "grad_norm": 1.1546231508255005,
+      "learning_rate": 9.32674699508987e-05,
+      "loss": 0.6572067737579346,
+      "step": 3252
+    },
+    {
+      "epoch": 1.4038834951456312,
+      "grad_norm": 1.1085712909698486,
+      "learning_rate": 9.325515938860988e-05,
+      "loss": 0.6182453036308289,
+      "step": 3254
+    },
+    {
+      "epoch": 1.4047464940668823,
+      "grad_norm": 1.2346831560134888,
+      "learning_rate": 9.324283839553437e-05,
+      "loss": 0.6558552980422974,
+      "step": 3256
+    },
+    {
+      "epoch": 1.4056094929881338,
+      "grad_norm": 1.1873912811279297,
+      "learning_rate": 9.323050697464331e-05,
+      "loss": 0.6406651735305786,
+      "step": 3258
+    },
+    {
+      "epoch": 1.4064724919093852,
+      "grad_norm": 1.5326859951019287,
+      "learning_rate": 9.32181651289104e-05,
+      "loss": 0.6869475841522217,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4073354908306364,
+      "grad_norm": 1.2141598463058472,
+      "learning_rate": 9.320581286131178e-05,
+      "loss": 0.6805770397186279,
+      "step": 3262
+    },
+    {
+      "epoch": 1.4081984897518878,
+      "grad_norm": 1.232690453529358,
+      "learning_rate": 9.319345017482615e-05,
+      "loss": 0.6193398833274841,
+      "step": 3264
+    },
+    {
+      "epoch": 1.4090614886731392,
+      "grad_norm": 1.0616317987442017,
+      "learning_rate": 9.318107707243475e-05,
+      "loss": 0.5868768095970154,
+      "step": 3266
+    },
+    {
+      "epoch": 1.4099244875943904,
+      "grad_norm": 1.091909408569336,
+      "learning_rate": 9.316869355712127e-05,
+      "loss": 0.6360217928886414,
+      "step": 3268
+    },
+    {
+      "epoch": 1.4107874865156418,
+      "grad_norm": 1.0247875452041626,
+      "learning_rate": 9.315629963187194e-05,
+      "loss": 0.6082828044891357,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4116504854368932,
+      "grad_norm": 1.2529716491699219,
+      "learning_rate": 9.314389529967551e-05,
+      "loss": 0.6340673565864563,
+      "step": 3272
+    },
+    {
+      "epoch": 1.4125134843581446,
+      "grad_norm": 0.9816965460777283,
+      "learning_rate": 9.31314805635232e-05,
+      "loss": 0.6333131790161133,
+      "step": 3274
+    },
+    {
+      "epoch": 1.413376483279396,
+      "grad_norm": 1.2140034437179565,
+      "learning_rate": 9.311905542640882e-05,
+      "loss": 0.7256702780723572,
+      "step": 3276
+    },
+    {
+      "epoch": 1.4142394822006472,
+      "grad_norm": 1.2171046733856201,
+      "learning_rate": 9.310661989132859e-05,
+      "loss": 0.7062730193138123,
+      "step": 3278
+    },
+    {
+      "epoch": 1.4151024811218986,
+      "grad_norm": 1.0224506855010986,
+      "learning_rate": 9.309417396128131e-05,
+      "loss": 0.6192088723182678,
+      "step": 3280
+    },
+    {
+      "epoch": 1.41596548004315,
+      "grad_norm": 1.1904140710830688,
+      "learning_rate": 9.308171763926823e-05,
+      "loss": 0.6522074341773987,
+      "step": 3282
+    },
+    {
+      "epoch": 1.4168284789644012,
+      "grad_norm": 1.1734132766723633,
+      "learning_rate": 9.306925092829317e-05,
+      "loss": 0.6472417712211609,
+      "step": 3284
+    },
+    {
+      "epoch": 1.4176914778856526,
+      "grad_norm": 1.2349748611450195,
+      "learning_rate": 9.30567738313624e-05,
+      "loss": 0.6990411877632141,
+      "step": 3286
+    },
+    {
+      "epoch": 1.418554476806904,
+      "grad_norm": 1.1732749938964844,
+      "learning_rate": 9.304428635148472e-05,
+      "loss": 0.6181937456130981,
+      "step": 3288
+    },
+    {
+      "epoch": 1.4194174757281552,
+      "grad_norm": 1.3659436702728271,
+      "learning_rate": 9.303178849167142e-05,
+      "loss": 0.7471824288368225,
+      "step": 3290
+    },
+    {
+      "epoch": 1.4202804746494067,
+      "grad_norm": 1.0865631103515625,
+      "learning_rate": 9.301928025493634e-05,
+      "loss": 0.5541732907295227,
+      "step": 3292
+    },
+    {
+      "epoch": 1.421143473570658,
+      "grad_norm": 1.2114793062210083,
+      "learning_rate": 9.300676164429574e-05,
+      "loss": 0.6934989094734192,
+      "step": 3294
+    },
+    {
+      "epoch": 1.4220064724919093,
+      "grad_norm": 1.1783509254455566,
+      "learning_rate": 9.299423266276846e-05,
+      "loss": 0.6887713670730591,
+      "step": 3296
+    },
+    {
+      "epoch": 1.4228694714131607,
+      "grad_norm": 1.3001595735549927,
+      "learning_rate": 9.298169331337577e-05,
+      "loss": 0.6956127882003784,
+      "step": 3298
+    },
+    {
+      "epoch": 1.423732470334412,
+      "grad_norm": 1.1657462120056152,
+      "learning_rate": 9.296914359914151e-05,
+      "loss": 0.690113365650177,
+      "step": 3300
+    },
+    {
+      "epoch": 1.423732470334412,
+      "eval_loss": 0.7086145281791687,
+      "eval_runtime": 838.887,
+      "eval_samples_per_second": 2.456,
+      "eval_steps_per_second": 2.456,
+      "step": 3300
+    },
+    {
+      "epoch": 1.4245954692556635,
+      "grad_norm": 1.1632269620895386,
+      "learning_rate": 9.295658352309198e-05,
+      "loss": 0.6801355481147766,
+      "step": 3302
+    },
+    {
+      "epoch": 1.425458468176915,
+      "grad_norm": 1.3355532884597778,
+      "learning_rate": 9.2944013088256e-05,
+      "loss": 0.6801818013191223,
+      "step": 3304
+    },
+    {
+      "epoch": 1.426321467098166,
+      "grad_norm": 1.0911835432052612,
+      "learning_rate": 9.293143229766484e-05,
+      "loss": 0.633465051651001,
+      "step": 3306
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 1.1873823404312134,
+      "learning_rate": 9.29188411543523e-05,
+      "loss": 0.6677361130714417,
+      "step": 3308
+    },
+    {
+      "epoch": 1.428047464940669,
+      "grad_norm": 1.2115873098373413,
+      "learning_rate": 9.290623966135472e-05,
+      "loss": 0.6741982102394104,
+      "step": 3310
+    },
+    {
+      "epoch": 1.42891046386192,
+      "grad_norm": 1.2389469146728516,
+      "learning_rate": 9.289362782171084e-05,
+      "loss": 0.7169438600540161,
+      "step": 3312
+    },
+    {
+      "epoch": 1.4297734627831715,
+      "grad_norm": 1.2088500261306763,
+      "learning_rate": 9.288100563846201e-05,
+      "loss": 0.7112365961074829,
+      "step": 3314
+    },
+    {
+      "epoch": 1.430636461704423,
+      "grad_norm": 1.1279292106628418,
+      "learning_rate": 9.286837311465196e-05,
+      "loss": 0.6910304427146912,
+      "step": 3316
+    },
+    {
+      "epoch": 1.4314994606256741,
+      "grad_norm": 1.1812585592269897,
+      "learning_rate": 9.285573025332698e-05,
+      "loss": 0.7288362979888916,
+      "step": 3318
+    },
+    {
+      "epoch": 1.4323624595469255,
+      "grad_norm": 1.5522732734680176,
+      "learning_rate": 9.284307705753587e-05,
+      "loss": 0.6918224692344666,
+      "step": 3320
+    },
+    {
+      "epoch": 1.433225458468177,
+      "grad_norm": 1.2825061082839966,
+      "learning_rate": 9.283041353032984e-05,
+      "loss": 0.6756533980369568,
+      "step": 3322
+    },
+    {
+      "epoch": 1.4340884573894281,
+      "grad_norm": 1.0962797403335571,
+      "learning_rate": 9.281773967476269e-05,
+      "loss": 0.6005351543426514,
+      "step": 3324
+    },
+    {
+      "epoch": 1.4349514563106796,
+      "grad_norm": 1.249868631362915,
+      "learning_rate": 9.280505549389063e-05,
+      "loss": 0.6576942205429077,
+      "step": 3326
+    },
+    {
+      "epoch": 1.435814455231931,
+      "grad_norm": 1.2587590217590332,
+      "learning_rate": 9.279236099077239e-05,
+      "loss": 0.6686707735061646,
+      "step": 3328
+    },
+    {
+      "epoch": 1.4366774541531824,
+      "grad_norm": 1.0961607694625854,
+      "learning_rate": 9.277965616846922e-05,
+      "loss": 0.5858182311058044,
+      "step": 3330
+    },
+    {
+      "epoch": 1.4375404530744338,
+      "grad_norm": 1.199666142463684,
+      "learning_rate": 9.276694103004481e-05,
+      "loss": 0.6661836504936218,
+      "step": 3332
+    },
+    {
+      "epoch": 1.438403451995685,
+      "grad_norm": 1.2444990873336792,
+      "learning_rate": 9.275421557856536e-05,
+      "loss": 0.6922421455383301,
+      "step": 3334
+    },
+    {
+      "epoch": 1.4392664509169364,
+      "grad_norm": 1.3198295831680298,
+      "learning_rate": 9.274147981709957e-05,
+      "loss": 0.750586986541748,
+      "step": 3336
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 1.0782018899917603,
+      "learning_rate": 9.27287337487186e-05,
+      "loss": 0.6817078590393066,
+      "step": 3338
+    },
+    {
+      "epoch": 1.440992448759439,
+      "grad_norm": 1.1463382244110107,
+      "learning_rate": 9.271597737649611e-05,
+      "loss": 0.6163688898086548,
+      "step": 3340
+    },
+    {
+      "epoch": 1.4418554476806904,
+      "grad_norm": 1.0618164539337158,
+      "learning_rate": 9.270321070350825e-05,
+      "loss": 0.6624670028686523,
+      "step": 3342
+    },
+    {
+      "epoch": 1.4427184466019418,
+      "grad_norm": 1.1389427185058594,
+      "learning_rate": 9.269043373283362e-05,
+      "loss": 0.6802749037742615,
+      "step": 3344
+    },
+    {
+      "epoch": 1.443581445523193,
+      "grad_norm": 1.082173466682434,
+      "learning_rate": 9.267764646755335e-05,
+      "loss": 0.6428707838058472,
+      "step": 3346
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 1.1741442680358887,
+      "learning_rate": 9.266484891075103e-05,
+      "loss": 0.6853904128074646,
+      "step": 3348
+    },
+    {
+      "epoch": 1.4453074433656958,
+      "grad_norm": 1.1328803300857544,
+      "learning_rate": 9.265204106551273e-05,
+      "loss": 0.6889524459838867,
+      "step": 3350
+    },
+    {
+      "epoch": 1.446170442286947,
+      "grad_norm": 1.0607874393463135,
+      "learning_rate": 9.2639222934927e-05,
+      "loss": 0.6037911176681519,
+      "step": 3352
+    },
+    {
+      "epoch": 1.4470334412081984,
+      "grad_norm": 1.1584129333496094,
+      "learning_rate": 9.262639452208487e-05,
+      "loss": 0.6671991944313049,
+      "step": 3354
+    },
+    {
+      "epoch": 1.4478964401294498,
+      "grad_norm": 1.1881403923034668,
+      "learning_rate": 9.261355583007987e-05,
+      "loss": 0.7165482044219971,
+      "step": 3356
+    },
+    {
+      "epoch": 1.4487594390507013,
+      "grad_norm": 1.2093054056167603,
+      "learning_rate": 9.260070686200797e-05,
+      "loss": 0.7108280062675476,
+      "step": 3358
+    },
+    {
+      "epoch": 1.4496224379719527,
+      "grad_norm": 1.1105186939239502,
+      "learning_rate": 9.258784762096768e-05,
+      "loss": 0.718152642250061,
+      "step": 3360
+    },
+    {
+      "epoch": 1.4504854368932039,
+      "grad_norm": 1.028255820274353,
+      "learning_rate": 9.25749781100599e-05,
+      "loss": 0.6449500322341919,
+      "step": 3362
+    },
+    {
+      "epoch": 1.4513484358144553,
+      "grad_norm": 1.235719919204712,
+      "learning_rate": 9.25620983323881e-05,
+      "loss": 0.6895458102226257,
+      "step": 3364
+    },
+    {
+      "epoch": 1.4522114347357067,
+      "grad_norm": 1.3693572282791138,
+      "learning_rate": 9.254920829105816e-05,
+      "loss": 0.7243028879165649,
+      "step": 3366
+    },
+    {
+      "epoch": 1.4530744336569579,
+      "grad_norm": 1.1795028448104858,
+      "learning_rate": 9.253630798917843e-05,
+      "loss": 0.6177746057510376,
+      "step": 3368
+    },
+    {
+      "epoch": 1.4539374325782093,
+      "grad_norm": 1.0759676694869995,
+      "learning_rate": 9.25233974298598e-05,
+      "loss": 0.6603507399559021,
+      "step": 3370
+    },
+    {
+      "epoch": 1.4548004314994607,
+      "grad_norm": 0.9495601058006287,
+      "learning_rate": 9.251047661621556e-05,
+      "loss": 0.5989331603050232,
+      "step": 3372
+    },
+    {
+      "epoch": 1.455663430420712,
+      "grad_norm": 1.1334787607192993,
+      "learning_rate": 9.249754555136154e-05,
+      "loss": 0.5743664503097534,
+      "step": 3374
+    },
+    {
+      "epoch": 1.4565264293419633,
+      "grad_norm": 1.114450216293335,
+      "learning_rate": 9.248460423841598e-05,
+      "loss": 0.5905263423919678,
+      "step": 3376
+    },
+    {
+      "epoch": 1.4573894282632147,
+      "grad_norm": 1.2400883436203003,
+      "learning_rate": 9.247165268049963e-05,
+      "loss": 0.6751091480255127,
+      "step": 3378
+    },
+    {
+      "epoch": 1.458252427184466,
+      "grad_norm": 1.2845438718795776,
+      "learning_rate": 9.245869088073569e-05,
+      "loss": 0.6872944235801697,
+      "step": 3380
+    },
+    {
+      "epoch": 1.4591154261057173,
+      "grad_norm": 1.054388165473938,
+      "learning_rate": 9.244571884224987e-05,
+      "loss": 0.6467806100845337,
+      "step": 3382
+    },
+    {
+      "epoch": 1.4599784250269687,
+      "grad_norm": 1.0959769487380981,
+      "learning_rate": 9.243273656817029e-05,
+      "loss": 0.6486450433731079,
+      "step": 3384
+    },
+    {
+      "epoch": 1.4608414239482201,
+      "grad_norm": 1.6487104892730713,
+      "learning_rate": 9.241974406162756e-05,
+      "loss": 0.7024005651473999,
+      "step": 3386
+    },
+    {
+      "epoch": 1.4617044228694713,
+      "grad_norm": 1.3018934726715088,
+      "learning_rate": 9.240674132575481e-05,
+      "loss": 0.7081685662269592,
+      "step": 3388
+    },
+    {
+      "epoch": 1.4625674217907227,
+      "grad_norm": 0.9295057058334351,
+      "learning_rate": 9.239372836368754e-05,
+      "loss": 0.5842618346214294,
+      "step": 3390
+    },
+    {
+      "epoch": 1.4634304207119742,
+      "grad_norm": 1.0556540489196777,
+      "learning_rate": 9.238070517856378e-05,
+      "loss": 0.6355351209640503,
+      "step": 3392
+    },
+    {
+      "epoch": 1.4642934196332256,
+      "grad_norm": 1.2247962951660156,
+      "learning_rate": 9.236767177352403e-05,
+      "loss": 0.6340579390525818,
+      "step": 3394
+    },
+    {
+      "epoch": 1.4651564185544768,
+      "grad_norm": 1.1625045537948608,
+      "learning_rate": 9.235462815171123e-05,
+      "loss": 0.6405525803565979,
+      "step": 3396
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 1.009028673171997,
+      "learning_rate": 9.234157431627077e-05,
+      "loss": 0.5884681940078735,
+      "step": 3398
+    },
+    {
+      "epoch": 1.4668824163969796,
+      "grad_norm": 1.174851894378662,
+      "learning_rate": 9.232851027035056e-05,
+      "loss": 0.6532622575759888,
+      "step": 3400
+    },
+    {
+      "epoch": 1.4668824163969796,
+      "eval_loss": 0.7067069411277771,
+      "eval_runtime": 829.9411,
+      "eval_samples_per_second": 2.482,
+      "eval_steps_per_second": 2.482,
+      "step": 3400
+    },
+    {
+      "epoch": 1.4677454153182308,
+      "grad_norm": 0.9843166470527649,
+      "learning_rate": 9.231543601710091e-05,
+      "loss": 0.6505036354064941,
+      "step": 3402
+    },
+    {
+      "epoch": 1.4686084142394822,
+      "grad_norm": 1.1260782480239868,
+      "learning_rate": 9.230235155967461e-05,
+      "loss": 0.5809159874916077,
+      "step": 3404
+    },
+    {
+      "epoch": 1.4694714131607336,
+      "grad_norm": 1.3986146450042725,
+      "learning_rate": 9.228925690122696e-05,
+      "loss": 0.6431854367256165,
+      "step": 3406
+    },
+    {
+      "epoch": 1.4703344120819848,
+      "grad_norm": 1.145108938217163,
+      "learning_rate": 9.227615204491562e-05,
+      "loss": 0.6126174330711365,
+      "step": 3408
+    },
+    {
+      "epoch": 1.4711974110032362,
+      "grad_norm": 1.0973509550094604,
+      "learning_rate": 9.22630369939008e-05,
+      "loss": 0.6271735429763794,
+      "step": 3410
+    },
+    {
+      "epoch": 1.4720604099244876,
+      "grad_norm": 1.0548614263534546,
+      "learning_rate": 9.224991175134515e-05,
+      "loss": 0.5894950032234192,
+      "step": 3412
+    },
+    {
+      "epoch": 1.472923408845739,
+      "grad_norm": 1.3288220167160034,
+      "learning_rate": 9.223677632041374e-05,
+      "loss": 0.6621272563934326,
+      "step": 3414
+    },
+    {
+      "epoch": 1.4737864077669902,
+      "grad_norm": 1.0312891006469727,
+      "learning_rate": 9.222363070427411e-05,
+      "loss": 0.6911466717720032,
+      "step": 3416
+    },
+    {
+      "epoch": 1.4746494066882416,
+      "grad_norm": 1.1039631366729736,
+      "learning_rate": 9.221047490609629e-05,
+      "loss": 0.5945214033126831,
+      "step": 3418
+    },
+    {
+      "epoch": 1.475512405609493,
+      "grad_norm": 1.1555835008621216,
+      "learning_rate": 9.219730892905273e-05,
+      "loss": 0.6231427192687988,
+      "step": 3420
+    },
+    {
+      "epoch": 1.4763754045307445,
+      "grad_norm": 1.0504940748214722,
+      "learning_rate": 9.218413277631835e-05,
+      "loss": 0.6413036584854126,
+      "step": 3422
+    },
+    {
+      "epoch": 1.4772384034519956,
+      "grad_norm": 1.0855275392532349,
+      "learning_rate": 9.217094645107053e-05,
+      "loss": 0.663884699344635,
+      "step": 3424
+    },
+    {
+      "epoch": 1.478101402373247,
+      "grad_norm": 1.068493366241455,
+      "learning_rate": 9.215774995648907e-05,
+      "loss": 0.6161345839500427,
+      "step": 3426
+    },
+    {
+      "epoch": 1.4789644012944985,
+      "grad_norm": 1.0962451696395874,
+      "learning_rate": 9.214454329575626e-05,
+      "loss": 0.6693077087402344,
+      "step": 3428
+    },
+    {
+      "epoch": 1.4798274002157497,
+      "grad_norm": 1.070582389831543,
+      "learning_rate": 9.213132647205682e-05,
+      "loss": 0.6362645030021667,
+      "step": 3430
+    },
+    {
+      "epoch": 1.480690399137001,
+      "grad_norm": 1.201202154159546,
+      "learning_rate": 9.211809948857793e-05,
+      "loss": 0.7378971576690674,
+      "step": 3432
+    },
+    {
+      "epoch": 1.4815533980582525,
+      "grad_norm": 1.1988805532455444,
+      "learning_rate": 9.210486234850921e-05,
+      "loss": 0.686621367931366,
+      "step": 3434
+    },
+    {
+      "epoch": 1.4824163969795037,
+      "grad_norm": 1.2886041402816772,
+      "learning_rate": 9.209161505504274e-05,
+      "loss": 0.6873248219490051,
+      "step": 3436
+    },
+    {
+      "epoch": 1.483279395900755,
+      "grad_norm": 1.0028132200241089,
+      "learning_rate": 9.207835761137307e-05,
+      "loss": 0.624991774559021,
+      "step": 3438
+    },
+    {
+      "epoch": 1.4841423948220065,
+      "grad_norm": 1.139201045036316,
+      "learning_rate": 9.206509002069715e-05,
+      "loss": 0.6064573526382446,
+      "step": 3440
+    },
+    {
+      "epoch": 1.4850053937432577,
+      "grad_norm": 1.0371732711791992,
+      "learning_rate": 9.205181228621438e-05,
+      "loss": 0.6139655709266663,
+      "step": 3442
+    },
+    {
+      "epoch": 1.485868392664509,
+      "grad_norm": 1.1208233833312988,
+      "learning_rate": 9.203852441112665e-05,
+      "loss": 0.6176325082778931,
+      "step": 3444
+    },
+    {
+      "epoch": 1.4867313915857605,
+      "grad_norm": 1.2438154220581055,
+      "learning_rate": 9.20252263986383e-05,
+      "loss": 0.6920754313468933,
+      "step": 3446
+    },
+    {
+      "epoch": 1.487594390507012,
+      "grad_norm": 1.2117916345596313,
+      "learning_rate": 9.201191825195604e-05,
+      "loss": 0.6859762072563171,
+      "step": 3448
+    },
+    {
+      "epoch": 1.4884573894282633,
+      "grad_norm": 1.2274118661880493,
+      "learning_rate": 9.199859997428908e-05,
+      "loss": 0.7358171939849854,
+      "step": 3450
+    },
+    {
+      "epoch": 1.4893203883495145,
+      "grad_norm": 1.2810527086257935,
+      "learning_rate": 9.198527156884907e-05,
+      "loss": 0.6878830194473267,
+      "step": 3452
+    },
+    {
+      "epoch": 1.490183387270766,
+      "grad_norm": 1.2344684600830078,
+      "learning_rate": 9.197193303885008e-05,
+      "loss": 0.6995986104011536,
+      "step": 3454
+    },
+    {
+      "epoch": 1.4910463861920173,
+      "grad_norm": 1.233686923980713,
+      "learning_rate": 9.195858438750866e-05,
+      "loss": 0.6423236131668091,
+      "step": 3456
+    },
+    {
+      "epoch": 1.4919093851132685,
+      "grad_norm": 1.3485876321792603,
+      "learning_rate": 9.194522561804377e-05,
+      "loss": 0.7513291835784912,
+      "step": 3458
+    },
+    {
+      "epoch": 1.49277238403452,
+      "grad_norm": 0.9913640022277832,
+      "learning_rate": 9.19318567336768e-05,
+      "loss": 0.6667183637619019,
+      "step": 3460
+    },
+    {
+      "epoch": 1.4936353829557714,
+      "grad_norm": 1.146099328994751,
+      "learning_rate": 9.191847773763161e-05,
+      "loss": 0.643630862236023,
+      "step": 3462
+    },
+    {
+      "epoch": 1.4944983818770226,
+      "grad_norm": 1.0300167798995972,
+      "learning_rate": 9.190508863313448e-05,
+      "loss": 0.655176043510437,
+      "step": 3464
+    },
+    {
+      "epoch": 1.495361380798274,
+      "grad_norm": 1.285008430480957,
+      "learning_rate": 9.189168942341416e-05,
+      "loss": 0.6792252659797668,
+      "step": 3466
+    },
+    {
+      "epoch": 1.4962243797195254,
+      "grad_norm": 1.1455490589141846,
+      "learning_rate": 9.187828011170175e-05,
+      "loss": 0.7081018686294556,
+      "step": 3468
+    },
+    {
+      "epoch": 1.4970873786407766,
+      "grad_norm": 1.2560410499572754,
+      "learning_rate": 9.186486070123091e-05,
+      "loss": 0.6613248586654663,
+      "step": 3470
+    },
+    {
+      "epoch": 1.497950377562028,
+      "grad_norm": 1.2391352653503418,
+      "learning_rate": 9.185143119523761e-05,
+      "loss": 0.6299360990524292,
+      "step": 3472
+    },
+    {
+      "epoch": 1.4988133764832794,
+      "grad_norm": 1.134585976600647,
+      "learning_rate": 9.183799159696035e-05,
+      "loss": 0.5910941362380981,
+      "step": 3474
+    },
+    {
+      "epoch": 1.4996763754045308,
+      "grad_norm": 1.1744983196258545,
+      "learning_rate": 9.182454190964005e-05,
+      "loss": 0.7077166438102722,
+      "step": 3476
+    },
+    {
+      "epoch": 1.5005393743257822,
+      "grad_norm": 1.1142816543579102,
+      "learning_rate": 9.181108213651999e-05,
+      "loss": 0.6168769598007202,
+      "step": 3478
+    },
+    {
+      "epoch": 1.5014023732470334,
+      "grad_norm": 1.1463894844055176,
+      "learning_rate": 9.179761228084597e-05,
+      "loss": 0.6587874293327332,
+      "step": 3480
+    },
+    {
+      "epoch": 1.5022653721682848,
+      "grad_norm": 1.3945616483688354,
+      "learning_rate": 9.178413234586616e-05,
+      "loss": 0.7696266770362854,
+      "step": 3482
+    },
+    {
+      "epoch": 1.5031283710895362,
+      "grad_norm": 1.3266242742538452,
+      "learning_rate": 9.177064233483121e-05,
+      "loss": 0.7142559885978699,
+      "step": 3484
+    },
+    {
+      "epoch": 1.5039913700107874,
+      "grad_norm": 1.1857004165649414,
+      "learning_rate": 9.175714225099417e-05,
+      "loss": 0.7085118293762207,
+      "step": 3486
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 1.1070430278778076,
+      "learning_rate": 9.174363209761049e-05,
+      "loss": 0.6643039584159851,
+      "step": 3488
+    },
+    {
+      "epoch": 1.5057173678532902,
+      "grad_norm": 1.0877243280410767,
+      "learning_rate": 9.173011187793814e-05,
+      "loss": 0.6282537579536438,
+      "step": 3490
+    },
+    {
+      "epoch": 1.5065803667745414,
+      "grad_norm": 1.2648247480392456,
+      "learning_rate": 9.171658159523741e-05,
+      "loss": 0.6811800599098206,
+      "step": 3492
+    },
+    {
+      "epoch": 1.5074433656957928,
+      "grad_norm": 1.1535227298736572,
+      "learning_rate": 9.170304125277111e-05,
+      "loss": 0.6442922353744507,
+      "step": 3494
+    },
+    {
+      "epoch": 1.5083063646170443,
+      "grad_norm": 1.1175323724746704,
+      "learning_rate": 9.16894908538044e-05,
+      "loss": 0.717490553855896,
+      "step": 3496
+    },
+    {
+      "epoch": 1.5091693635382954,
+      "grad_norm": 1.1923470497131348,
+      "learning_rate": 9.167593040160491e-05,
+      "loss": 0.6537191867828369,
+      "step": 3498
+    },
+    {
+      "epoch": 1.510032362459547,
+      "grad_norm": 1.3302042484283447,
+      "learning_rate": 9.166235989944268e-05,
+      "loss": 0.7314199805259705,
+      "step": 3500
+    },
+    {
+      "epoch": 1.510032362459547,
+      "eval_loss": 0.7047909498214722,
+      "eval_runtime": 821.8811,
+      "eval_samples_per_second": 2.506,
+      "eval_steps_per_second": 2.506,
+      "step": 3500
+    },
+    {
+      "epoch": 1.5108953613807983,
+      "grad_norm": 1.1480292081832886,
+      "learning_rate": 9.164877935059017e-05,
+      "loss": 0.6363071203231812,
+      "step": 3502
+    },
+    {
+      "epoch": 1.5117583603020495,
+      "grad_norm": 1.1068446636199951,
+      "learning_rate": 9.163518875832227e-05,
+      "loss": 0.6020049452781677,
+      "step": 3504
+    },
+    {
+      "epoch": 1.512621359223301,
+      "grad_norm": 1.2131398916244507,
+      "learning_rate": 9.16215881259163e-05,
+      "loss": 0.7053841352462769,
+      "step": 3506
+    },
+    {
+      "epoch": 1.5134843581445523,
+      "grad_norm": 1.1875312328338623,
+      "learning_rate": 9.160797745665199e-05,
+      "loss": 0.6994143724441528,
+      "step": 3508
+    },
+    {
+      "epoch": 1.5143473570658037,
+      "grad_norm": 1.301967740058899,
+      "learning_rate": 9.159435675381147e-05,
+      "loss": 0.7001288533210754,
+      "step": 3510
+    },
+    {
+      "epoch": 1.5152103559870551,
+      "grad_norm": 1.1110293865203857,
+      "learning_rate": 9.158072602067933e-05,
+      "loss": 0.6847118139266968,
+      "step": 3512
+    },
+    {
+      "epoch": 1.5160733549083063,
+      "grad_norm": 1.2346487045288086,
+      "learning_rate": 9.156708526054256e-05,
+      "loss": 0.711972713470459,
+      "step": 3514
+    },
+    {
+      "epoch": 1.5169363538295577,
+      "grad_norm": 1.079206109046936,
+      "learning_rate": 9.155343447669058e-05,
+      "loss": 0.6406894326210022,
+      "step": 3516
+    },
+    {
+      "epoch": 1.5177993527508091,
+      "grad_norm": 1.0806539058685303,
+      "learning_rate": 9.153977367241519e-05,
+      "loss": 0.6568963527679443,
+      "step": 3518
+    },
+    {
+      "epoch": 1.5186623516720603,
+      "grad_norm": 1.244635820388794,
+      "learning_rate": 9.152610285101064e-05,
+      "loss": 0.6841240525245667,
+      "step": 3520
+    },
+    {
+      "epoch": 1.5195253505933117,
+      "grad_norm": 1.1596359014511108,
+      "learning_rate": 9.151242201577359e-05,
+      "loss": 0.619236946105957,
+      "step": 3522
+    },
+    {
+      "epoch": 1.5203883495145631,
+      "grad_norm": 1.2404531240463257,
+      "learning_rate": 9.14987311700031e-05,
+      "loss": 0.6059816479682922,
+      "step": 3524
+    },
+    {
+      "epoch": 1.5212513484358143,
+      "grad_norm": 1.0919517278671265,
+      "learning_rate": 9.148503031700067e-05,
+      "loss": 0.606432318687439,
+      "step": 3526
+    },
+    {
+      "epoch": 1.522114347357066,
+      "grad_norm": 1.113590121269226,
+      "learning_rate": 9.147131946007019e-05,
+      "loss": 0.6507619023323059,
+      "step": 3528
+    },
+    {
+      "epoch": 1.5229773462783172,
+      "grad_norm": 1.309885025024414,
+      "learning_rate": 9.145759860251799e-05,
+      "loss": 0.7071492075920105,
+      "step": 3530
+    },
+    {
+      "epoch": 1.5238403451995683,
+      "grad_norm": 1.1807539463043213,
+      "learning_rate": 9.144386774765278e-05,
+      "loss": 0.6297637224197388,
+      "step": 3532
+    },
+    {
+      "epoch": 1.52470334412082,
+      "grad_norm": 1.1587409973144531,
+      "learning_rate": 9.143012689878568e-05,
+      "loss": 0.6588859558105469,
+      "step": 3534
+    },
+    {
+      "epoch": 1.5255663430420712,
+      "grad_norm": 1.2401988506317139,
+      "learning_rate": 9.141637605923026e-05,
+      "loss": 0.6460821628570557,
+      "step": 3536
+    },
+    {
+      "epoch": 1.5264293419633226,
+      "grad_norm": 1.0708749294281006,
+      "learning_rate": 9.140261523230244e-05,
+      "loss": 0.6757200956344604,
+      "step": 3538
+    },
+    {
+      "epoch": 1.527292340884574,
+      "grad_norm": 1.2433289289474487,
+      "learning_rate": 9.138884442132061e-05,
+      "loss": 0.63701993227005,
+      "step": 3540
+    },
+    {
+      "epoch": 1.5281553398058252,
+      "grad_norm": 1.376205563545227,
+      "learning_rate": 9.137506362960553e-05,
+      "loss": 0.6708084344863892,
+      "step": 3542
+    },
+    {
+      "epoch": 1.5290183387270766,
+      "grad_norm": 1.1187406778335571,
+      "learning_rate": 9.136127286048038e-05,
+      "loss": 0.6108639240264893,
+      "step": 3544
+    },
+    {
+      "epoch": 1.529881337648328,
+      "grad_norm": 1.0509378910064697,
+      "learning_rate": 9.134747211727074e-05,
+      "loss": 0.58659827709198,
+      "step": 3546
+    },
+    {
+      "epoch": 1.5307443365695792,
+      "grad_norm": 1.1126946210861206,
+      "learning_rate": 9.133366140330458e-05,
+      "loss": 0.6567258238792419,
+      "step": 3548
+    },
+    {
+      "epoch": 1.5316073354908306,
+      "grad_norm": 1.235511302947998,
+      "learning_rate": 9.131984072191232e-05,
+      "loss": 0.7188405990600586,
+      "step": 3550
+    },
+    {
+      "epoch": 1.532470334412082,
+      "grad_norm": 1.1522574424743652,
+      "learning_rate": 9.130601007642674e-05,
+      "loss": 0.7415963411331177,
+      "step": 3552
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 1.1485437154769897,
+      "learning_rate": 9.129216947018302e-05,
+      "loss": 0.6286011934280396,
+      "step": 3554
+    },
+    {
+      "epoch": 1.5341963322545846,
+      "grad_norm": 1.1583795547485352,
+      "learning_rate": 9.127831890651879e-05,
+      "loss": 0.6919670701026917,
+      "step": 3556
+    },
+    {
+      "epoch": 1.535059331175836,
+      "grad_norm": 1.1450461149215698,
+      "learning_rate": 9.126445838877404e-05,
+      "loss": 0.6787450313568115,
+      "step": 3558
+    },
+    {
+      "epoch": 1.5359223300970872,
+      "grad_norm": 1.0954205989837646,
+      "learning_rate": 9.125058792029118e-05,
+      "loss": 0.6718940138816833,
+      "step": 3560
+    },
+    {
+      "epoch": 1.5367853290183389,
+      "grad_norm": 1.2366043329238892,
+      "learning_rate": 9.1236707504415e-05,
+      "loss": 0.7359552383422852,
+      "step": 3562
+    },
+    {
+      "epoch": 1.53764832793959,
+      "grad_norm": 1.1417096853256226,
+      "learning_rate": 9.122281714449268e-05,
+      "loss": 0.654550313949585,
+      "step": 3564
+    },
+    {
+      "epoch": 1.5385113268608415,
+      "grad_norm": 1.006664752960205,
+      "learning_rate": 9.120891684387386e-05,
+      "loss": 0.7129375338554382,
+      "step": 3566
+    },
+    {
+      "epoch": 1.5393743257820929,
+      "grad_norm": 1.2887771129608154,
+      "learning_rate": 9.11950066059105e-05,
+      "loss": 0.728861391544342,
+      "step": 3568
+    },
+    {
+      "epoch": 1.540237324703344,
+      "grad_norm": 1.194985270500183,
+      "learning_rate": 9.118108643395702e-05,
+      "loss": 0.6207047700881958,
+      "step": 3570
+    },
+    {
+      "epoch": 1.5411003236245955,
+      "grad_norm": 1.1913846731185913,
+      "learning_rate": 9.116715633137019e-05,
+      "loss": 0.6507744789123535,
+      "step": 3572
+    },
+    {
+      "epoch": 1.541963322545847,
+      "grad_norm": 1.1695647239685059,
+      "learning_rate": 9.115321630150919e-05,
+      "loss": 0.6736630201339722,
+      "step": 3574
+    },
+    {
+      "epoch": 1.542826321467098,
+      "grad_norm": 1.2866519689559937,
+      "learning_rate": 9.113926634773558e-05,
+      "loss": 0.6998549699783325,
+      "step": 3576
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 1.1755015850067139,
+      "learning_rate": 9.112530647341337e-05,
+      "loss": 0.6986453533172607,
+      "step": 3578
+    },
+    {
+      "epoch": 1.544552319309601,
+      "grad_norm": 1.2085564136505127,
+      "learning_rate": 9.111133668190889e-05,
+      "loss": 0.6959735155105591,
+      "step": 3580
+    },
+    {
+      "epoch": 1.545415318230852,
+      "grad_norm": 1.2213492393493652,
+      "learning_rate": 9.109735697659089e-05,
+      "loss": 0.6336050033569336,
+      "step": 3582
+    },
+    {
+      "epoch": 1.5462783171521035,
+      "grad_norm": 1.1609338521957397,
+      "learning_rate": 9.108336736083054e-05,
+      "loss": 0.652992844581604,
+      "step": 3584
+    },
+    {
+      "epoch": 1.547141316073355,
+      "grad_norm": 1.1483949422836304,
+      "learning_rate": 9.106936783800136e-05,
+      "loss": 0.5593333840370178,
+      "step": 3586
+    },
+    {
+      "epoch": 1.5480043149946061,
+      "grad_norm": 1.0852224826812744,
+      "learning_rate": 9.105535841147925e-05,
+      "loss": 0.6488125920295715,
+      "step": 3588
+    },
+    {
+      "epoch": 1.5488673139158577,
+      "grad_norm": 1.1804794073104858,
+      "learning_rate": 9.104133908464254e-05,
+      "loss": 0.5877764225006104,
+      "step": 3590
+    },
+    {
+      "epoch": 1.549730312837109,
+      "grad_norm": 1.0295292139053345,
+      "learning_rate": 9.102730986087194e-05,
+      "loss": 0.5924050211906433,
+      "step": 3592
+    },
+    {
+      "epoch": 1.5505933117583603,
+      "grad_norm": 1.524871587753296,
+      "learning_rate": 9.101327074355051e-05,
+      "loss": 0.6224870085716248,
+      "step": 3594
+    },
+    {
+      "epoch": 1.5514563106796118,
+      "grad_norm": 1.3207813501358032,
+      "learning_rate": 9.099922173606373e-05,
+      "loss": 0.5918852090835571,
+      "step": 3596
+    },
+    {
+      "epoch": 1.552319309600863,
+      "grad_norm": 1.2524348497390747,
+      "learning_rate": 9.098516284179945e-05,
+      "loss": 0.6756706237792969,
+      "step": 3598
+    },
+    {
+      "epoch": 1.5531823085221144,
+      "grad_norm": 1.2434089183807373,
+      "learning_rate": 9.097109406414792e-05,
+      "loss": 0.7416443824768066,
+      "step": 3600
+    },
+    {
+      "epoch": 1.5531823085221144,
+      "eval_loss": 0.701831042766571,
+      "eval_runtime": 836.0199,
+      "eval_samples_per_second": 2.464,
+      "eval_steps_per_second": 2.464,
+      "step": 3600
+    },
+    {
+      "epoch": 1.5540453074433658,
+      "grad_norm": 1.0528122186660767,
+      "learning_rate": 9.095701540650175e-05,
+      "loss": 0.5911263227462769,
+      "step": 3602
+    },
+    {
+      "epoch": 1.554908306364617,
+      "grad_norm": 1.1165716648101807,
+      "learning_rate": 9.094292687225595e-05,
+      "loss": 0.5925720930099487,
+      "step": 3604
+    },
+    {
+      "epoch": 1.5557713052858684,
+      "grad_norm": 1.1127078533172607,
+      "learning_rate": 9.09288284648079e-05,
+      "loss": 0.6038441061973572,
+      "step": 3606
+    },
+    {
+      "epoch": 1.5566343042071198,
+      "grad_norm": 1.195662498474121,
+      "learning_rate": 9.091472018755738e-05,
+      "loss": 0.6781054735183716,
+      "step": 3608
+    },
+    {
+      "epoch": 1.557497303128371,
+      "grad_norm": 1.1671538352966309,
+      "learning_rate": 9.090060204390652e-05,
+      "loss": 0.7171973586082458,
+      "step": 3610
+    },
+    {
+      "epoch": 1.5583603020496224,
+      "grad_norm": 1.2814654111862183,
+      "learning_rate": 9.088647403725987e-05,
+      "loss": 0.6465495228767395,
+      "step": 3612
+    },
+    {
+      "epoch": 1.5592233009708738,
+      "grad_norm": 1.2238248586654663,
+      "learning_rate": 9.087233617102433e-05,
+      "loss": 0.7105259299278259,
+      "step": 3614
+    },
+    {
+      "epoch": 1.560086299892125,
+      "grad_norm": 1.1868736743927002,
+      "learning_rate": 9.085818844860916e-05,
+      "loss": 0.6721329092979431,
+      "step": 3616
+    },
+    {
+      "epoch": 1.5609492988133766,
+      "grad_norm": 1.1358642578125,
+      "learning_rate": 9.084403087342606e-05,
+      "loss": 0.6684948801994324,
+      "step": 3618
+    },
+    {
+      "epoch": 1.5618122977346278,
+      "grad_norm": 1.1101531982421875,
+      "learning_rate": 9.082986344888901e-05,
+      "loss": 0.653505802154541,
+      "step": 3620
+    },
+    {
+      "epoch": 1.5626752966558792,
+      "grad_norm": 1.3343961238861084,
+      "learning_rate": 9.081568617841447e-05,
+      "loss": 0.673059344291687,
+      "step": 3622
+    },
+    {
+      "epoch": 1.5635382955771306,
+      "grad_norm": 1.0680615901947021,
+      "learning_rate": 9.080149906542119e-05,
+      "loss": 0.6429969668388367,
+      "step": 3624
+    },
+    {
+      "epoch": 1.5644012944983818,
+      "grad_norm": 1.0026768445968628,
+      "learning_rate": 9.078730211333038e-05,
+      "loss": 0.694593608379364,
+      "step": 3626
+    },
+    {
+      "epoch": 1.5652642934196332,
+      "grad_norm": 1.100269079208374,
+      "learning_rate": 9.077309532556551e-05,
+      "loss": 0.6931326985359192,
+      "step": 3628
+    },
+    {
+      "epoch": 1.5661272923408847,
+      "grad_norm": 1.1724740266799927,
+      "learning_rate": 9.07588787055525e-05,
+      "loss": 0.6182643175125122,
+      "step": 3630
+    },
+    {
+      "epoch": 1.5669902912621358,
+      "grad_norm": 1.0137040615081787,
+      "learning_rate": 9.074465225671965e-05,
+      "loss": 0.6307282447814941,
+      "step": 3632
+    },
+    {
+      "epoch": 1.5678532901833873,
+      "grad_norm": 1.0997183322906494,
+      "learning_rate": 9.073041598249756e-05,
+      "loss": 0.6049362421035767,
+      "step": 3634
+    },
+    {
+      "epoch": 1.5687162891046387,
+      "grad_norm": 1.1688816547393799,
+      "learning_rate": 9.07161698863193e-05,
+      "loss": 0.6583153009414673,
+      "step": 3636
+    },
+    {
+      "epoch": 1.5695792880258899,
+      "grad_norm": 1.2558770179748535,
+      "learning_rate": 9.07019139716202e-05,
+      "loss": 0.6346786022186279,
+      "step": 3638
+    },
+    {
+      "epoch": 1.5704422869471413,
+      "grad_norm": 1.037838101387024,
+      "learning_rate": 9.068764824183804e-05,
+      "loss": 0.6214386224746704,
+      "step": 3640
+    },
+    {
+      "epoch": 1.5713052858683927,
+      "grad_norm": 1.2514159679412842,
+      "learning_rate": 9.067337270041293e-05,
+      "loss": 0.6745668649673462,
+      "step": 3642
+    },
+    {
+      "epoch": 1.5721682847896439,
+      "grad_norm": 1.2718034982681274,
+      "learning_rate": 9.065908735078736e-05,
+      "loss": 0.6409499049186707,
+      "step": 3644
+    },
+    {
+      "epoch": 1.5730312837108955,
+      "grad_norm": 1.1754566431045532,
+      "learning_rate": 9.064479219640615e-05,
+      "loss": 0.5885956883430481,
+      "step": 3646
+    },
+    {
+      "epoch": 1.5738942826321467,
+      "grad_norm": 1.3022228479385376,
+      "learning_rate": 9.063048724071653e-05,
+      "loss": 0.6472923159599304,
+      "step": 3648
+    },
+    {
+      "epoch": 1.574757281553398,
+      "grad_norm": 1.2492727041244507,
+      "learning_rate": 9.061617248716808e-05,
+      "loss": 0.6977139711380005,
+      "step": 3650
+    },
+    {
+      "epoch": 1.5756202804746495,
+      "grad_norm": 1.0432982444763184,
+      "learning_rate": 9.060184793921274e-05,
+      "loss": 0.6403328776359558,
+      "step": 3652
+    },
+    {
+      "epoch": 1.5764832793959007,
+      "grad_norm": 1.0898420810699463,
+      "learning_rate": 9.05875136003048e-05,
+      "loss": 0.6726073026657104,
+      "step": 3654
+    },
+    {
+      "epoch": 1.5773462783171521,
+      "grad_norm": 1.1885254383087158,
+      "learning_rate": 9.057316947390091e-05,
+      "loss": 0.6861021518707275,
+      "step": 3656
+    },
+    {
+      "epoch": 1.5782092772384035,
+      "grad_norm": 1.0795915126800537,
+      "learning_rate": 9.055881556346012e-05,
+      "loss": 0.6786487102508545,
+      "step": 3658
+    },
+    {
+      "epoch": 1.5790722761596547,
+      "grad_norm": 1.0702277421951294,
+      "learning_rate": 9.054445187244379e-05,
+      "loss": 0.6757274866104126,
+      "step": 3660
+    },
+    {
+      "epoch": 1.5799352750809061,
+      "grad_norm": 0.9469023942947388,
+      "learning_rate": 9.053007840431566e-05,
+      "loss": 0.6596087217330933,
+      "step": 3662
+    },
+    {
+      "epoch": 1.5807982740021576,
+      "grad_norm": 0.979669988155365,
+      "learning_rate": 9.051569516254185e-05,
+      "loss": 0.6254066228866577,
+      "step": 3664
+    },
+    {
+      "epoch": 1.5816612729234087,
+      "grad_norm": 1.294577717781067,
+      "learning_rate": 9.050130215059081e-05,
+      "loss": 0.6463974118232727,
+      "step": 3666
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 1.1566381454467773,
+      "learning_rate": 9.04868993719333e-05,
+      "loss": 0.6496619582176208,
+      "step": 3668
+    },
+    {
+      "epoch": 1.5833872707659116,
+      "grad_norm": 1.0207090377807617,
+      "learning_rate": 9.047248683004254e-05,
+      "loss": 0.5987569093704224,
+      "step": 3670
+    },
+    {
+      "epoch": 1.5842502696871628,
+      "grad_norm": 1.1533408164978027,
+      "learning_rate": 9.045806452839404e-05,
+      "loss": 0.6392318606376648,
+      "step": 3672
+    },
+    {
+      "epoch": 1.5851132686084144,
+      "grad_norm": 1.0433615446090698,
+      "learning_rate": 9.044363247046565e-05,
+      "loss": 0.6197013854980469,
+      "step": 3674
+    },
+    {
+      "epoch": 1.5859762675296656,
+      "grad_norm": 1.2047847509384155,
+      "learning_rate": 9.042919065973763e-05,
+      "loss": 0.620261549949646,
+      "step": 3676
+    },
+    {
+      "epoch": 1.5868392664509168,
+      "grad_norm": 1.232506513595581,
+      "learning_rate": 9.041473909969252e-05,
+      "loss": 0.6751506924629211,
+      "step": 3678
+    },
+    {
+      "epoch": 1.5877022653721684,
+      "grad_norm": 1.1659306287765503,
+      "learning_rate": 9.040027779381528e-05,
+      "loss": 0.6619004011154175,
+      "step": 3680
+    },
+    {
+      "epoch": 1.5885652642934196,
+      "grad_norm": 1.0183755159378052,
+      "learning_rate": 9.038580674559316e-05,
+      "loss": 0.6120786666870117,
+      "step": 3682
+    },
+    {
+      "epoch": 1.589428263214671,
+      "grad_norm": 1.194318175315857,
+      "learning_rate": 9.037132595851581e-05,
+      "loss": 0.616247296333313,
+      "step": 3684
+    },
+    {
+      "epoch": 1.5902912621359224,
+      "grad_norm": 1.500475525856018,
+      "learning_rate": 9.035683543607522e-05,
+      "loss": 0.7229626774787903,
+      "step": 3686
+    },
+    {
+      "epoch": 1.5911542610571736,
+      "grad_norm": 1.047954797744751,
+      "learning_rate": 9.034233518176567e-05,
+      "loss": 0.6193928718566895,
+      "step": 3688
+    },
+    {
+      "epoch": 1.592017259978425,
+      "grad_norm": 1.3852946758270264,
+      "learning_rate": 9.032782519908386e-05,
+      "loss": 0.7276740074157715,
+      "step": 3690
+    },
+    {
+      "epoch": 1.5928802588996764,
+      "grad_norm": 1.2648470401763916,
+      "learning_rate": 9.031330549152882e-05,
+      "loss": 0.6635852456092834,
+      "step": 3692
+    },
+    {
+      "epoch": 1.5937432578209276,
+      "grad_norm": 1.1195251941680908,
+      "learning_rate": 9.029877606260186e-05,
+      "loss": 0.6905328631401062,
+      "step": 3694
+    },
+    {
+      "epoch": 1.594606256742179,
+      "grad_norm": 1.0823792219161987,
+      "learning_rate": 9.028423691580675e-05,
+      "loss": 0.6474795341491699,
+      "step": 3696
+    },
+    {
+      "epoch": 1.5954692556634305,
+      "grad_norm": 1.1743394136428833,
+      "learning_rate": 9.026968805464948e-05,
+      "loss": 0.6740748882293701,
+      "step": 3698
+    },
+    {
+      "epoch": 1.5963322545846816,
+      "grad_norm": 1.19667387008667,
+      "learning_rate": 9.02551294826385e-05,
+      "loss": 0.5662298798561096,
+      "step": 3700
+    },
+    {
+      "epoch": 1.5963322545846816,
+      "eval_loss": 0.6999035477638245,
+      "eval_runtime": 836.4944,
+      "eval_samples_per_second": 2.463,
+      "eval_steps_per_second": 2.463,
+      "step": 3700
+    },
+    {
+      "epoch": 1.5971952535059333,
+      "grad_norm": 0.9947987794876099,
+      "learning_rate": 9.02405612032845e-05,
+      "loss": 0.6141433119773865,
+      "step": 3702
+    },
+    {
+      "epoch": 1.5980582524271845,
+      "grad_norm": 1.0307772159576416,
+      "learning_rate": 9.022598322010059e-05,
+      "loss": 0.6011156439781189,
+      "step": 3704
+    },
+    {
+      "epoch": 1.5989212513484357,
+      "grad_norm": 1.3516873121261597,
+      "learning_rate": 9.021139553660214e-05,
+      "loss": 0.7671760320663452,
+      "step": 3706
+    },
+    {
+      "epoch": 1.5997842502696873,
+      "grad_norm": 1.0319154262542725,
+      "learning_rate": 9.019679815630695e-05,
+      "loss": 0.5630212426185608,
+      "step": 3708
+    },
+    {
+      "epoch": 1.6006472491909385,
+      "grad_norm": 1.34807550907135,
+      "learning_rate": 9.01821910827351e-05,
+      "loss": 0.7143195271492004,
+      "step": 3710
+    },
+    {
+      "epoch": 1.60151024811219,
+      "grad_norm": 1.0618840456008911,
+      "learning_rate": 9.016757431940898e-05,
+      "loss": 0.6624904274940491,
+      "step": 3712
+    },
+    {
+      "epoch": 1.6023732470334413,
+      "grad_norm": 0.987445056438446,
+      "learning_rate": 9.015294786985342e-05,
+      "loss": 0.631399929523468,
+      "step": 3714
+    },
+    {
+      "epoch": 1.6032362459546925,
+      "grad_norm": 1.0872961282730103,
+      "learning_rate": 9.013831173759548e-05,
+      "loss": 0.646812379360199,
+      "step": 3716
+    },
+    {
+      "epoch": 1.604099244875944,
+      "grad_norm": 0.9798976182937622,
+      "learning_rate": 9.012366592616461e-05,
+      "loss": 0.6353676915168762,
+      "step": 3718
+    },
+    {
+      "epoch": 1.6049622437971953,
+      "grad_norm": 1.3835123777389526,
+      "learning_rate": 9.010901043909258e-05,
+      "loss": 0.5802496671676636,
+      "step": 3720
+    },
+    {
+      "epoch": 1.6058252427184465,
+      "grad_norm": 1.0870599746704102,
+      "learning_rate": 9.009434527991349e-05,
+      "loss": 0.6140236854553223,
+      "step": 3722
+    },
+    {
+      "epoch": 1.606688241639698,
+      "grad_norm": 1.0582176446914673,
+      "learning_rate": 9.007967045216379e-05,
+      "loss": 0.6562411785125732,
+      "step": 3724
+    },
+    {
+      "epoch": 1.6075512405609493,
+      "grad_norm": 1.0486522912979126,
+      "learning_rate": 9.006498595938224e-05,
+      "loss": 0.5995528697967529,
+      "step": 3726
+    },
+    {
+      "epoch": 1.6084142394822005,
+      "grad_norm": 1.1152576208114624,
+      "learning_rate": 9.005029180510994e-05,
+      "loss": 0.6296200752258301,
+      "step": 3728
+    },
+    {
+      "epoch": 1.6092772384034522,
+      "grad_norm": 1.2114241123199463,
+      "learning_rate": 9.003558799289031e-05,
+      "loss": 0.7050777077674866,
+      "step": 3730
+    },
+    {
+      "epoch": 1.6101402373247033,
+      "grad_norm": 1.1112302541732788,
+      "learning_rate": 9.002087452626913e-05,
+      "loss": 0.6323828101158142,
+      "step": 3732
+    },
+    {
+      "epoch": 1.6110032362459545,
+      "grad_norm": 1.102839469909668,
+      "learning_rate": 9.000615140879446e-05,
+      "loss": 0.7021811008453369,
+      "step": 3734
+    },
+    {
+      "epoch": 1.6118662351672062,
+      "grad_norm": 0.8860642910003662,
+      "learning_rate": 8.999141864401672e-05,
+      "loss": 0.5094663500785828,
+      "step": 3736
+    },
+    {
+      "epoch": 1.6127292340884574,
+      "grad_norm": 1.1873432397842407,
+      "learning_rate": 8.997667623548869e-05,
+      "loss": 0.6481591463088989,
+      "step": 3738
+    },
+    {
+      "epoch": 1.6135922330097088,
+      "grad_norm": 0.9258447885513306,
+      "learning_rate": 8.996192418676537e-05,
+      "loss": 0.6846200227737427,
+      "step": 3740
+    },
+    {
+      "epoch": 1.6144552319309602,
+      "grad_norm": 1.1108704805374146,
+      "learning_rate": 8.994716250140419e-05,
+      "loss": 0.7081074118614197,
+      "step": 3742
+    },
+    {
+      "epoch": 1.6153182308522114,
+      "grad_norm": 1.01946222782135,
+      "learning_rate": 8.993239118296486e-05,
+      "loss": 0.683017373085022,
+      "step": 3744
+    },
+    {
+      "epoch": 1.6161812297734628,
+      "grad_norm": 1.0794068574905396,
+      "learning_rate": 8.991761023500942e-05,
+      "loss": 0.6625820994377136,
+      "step": 3746
+    },
+    {
+      "epoch": 1.6170442286947142,
+      "grad_norm": 1.0697723627090454,
+      "learning_rate": 8.990281966110221e-05,
+      "loss": 0.6714970469474792,
+      "step": 3748
+    },
+    {
+      "epoch": 1.6179072276159654,
+      "grad_norm": 1.118213176727295,
+      "learning_rate": 8.988801946480991e-05,
+      "loss": 0.6702249646186829,
+      "step": 3750
+    },
+    {
+      "epoch": 1.6187702265372168,
+      "grad_norm": 1.3541207313537598,
+      "learning_rate": 8.987320964970156e-05,
+      "loss": 0.6783145070075989,
+      "step": 3752
+    },
+    {
+      "epoch": 1.6196332254584682,
+      "grad_norm": 1.2015430927276611,
+      "learning_rate": 8.985839021934843e-05,
+      "loss": 0.6774528622627258,
+      "step": 3754
+    },
+    {
+      "epoch": 1.6204962243797194,
+      "grad_norm": 1.2218681573867798,
+      "learning_rate": 8.984356117732419e-05,
+      "loss": 0.6122720241546631,
+      "step": 3756
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 1.0959049463272095,
+      "learning_rate": 8.982872252720478e-05,
+      "loss": 0.6658767461776733,
+      "step": 3758
+    },
+    {
+      "epoch": 1.6222222222222222,
+      "grad_norm": 1.0399792194366455,
+      "learning_rate": 8.981387427256849e-05,
+      "loss": 0.6943061947822571,
+      "step": 3760
+    },
+    {
+      "epoch": 1.6230852211434734,
+      "grad_norm": 1.1107885837554932,
+      "learning_rate": 8.979901641699591e-05,
+      "loss": 0.6100694537162781,
+      "step": 3762
+    },
+    {
+      "epoch": 1.623948220064725,
+      "grad_norm": 1.2478004693984985,
+      "learning_rate": 8.978414896406993e-05,
+      "loss": 0.7148857712745667,
+      "step": 3764
+    },
+    {
+      "epoch": 1.6248112189859762,
+      "grad_norm": 1.333235502243042,
+      "learning_rate": 8.976927191737577e-05,
+      "loss": 0.6626213788986206,
+      "step": 3766
+    },
+    {
+      "epoch": 1.6256742179072277,
+      "grad_norm": 1.2001140117645264,
+      "learning_rate": 8.975438528050095e-05,
+      "loss": 0.6073925495147705,
+      "step": 3768
+    },
+    {
+      "epoch": 1.626537216828479,
+      "grad_norm": 0.8740705847740173,
+      "learning_rate": 8.973948905703535e-05,
+      "loss": 0.5861713290214539,
+      "step": 3770
+    },
+    {
+      "epoch": 1.6274002157497303,
+      "grad_norm": 1.0697330236434937,
+      "learning_rate": 8.972458325057111e-05,
+      "loss": 0.7104416489601135,
+      "step": 3772
+    },
+    {
+      "epoch": 1.6282632146709817,
+      "grad_norm": 1.2699353694915771,
+      "learning_rate": 8.970966786470271e-05,
+      "loss": 0.652777373790741,
+      "step": 3774
+    },
+    {
+      "epoch": 1.629126213592233,
+      "grad_norm": 1.2045143842697144,
+      "learning_rate": 8.969474290302691e-05,
+      "loss": 0.6303225159645081,
+      "step": 3776
+    },
+    {
+      "epoch": 1.6299892125134843,
+      "grad_norm": 1.2175726890563965,
+      "learning_rate": 8.96798083691428e-05,
+      "loss": 0.594299852848053,
+      "step": 3778
+    },
+    {
+      "epoch": 1.6308522114347357,
+      "grad_norm": 0.9873425960540771,
+      "learning_rate": 8.966486426665178e-05,
+      "loss": 0.566745400428772,
+      "step": 3780
+    },
+    {
+      "epoch": 1.631715210355987,
+      "grad_norm": 1.4304969310760498,
+      "learning_rate": 8.964991059915756e-05,
+      "loss": 0.76564621925354,
+      "step": 3782
+    },
+    {
+      "epoch": 1.6325782092772383,
+      "grad_norm": 1.1225755214691162,
+      "learning_rate": 8.963494737026611e-05,
+      "loss": 0.7026722431182861,
+      "step": 3784
+    },
+    {
+      "epoch": 1.6334412081984897,
+      "grad_norm": 1.251551628112793,
+      "learning_rate": 8.961997458358581e-05,
+      "loss": 0.6778496503829956,
+      "step": 3786
+    },
+    {
+      "epoch": 1.6343042071197411,
+      "grad_norm": 0.9573827981948853,
+      "learning_rate": 8.960499224272725e-05,
+      "loss": 0.5866316556930542,
+      "step": 3788
+    },
+    {
+      "epoch": 1.6351672060409923,
+      "grad_norm": 1.0466986894607544,
+      "learning_rate": 8.959000035130334e-05,
+      "loss": 0.6244695782661438,
+      "step": 3790
+    },
+    {
+      "epoch": 1.636030204962244,
+      "grad_norm": 1.0724520683288574,
+      "learning_rate": 8.957499891292931e-05,
+      "loss": 0.6380400061607361,
+      "step": 3792
+    },
+    {
+      "epoch": 1.6368932038834951,
+      "grad_norm": 1.1590946912765503,
+      "learning_rate": 8.955998793122271e-05,
+      "loss": 0.625150203704834,
+      "step": 3794
+    },
+    {
+      "epoch": 1.6377562028047465,
+      "grad_norm": 1.3405407667160034,
+      "learning_rate": 8.954496740980334e-05,
+      "loss": 0.7554174065589905,
+      "step": 3796
+    },
+    {
+      "epoch": 1.638619201725998,
+      "grad_norm": 1.2695956230163574,
+      "learning_rate": 8.952993735229338e-05,
+      "loss": 0.6710763573646545,
+      "step": 3798
+    },
+    {
+      "epoch": 1.6394822006472491,
+      "grad_norm": 1.4330781698226929,
+      "learning_rate": 8.95148977623172e-05,
+      "loss": 0.7102031111717224,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6394822006472491,
+      "eval_loss": 0.6974969506263733,
+      "eval_runtime": 823.7017,
+      "eval_samples_per_second": 2.501,
+      "eval_steps_per_second": 2.501,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6403451995685006,
+      "grad_norm": 0.9818692803382874,
+      "learning_rate": 8.949984864350157e-05,
+      "loss": 0.6028781533241272,
+      "step": 3802
+    },
+    {
+      "epoch": 1.641208198489752,
+      "grad_norm": 1.0779424905776978,
+      "learning_rate": 8.94847899994755e-05,
+      "loss": 0.649263322353363,
+      "step": 3804
+    },
+    {
+      "epoch": 1.6420711974110032,
+      "grad_norm": 1.2096810340881348,
+      "learning_rate": 8.946972183387033e-05,
+      "loss": 0.6424242258071899,
+      "step": 3806
+    },
+    {
+      "epoch": 1.6429341963322546,
+      "grad_norm": 1.1697514057159424,
+      "learning_rate": 8.945464415031966e-05,
+      "loss": 0.6751719117164612,
+      "step": 3808
+    },
+    {
+      "epoch": 1.643797195253506,
+      "grad_norm": 1.6524088382720947,
+      "learning_rate": 8.943955695245942e-05,
+      "loss": 0.616973340511322,
+      "step": 3810
+    },
+    {
+      "epoch": 1.6446601941747572,
+      "grad_norm": 1.1154218912124634,
+      "learning_rate": 8.942446024392782e-05,
+      "loss": 0.6577548980712891,
+      "step": 3812
+    },
+    {
+      "epoch": 1.6455231930960086,
+      "grad_norm": 1.1747835874557495,
+      "learning_rate": 8.940935402836535e-05,
+      "loss": 0.6534224152565002,
+      "step": 3814
+    },
+    {
+      "epoch": 1.64638619201726,
+      "grad_norm": 1.189562439918518,
+      "learning_rate": 8.939423830941482e-05,
+      "loss": 0.6356330513954163,
+      "step": 3816
+    },
+    {
+      "epoch": 1.6472491909385112,
+      "grad_norm": 1.0842949151992798,
+      "learning_rate": 8.937911309072132e-05,
+      "loss": 0.6097415685653687,
+      "step": 3818
+    },
+    {
+      "epoch": 1.6481121898597628,
+      "grad_norm": 1.1931639909744263,
+      "learning_rate": 8.936397837593221e-05,
+      "loss": 0.6339239478111267,
+      "step": 3820
+    },
+    {
+      "epoch": 1.648975188781014,
+      "grad_norm": 1.4980956315994263,
+      "learning_rate": 8.934883416869717e-05,
+      "loss": 0.6588500142097473,
+      "step": 3822
+    },
+    {
+      "epoch": 1.6498381877022654,
+      "grad_norm": 1.2506407499313354,
+      "learning_rate": 8.933368047266814e-05,
+      "loss": 0.6418336033821106,
+      "step": 3824
+    },
+    {
+      "epoch": 1.6507011866235168,
+      "grad_norm": 1.3003007173538208,
+      "learning_rate": 8.93185172914994e-05,
+      "loss": 0.6475774049758911,
+      "step": 3826
+    },
+    {
+      "epoch": 1.651564185544768,
+      "grad_norm": 1.2024277448654175,
+      "learning_rate": 8.930334462884747e-05,
+      "loss": 0.7352058291435242,
+      "step": 3828
+    },
+    {
+      "epoch": 1.6524271844660194,
+      "grad_norm": 1.0713937282562256,
+      "learning_rate": 8.928816248837115e-05,
+      "loss": 0.6267798542976379,
+      "step": 3830
+    },
+    {
+      "epoch": 1.6532901833872709,
+      "grad_norm": 1.3538615703582764,
+      "learning_rate": 8.927297087373157e-05,
+      "loss": 0.7097233533859253,
+      "step": 3832
+    },
+    {
+      "epoch": 1.654153182308522,
+      "grad_norm": 1.1811957359313965,
+      "learning_rate": 8.92577697885921e-05,
+      "loss": 0.6012749671936035,
+      "step": 3834
+    },
+    {
+      "epoch": 1.6550161812297735,
+      "grad_norm": 1.064659595489502,
+      "learning_rate": 8.924255923661841e-05,
+      "loss": 0.6257606744766235,
+      "step": 3836
+    },
+    {
+      "epoch": 1.6558791801510249,
+      "grad_norm": 1.1258097887039185,
+      "learning_rate": 8.922733922147848e-05,
+      "loss": 0.7108950018882751,
+      "step": 3838
+    },
+    {
+      "epoch": 1.656742179072276,
+      "grad_norm": 1.4515002965927124,
+      "learning_rate": 8.921210974684253e-05,
+      "loss": 0.6432359218597412,
+      "step": 3840
+    },
+    {
+      "epoch": 1.6576051779935275,
+      "grad_norm": 1.040401816368103,
+      "learning_rate": 8.919687081638308e-05,
+      "loss": 0.5962746739387512,
+      "step": 3842
+    },
+    {
+      "epoch": 1.6584681769147789,
+      "grad_norm": 1.0090585947036743,
+      "learning_rate": 8.918162243377493e-05,
+      "loss": 0.6396839022636414,
+      "step": 3844
+    },
+    {
+      "epoch": 1.65933117583603,
+      "grad_norm": 1.2226985692977905,
+      "learning_rate": 8.916636460269517e-05,
+      "loss": 0.7058185935020447,
+      "step": 3846
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 1.117598295211792,
+      "learning_rate": 8.915109732682316e-05,
+      "loss": 0.6138123273849487,
+      "step": 3848
+    },
+    {
+      "epoch": 1.661057173678533,
+      "grad_norm": 1.1237187385559082,
+      "learning_rate": 8.913582060984052e-05,
+      "loss": 0.6389812231063843,
+      "step": 3850
+    },
+    {
+      "epoch": 1.661920172599784,
+      "grad_norm": 1.148144245147705,
+      "learning_rate": 8.912053445543114e-05,
+      "loss": 0.7189341187477112,
+      "step": 3852
+    },
+    {
+      "epoch": 1.6627831715210357,
+      "grad_norm": 1.3115259408950806,
+      "learning_rate": 8.910523886728126e-05,
+      "loss": 0.671591579914093,
+      "step": 3854
+    },
+    {
+      "epoch": 1.663646170442287,
+      "grad_norm": 1.0031025409698486,
+      "learning_rate": 8.908993384907928e-05,
+      "loss": 0.6695008873939514,
+      "step": 3856
+    },
+    {
+      "epoch": 1.6645091693635383,
+      "grad_norm": 1.2426995038986206,
+      "learning_rate": 8.907461940451601e-05,
+      "loss": 0.6557230353355408,
+      "step": 3858
+    },
+    {
+      "epoch": 1.6653721682847897,
+      "grad_norm": 1.1471003293991089,
+      "learning_rate": 8.90592955372844e-05,
+      "loss": 0.6126149296760559,
+      "step": 3860
+    },
+    {
+      "epoch": 1.666235167206041,
+      "grad_norm": 1.2763721942901611,
+      "learning_rate": 8.904396225107976e-05,
+      "loss": 0.742831826210022,
+      "step": 3862
+    },
+    {
+      "epoch": 1.6670981661272923,
+      "grad_norm": 1.0612274408340454,
+      "learning_rate": 8.902861954959963e-05,
+      "loss": 0.6506689190864563,
+      "step": 3864
+    },
+    {
+      "epoch": 1.6679611650485437,
+      "grad_norm": 1.160914421081543,
+      "learning_rate": 8.901326743654384e-05,
+      "loss": 0.6939412355422974,
+      "step": 3866
+    },
+    {
+      "epoch": 1.668824163969795,
+      "grad_norm": 1.2517154216766357,
+      "learning_rate": 8.899790591561447e-05,
+      "loss": 0.6331133842468262,
+      "step": 3868
+    },
+    {
+      "epoch": 1.6696871628910464,
+      "grad_norm": 0.9491620063781738,
+      "learning_rate": 8.89825349905159e-05,
+      "loss": 0.6425275206565857,
+      "step": 3870
+    },
+    {
+      "epoch": 1.6705501618122978,
+      "grad_norm": 0.9615546464920044,
+      "learning_rate": 8.896715466495476e-05,
+      "loss": 0.5937016606330872,
+      "step": 3872
+    },
+    {
+      "epoch": 1.671413160733549,
+      "grad_norm": 1.1985108852386475,
+      "learning_rate": 8.895176494263992e-05,
+      "loss": 0.7523217797279358,
+      "step": 3874
+    },
+    {
+      "epoch": 1.6722761596548006,
+      "grad_norm": 1.1150988340377808,
+      "learning_rate": 8.893636582728257e-05,
+      "loss": 0.6035639047622681,
+      "step": 3876
+    },
+    {
+      "epoch": 1.6731391585760518,
+      "grad_norm": 0.9545241594314575,
+      "learning_rate": 8.892095732259612e-05,
+      "loss": 0.6141317486763,
+      "step": 3878
+    },
+    {
+      "epoch": 1.674002157497303,
+      "grad_norm": 1.2937493324279785,
+      "learning_rate": 8.890553943229627e-05,
+      "loss": 0.6015914678573608,
+      "step": 3880
+    },
+    {
+      "epoch": 1.6748651564185546,
+      "grad_norm": 1.0606268644332886,
+      "learning_rate": 8.889011216010099e-05,
+      "loss": 0.6010603904724121,
+      "step": 3882
+    },
+    {
+      "epoch": 1.6757281553398058,
+      "grad_norm": 1.0271397829055786,
+      "learning_rate": 8.887467550973046e-05,
+      "loss": 0.6256402134895325,
+      "step": 3884
+    },
+    {
+      "epoch": 1.6765911542610572,
+      "grad_norm": 1.096360683441162,
+      "learning_rate": 8.885922948490719e-05,
+      "loss": 0.6119387745857239,
+      "step": 3886
+    },
+    {
+      "epoch": 1.6774541531823086,
+      "grad_norm": 1.1860841512680054,
+      "learning_rate": 8.884377408935589e-05,
+      "loss": 0.6526403427124023,
+      "step": 3888
+    },
+    {
+      "epoch": 1.6783171521035598,
+      "grad_norm": 1.1262441873550415,
+      "learning_rate": 8.882830932680359e-05,
+      "loss": 0.5966535806655884,
+      "step": 3890
+    },
+    {
+      "epoch": 1.6791801510248112,
+      "grad_norm": 1.0326828956604004,
+      "learning_rate": 8.881283520097953e-05,
+      "loss": 0.6420881748199463,
+      "step": 3892
+    },
+    {
+      "epoch": 1.6800431499460626,
+      "grad_norm": 1.1928532123565674,
+      "learning_rate": 8.879735171561525e-05,
+      "loss": 0.6054600477218628,
+      "step": 3894
+    },
+    {
+      "epoch": 1.6809061488673138,
+      "grad_norm": 1.2887117862701416,
+      "learning_rate": 8.878185887444448e-05,
+      "loss": 0.7106166481971741,
+      "step": 3896
+    },
+    {
+      "epoch": 1.6817691477885652,
+      "grad_norm": 1.1483163833618164,
+      "learning_rate": 8.876635668120327e-05,
+      "loss": 0.7001441717147827,
+      "step": 3898
+    },
+    {
+      "epoch": 1.6826321467098166,
+      "grad_norm": 1.2908320426940918,
+      "learning_rate": 8.875084513962993e-05,
+      "loss": 0.6607111692428589,
+      "step": 3900
+    },
+    {
+      "epoch": 1.6826321467098166,
+      "eval_loss": 0.6949095726013184,
+      "eval_runtime": 817.4497,
+      "eval_samples_per_second": 2.52,
+      "eval_steps_per_second": 2.52,
+      "step": 3900
+    },
+    {
+      "epoch": 1.6834951456310678,
+      "grad_norm": 1.00212824344635,
+      "learning_rate": 8.873532425346496e-05,
+      "loss": 0.6463292837142944,
+      "step": 3902
+    },
+    {
+      "epoch": 1.6843581445523195,
+      "grad_norm": 1.2223109006881714,
+      "learning_rate": 8.871979402645116e-05,
+      "loss": 0.6627839207649231,
+      "step": 3904
+    },
+    {
+      "epoch": 1.6852211434735707,
+      "grad_norm": 1.047956109046936,
+      "learning_rate": 8.870425446233359e-05,
+      "loss": 0.6311076879501343,
+      "step": 3906
+    },
+    {
+      "epoch": 1.6860841423948218,
+      "grad_norm": 1.178468108177185,
+      "learning_rate": 8.868870556485954e-05,
+      "loss": 0.6916288137435913,
+      "step": 3908
+    },
+    {
+      "epoch": 1.6869471413160735,
+      "grad_norm": 6.569681644439697,
+      "learning_rate": 8.867314733777854e-05,
+      "loss": 0.696479856967926,
+      "step": 3910
+    },
+    {
+      "epoch": 1.6878101402373247,
+      "grad_norm": 1.100183129310608,
+      "learning_rate": 8.865757978484241e-05,
+      "loss": 0.6245654225349426,
+      "step": 3912
+    },
+    {
+      "epoch": 1.688673139158576,
+      "grad_norm": 1.222562313079834,
+      "learning_rate": 8.864200290980518e-05,
+      "loss": 0.6178512573242188,
+      "step": 3914
+    },
+    {
+      "epoch": 1.6895361380798275,
+      "grad_norm": 1.235471487045288,
+      "learning_rate": 8.862641671642314e-05,
+      "loss": 0.6598679423332214,
+      "step": 3916
+    },
+    {
+      "epoch": 1.6903991370010787,
+      "grad_norm": 1.2539454698562622,
+      "learning_rate": 8.861082120845486e-05,
+      "loss": 0.6503698825836182,
+      "step": 3918
+    },
+    {
+      "epoch": 1.69126213592233,
+      "grad_norm": 1.1766268014907837,
+      "learning_rate": 8.859521638966107e-05,
+      "loss": 0.6858329176902771,
+      "step": 3920
+    },
+    {
+      "epoch": 1.6921251348435815,
+      "grad_norm": 1.3463754653930664,
+      "learning_rate": 8.857960226380486e-05,
+      "loss": 0.679460883140564,
+      "step": 3922
+    },
+    {
+      "epoch": 1.6929881337648327,
+      "grad_norm": 1.3011246919631958,
+      "learning_rate": 8.856397883465149e-05,
+      "loss": 0.6901042461395264,
+      "step": 3924
+    },
+    {
+      "epoch": 1.6938511326860841,
+      "grad_norm": 0.9615590572357178,
+      "learning_rate": 8.854834610596845e-05,
+      "loss": 0.6756444573402405,
+      "step": 3926
+    },
+    {
+      "epoch": 1.6947141316073355,
+      "grad_norm": 0.9635165333747864,
+      "learning_rate": 8.853270408152552e-05,
+      "loss": 0.6690155863761902,
+      "step": 3928
+    },
+    {
+      "epoch": 1.6955771305285867,
+      "grad_norm": 1.2924212217330933,
+      "learning_rate": 8.851705276509472e-05,
+      "loss": 0.6412895321846008,
+      "step": 3930
+    },
+    {
+      "epoch": 1.6964401294498384,
+      "grad_norm": 1.0281823873519897,
+      "learning_rate": 8.850139216045028e-05,
+      "loss": 0.6250708699226379,
+      "step": 3932
+    },
+    {
+      "epoch": 1.6973031283710895,
+      "grad_norm": 1.321605920791626,
+      "learning_rate": 8.848572227136868e-05,
+      "loss": 0.6545571684837341,
+      "step": 3934
+    },
+    {
+      "epoch": 1.6981661272923407,
+      "grad_norm": 1.1515092849731445,
+      "learning_rate": 8.847004310162866e-05,
+      "loss": 0.6466772556304932,
+      "step": 3936
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 1.1318165063858032,
+      "learning_rate": 8.845435465501116e-05,
+      "loss": 0.679786741733551,
+      "step": 3938
+    },
+    {
+      "epoch": 1.6998921251348436,
+      "grad_norm": 0.973469078540802,
+      "learning_rate": 8.84386569352994e-05,
+      "loss": 0.6546611189842224,
+      "step": 3940
+    },
+    {
+      "epoch": 1.700755124056095,
+      "grad_norm": 1.0992401838302612,
+      "learning_rate": 8.842294994627879e-05,
+      "loss": 0.6207357048988342,
+      "step": 3942
+    },
+    {
+      "epoch": 1.7016181229773464,
+      "grad_norm": 1.1114588975906372,
+      "learning_rate": 8.840723369173702e-05,
+      "loss": 0.6113827228546143,
+      "step": 3944
+    },
+    {
+      "epoch": 1.7024811218985976,
+      "grad_norm": 1.171692132949829,
+      "learning_rate": 8.839150817546398e-05,
+      "loss": 0.6246283650398254,
+      "step": 3946
+    },
+    {
+      "epoch": 1.703344120819849,
+      "grad_norm": 1.249851942062378,
+      "learning_rate": 8.837577340125181e-05,
+      "loss": 0.7027832269668579,
+      "step": 3948
+    },
+    {
+      "epoch": 1.7042071197411004,
+      "grad_norm": 1.2475496530532837,
+      "learning_rate": 8.836002937289489e-05,
+      "loss": 0.7121281027793884,
+      "step": 3950
+    },
+    {
+      "epoch": 1.7050701186623516,
+      "grad_norm": 1.048726201057434,
+      "learning_rate": 8.834427609418981e-05,
+      "loss": 0.6521060466766357,
+      "step": 3952
+    },
+    {
+      "epoch": 1.705933117583603,
+      "grad_norm": 1.1689599752426147,
+      "learning_rate": 8.832851356893541e-05,
+      "loss": 0.6735731959342957,
+      "step": 3954
+    },
+    {
+      "epoch": 1.7067961165048544,
+      "grad_norm": 1.4230015277862549,
+      "learning_rate": 8.831274180093275e-05,
+      "loss": 0.6530178189277649,
+      "step": 3956
+    },
+    {
+      "epoch": 1.7076591154261056,
+      "grad_norm": 1.2237123250961304,
+      "learning_rate": 8.829696079398511e-05,
+      "loss": 0.6858299970626831,
+      "step": 3958
+    },
+    {
+      "epoch": 1.708522114347357,
+      "grad_norm": 1.1330037117004395,
+      "learning_rate": 8.828117055189803e-05,
+      "loss": 0.6567219495773315,
+      "step": 3960
+    },
+    {
+      "epoch": 1.7093851132686084,
+      "grad_norm": 1.2441636323928833,
+      "learning_rate": 8.826537107847925e-05,
+      "loss": 0.6207150816917419,
+      "step": 3962
+    },
+    {
+      "epoch": 1.7102481121898596,
+      "grad_norm": 0.9952784776687622,
+      "learning_rate": 8.824956237753872e-05,
+      "loss": 0.6192324161529541,
+      "step": 3964
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 1.0648247003555298,
+      "learning_rate": 8.823374445288865e-05,
+      "loss": 0.6485670208930969,
+      "step": 3966
+    },
+    {
+      "epoch": 1.7119741100323624,
+      "grad_norm": 1.2040860652923584,
+      "learning_rate": 8.821791730834347e-05,
+      "loss": 0.6591274738311768,
+      "step": 3968
+    },
+    {
+      "epoch": 1.7128371089536139,
+      "grad_norm": 1.2346793413162231,
+      "learning_rate": 8.820208094771982e-05,
+      "loss": 0.6427604556083679,
+      "step": 3970
+    },
+    {
+      "epoch": 1.7137001078748653,
+      "grad_norm": 1.0716660022735596,
+      "learning_rate": 8.818623537483658e-05,
+      "loss": 0.5952412486076355,
+      "step": 3972
+    },
+    {
+      "epoch": 1.7145631067961165,
+      "grad_norm": 1.1809104681015015,
+      "learning_rate": 8.817038059351481e-05,
+      "loss": 0.6808739900588989,
+      "step": 3974
+    },
+    {
+      "epoch": 1.7154261057173679,
+      "grad_norm": 1.224974513053894,
+      "learning_rate": 8.815451660757783e-05,
+      "loss": 0.6927714347839355,
+      "step": 3976
+    },
+    {
+      "epoch": 1.7162891046386193,
+      "grad_norm": 1.141288161277771,
+      "learning_rate": 8.813864342085117e-05,
+      "loss": 0.6450885534286499,
+      "step": 3978
+    },
+    {
+      "epoch": 1.7171521035598705,
+      "grad_norm": 1.0749722719192505,
+      "learning_rate": 8.812276103716259e-05,
+      "loss": 0.6775641441345215,
+      "step": 3980
+    },
+    {
+      "epoch": 1.7180151024811219,
+      "grad_norm": 1.13978111743927,
+      "learning_rate": 8.810686946034203e-05,
+      "loss": 0.6083413362503052,
+      "step": 3982
+    },
+    {
+      "epoch": 1.7188781014023733,
+      "grad_norm": 1.3158389329910278,
+      "learning_rate": 8.80909686942217e-05,
+      "loss": 0.6711311936378479,
+      "step": 3984
+    },
+    {
+      "epoch": 1.7197411003236245,
+      "grad_norm": 1.0368914604187012,
+      "learning_rate": 8.807505874263597e-05,
+      "loss": 0.6818822026252747,
+      "step": 3986
+    },
+    {
+      "epoch": 1.720604099244876,
+      "grad_norm": 1.249948263168335,
+      "learning_rate": 8.805913960942148e-05,
+      "loss": 0.6346645355224609,
+      "step": 3988
+    },
+    {
+      "epoch": 1.7214670981661273,
+      "grad_norm": 1.0894579887390137,
+      "learning_rate": 8.804321129841703e-05,
+      "loss": 0.6550438404083252,
+      "step": 3990
+    },
+    {
+      "epoch": 1.7223300970873785,
+      "grad_norm": 1.2956905364990234,
+      "learning_rate": 8.80272738134637e-05,
+      "loss": 0.6257092952728271,
+      "step": 3992
+    },
+    {
+      "epoch": 1.7231930960086301,
+      "grad_norm": 1.0089573860168457,
+      "learning_rate": 8.80113271584047e-05,
+      "loss": 0.5877772569656372,
+      "step": 3994
+    },
+    {
+      "epoch": 1.7240560949298813,
+      "grad_norm": 1.0698174238204956,
+      "learning_rate": 8.799537133708552e-05,
+      "loss": 0.6440973281860352,
+      "step": 3996
+    },
+    {
+      "epoch": 1.7249190938511327,
+      "grad_norm": 1.1145644187927246,
+      "learning_rate": 8.797940635335382e-05,
+      "loss": 0.6270571351051331,
+      "step": 3998
+    },
+    {
+      "epoch": 1.7257820927723841,
+      "grad_norm": 1.0673385858535767,
+      "learning_rate": 8.796343221105949e-05,
+      "loss": 0.6463025212287903,
+      "step": 4000
+    },
+    {
+      "epoch": 1.7257820927723841,
+      "eval_loss": 0.6941319704055786,
+      "eval_runtime": 498.1423,
+      "eval_samples_per_second": 4.135,
+      "eval_steps_per_second": 4.135,
+      "step": 4000
+    },
+    {
+      "epoch": 1.7266450916936353,
+      "grad_norm": 1.1285523176193237,
+      "learning_rate": 8.794744891405461e-05,
+      "loss": 0.6483243107795715,
+      "step": 4002
+    },
+    {
+      "epoch": 1.7275080906148867,
+      "grad_norm": 1.0747114419937134,
+      "learning_rate": 8.793145646619352e-05,
+      "loss": 0.6334620714187622,
+      "step": 4004
+    },
+    {
+      "epoch": 1.7283710895361382,
+      "grad_norm": 1.0277562141418457,
+      "learning_rate": 8.791545487133269e-05,
+      "loss": 0.6604806780815125,
+      "step": 4006
+    },
+    {
+      "epoch": 1.7292340884573894,
+      "grad_norm": 1.0783002376556396,
+      "learning_rate": 8.789944413333083e-05,
+      "loss": 0.6033409237861633,
+      "step": 4008
+    },
+    {
+      "epoch": 1.7300970873786408,
+      "grad_norm": 1.2508432865142822,
+      "learning_rate": 8.788342425604887e-05,
+      "loss": 0.6360779404640198,
+      "step": 4010
+    },
+    {
+      "epoch": 1.7309600862998922,
+      "grad_norm": 1.0905060768127441,
+      "learning_rate": 8.786739524334993e-05,
+      "loss": 0.5815097689628601,
+      "step": 4012
+    },
+    {
+      "epoch": 1.7318230852211434,
+      "grad_norm": 1.1219831705093384,
+      "learning_rate": 8.785135709909935e-05,
+      "loss": 0.7228885889053345,
+      "step": 4014
+    },
+    {
+      "epoch": 1.7326860841423948,
+      "grad_norm": 1.2444658279418945,
+      "learning_rate": 8.783530982716463e-05,
+      "loss": 0.6539638638496399,
+      "step": 4016
+    },
+    {
+      "epoch": 1.7335490830636462,
+      "grad_norm": 1.158237338066101,
+      "learning_rate": 8.781925343141551e-05,
+      "loss": 0.6482524275779724,
+      "step": 4018
+    },
+    {
+      "epoch": 1.7344120819848974,
+      "grad_norm": 1.288036584854126,
+      "learning_rate": 8.78031879157239e-05,
+      "loss": 0.699881374835968,
+      "step": 4020
+    },
+    {
+      "epoch": 1.735275080906149,
+      "grad_norm": 1.0680478811264038,
+      "learning_rate": 8.778711328396396e-05,
+      "loss": 0.6012337803840637,
+      "step": 4022
+    },
+    {
+      "epoch": 1.7361380798274002,
+      "grad_norm": 1.3618063926696777,
+      "learning_rate": 8.7771029540012e-05,
+      "loss": 0.7336176037788391,
+      "step": 4024
+    },
+    {
+      "epoch": 1.7370010787486516,
+      "grad_norm": 1.1720085144042969,
+      "learning_rate": 8.775493668774652e-05,
+      "loss": 0.6370863914489746,
+      "step": 4026
+    },
+    {
+      "epoch": 1.737864077669903,
+      "grad_norm": 1.1800546646118164,
+      "learning_rate": 8.773883473104827e-05,
+      "loss": 0.6755601763725281,
+      "step": 4028
+    },
+    {
+      "epoch": 1.7387270765911542,
+      "grad_norm": 1.1262768507003784,
+      "learning_rate": 8.772272367380015e-05,
+      "loss": 0.6209036707878113,
+      "step": 4030
+    },
+    {
+      "epoch": 1.7395900755124056,
+      "grad_norm": 1.2891435623168945,
+      "learning_rate": 8.770660351988727e-05,
+      "loss": 0.6727437973022461,
+      "step": 4032
+    },
+    {
+      "epoch": 1.740453074433657,
+      "grad_norm": 1.1410391330718994,
+      "learning_rate": 8.769047427319694e-05,
+      "loss": 0.6565355658531189,
+      "step": 4034
+    },
+    {
+      "epoch": 1.7413160733549082,
+      "grad_norm": 1.075622797012329,
+      "learning_rate": 8.767433593761864e-05,
+      "loss": 0.7123345136642456,
+      "step": 4036
+    },
+    {
+      "epoch": 1.7421790722761596,
+      "grad_norm": 1.1938068866729736,
+      "learning_rate": 8.765818851704408e-05,
+      "loss": 0.6391130685806274,
+      "step": 4038
+    },
+    {
+      "epoch": 1.743042071197411,
+      "grad_norm": 1.2302309274673462,
+      "learning_rate": 8.76420320153671e-05,
+      "loss": 0.7138327956199646,
+      "step": 4040
+    },
+    {
+      "epoch": 1.7439050701186622,
+      "grad_norm": 1.3204374313354492,
+      "learning_rate": 8.76258664364838e-05,
+      "loss": 0.6010884046554565,
+      "step": 4042
+    },
+    {
+      "epoch": 1.7447680690399137,
+      "grad_norm": 1.2324718236923218,
+      "learning_rate": 8.760969178429244e-05,
+      "loss": 0.6665774583816528,
+      "step": 4044
+    },
+    {
+      "epoch": 1.745631067961165,
+      "grad_norm": 1.0836756229400635,
+      "learning_rate": 8.759350806269342e-05,
+      "loss": 0.644280731678009,
+      "step": 4046
+    },
+    {
+      "epoch": 1.7464940668824163,
+      "grad_norm": 1.00897216796875,
+      "learning_rate": 8.757731527558943e-05,
+      "loss": 0.5871413946151733,
+      "step": 4048
+    },
+    {
+      "epoch": 1.747357065803668,
+      "grad_norm": 0.9774143695831299,
+      "learning_rate": 8.756111342688527e-05,
+      "loss": 0.5935707688331604,
+      "step": 4050
+    },
+    {
+      "epoch": 1.748220064724919,
+      "grad_norm": 1.0634222030639648,
+      "learning_rate": 8.75449025204879e-05,
+      "loss": 0.6011680960655212,
+      "step": 4052
+    },
+    {
+      "epoch": 1.7490830636461703,
+      "grad_norm": 1.2054167985916138,
+      "learning_rate": 8.752868256030658e-05,
+      "loss": 0.6434821486473083,
+      "step": 4054
+    },
+    {
+      "epoch": 1.749946062567422,
+      "grad_norm": 1.2121766805648804,
+      "learning_rate": 8.751245355025261e-05,
+      "loss": 0.6644792556762695,
+      "step": 4056
+    },
+    {
+      "epoch": 1.750809061488673,
+      "grad_norm": 1.186061143875122,
+      "learning_rate": 8.749621549423959e-05,
+      "loss": 0.6489604711532593,
+      "step": 4058
+    },
+    {
+      "epoch": 1.7516720604099245,
+      "grad_norm": 1.074569821357727,
+      "learning_rate": 8.747996839618322e-05,
+      "loss": 0.6018257737159729,
+      "step": 4060
+    },
+    {
+      "epoch": 1.752535059331176,
+      "grad_norm": 1.0770072937011719,
+      "learning_rate": 8.746371226000145e-05,
+      "loss": 0.6242427825927734,
+      "step": 4062
+    },
+    {
+      "epoch": 1.7533980582524271,
+      "grad_norm": 1.070820927619934,
+      "learning_rate": 8.744744708961435e-05,
+      "loss": 0.6011215448379517,
+      "step": 4064
+    },
+    {
+      "epoch": 1.7542610571736785,
+      "grad_norm": 1.163550853729248,
+      "learning_rate": 8.743117288894419e-05,
+      "loss": 0.6660047769546509,
+      "step": 4066
+    },
+    {
+      "epoch": 1.75512405609493,
+      "grad_norm": 1.2277700901031494,
+      "learning_rate": 8.741488966191544e-05,
+      "loss": 0.6795720458030701,
+      "step": 4068
+    },
+    {
+      "epoch": 1.7559870550161811,
+      "grad_norm": 1.0723415613174438,
+      "learning_rate": 8.73985974124547e-05,
+      "loss": 0.635612428188324,
+      "step": 4070
+    },
+    {
+      "epoch": 1.7568500539374325,
+      "grad_norm": 1.1380949020385742,
+      "learning_rate": 8.73822961444908e-05,
+      "loss": 0.5903881788253784,
+      "step": 4072
+    },
+    {
+      "epoch": 1.757713052858684,
+      "grad_norm": 1.1686267852783203,
+      "learning_rate": 8.736598586195471e-05,
+      "loss": 0.6036412119865417,
+      "step": 4074
+    },
+    {
+      "epoch": 1.7585760517799351,
+      "grad_norm": 1.1719619035720825,
+      "learning_rate": 8.734966656877956e-05,
+      "loss": 0.6732221245765686,
+      "step": 4076
+    },
+    {
+      "epoch": 1.7594390507011868,
+      "grad_norm": 1.0898852348327637,
+      "learning_rate": 8.733333826890071e-05,
+      "loss": 0.6788899898529053,
+      "step": 4078
+    },
+    {
+      "epoch": 1.760302049622438,
+      "grad_norm": 1.405272126197815,
+      "learning_rate": 8.731700096625563e-05,
+      "loss": 0.7186161875724792,
+      "step": 4080
+    },
+    {
+      "epoch": 1.7611650485436892,
+      "grad_norm": 1.287775993347168,
+      "learning_rate": 8.7300654664784e-05,
+      "loss": 0.7305195331573486,
+      "step": 4082
+    },
+    {
+      "epoch": 1.7620280474649408,
+      "grad_norm": 1.2288155555725098,
+      "learning_rate": 8.728429936842762e-05,
+      "loss": 0.6772475838661194,
+      "step": 4084
+    },
+    {
+      "epoch": 1.762891046386192,
+      "grad_norm": 1.2059677839279175,
+      "learning_rate": 8.726793508113054e-05,
+      "loss": 0.6882839202880859,
+      "step": 4086
+    },
+    {
+      "epoch": 1.7637540453074434,
+      "grad_norm": 1.3448551893234253,
+      "learning_rate": 8.725156180683893e-05,
+      "loss": 0.72773277759552,
+      "step": 4088
+    },
+    {
+      "epoch": 1.7646170442286948,
+      "grad_norm": 1.088852882385254,
+      "learning_rate": 8.723517954950111e-05,
+      "loss": 0.6122981309890747,
+      "step": 4090
+    },
+    {
+      "epoch": 1.765480043149946,
+      "grad_norm": 1.0686509609222412,
+      "learning_rate": 8.72187883130676e-05,
+      "loss": 0.6185593605041504,
+      "step": 4092
+    },
+    {
+      "epoch": 1.7663430420711974,
+      "grad_norm": 1.3372819423675537,
+      "learning_rate": 8.720238810149109e-05,
+      "loss": 0.7149755358695984,
+      "step": 4094
+    },
+    {
+      "epoch": 1.7672060409924488,
+      "grad_norm": 1.2540380954742432,
+      "learning_rate": 8.718597891872638e-05,
+      "loss": 0.7330698370933533,
+      "step": 4096
+    },
+    {
+      "epoch": 1.7680690399137,
+      "grad_norm": 1.1653515100479126,
+      "learning_rate": 8.716956076873049e-05,
+      "loss": 0.6567472815513611,
+      "step": 4098
+    },
+    {
+      "epoch": 1.7689320388349514,
+      "grad_norm": 1.0333259105682373,
+      "learning_rate": 8.715313365546258e-05,
+      "loss": 0.612649142742157,
+      "step": 4100
+    },
+    {
+      "epoch": 1.7689320388349514,
+      "eval_loss": 0.6929343938827515,
+      "eval_runtime": 497.805,
+      "eval_samples_per_second": 4.138,
+      "eval_steps_per_second": 4.138,
+      "step": 4100
+    },
+    {
+      "epoch": 1.7697950377562028,
+      "grad_norm": 1.1154836416244507,
+      "learning_rate": 8.713669758288398e-05,
+      "loss": 0.6242305636405945,
+      "step": 4102
+    },
+    {
+      "epoch": 1.770658036677454,
+      "grad_norm": 1.2573596239089966,
+      "learning_rate": 8.712025255495817e-05,
+      "loss": 0.6430796384811401,
+      "step": 4104
+    },
+    {
+      "epoch": 1.7715210355987057,
+      "grad_norm": 1.407214641571045,
+      "learning_rate": 8.71037985756508e-05,
+      "loss": 0.6321606040000916,
+      "step": 4106
+    },
+    {
+      "epoch": 1.7723840345199569,
+      "grad_norm": 1.1721560955047607,
+      "learning_rate": 8.708733564892969e-05,
+      "loss": 0.6277352571487427,
+      "step": 4108
+    },
+    {
+      "epoch": 1.773247033441208,
+      "grad_norm": 1.0967402458190918,
+      "learning_rate": 8.707086377876475e-05,
+      "loss": 0.6275392770767212,
+      "step": 4110
+    },
+    {
+      "epoch": 1.7741100323624597,
+      "grad_norm": 1.0293948650360107,
+      "learning_rate": 8.705438296912813e-05,
+      "loss": 0.5985271334648132,
+      "step": 4112
+    },
+    {
+      "epoch": 1.7749730312837109,
+      "grad_norm": 1.1681201457977295,
+      "learning_rate": 8.703789322399409e-05,
+      "loss": 0.6175792217254639,
+      "step": 4114
+    },
+    {
+      "epoch": 1.7758360302049623,
+      "grad_norm": 1.3726415634155273,
+      "learning_rate": 8.702139454733907e-05,
+      "loss": 0.6355584859848022,
+      "step": 4116
+    },
+    {
+      "epoch": 1.7766990291262137,
+      "grad_norm": 1.1292396783828735,
+      "learning_rate": 8.700488694314166e-05,
+      "loss": 0.5877590179443359,
+      "step": 4118
+    },
+    {
+      "epoch": 1.7775620280474649,
+      "grad_norm": 1.086949348449707,
+      "learning_rate": 8.698837041538259e-05,
+      "loss": 0.6119363307952881,
+      "step": 4120
+    },
+    {
+      "epoch": 1.7784250269687163,
+      "grad_norm": 1.0434467792510986,
+      "learning_rate": 8.697184496804472e-05,
+      "loss": 0.7157787084579468,
+      "step": 4122
+    },
+    {
+      "epoch": 1.7792880258899677,
+      "grad_norm": 1.0321751832962036,
+      "learning_rate": 8.69553106051131e-05,
+      "loss": 0.6830587983131409,
+      "step": 4124
+    },
+    {
+      "epoch": 1.780151024811219,
+      "grad_norm": 1.3480322360992432,
+      "learning_rate": 8.693876733057493e-05,
+      "loss": 0.7015431523323059,
+      "step": 4126
+    },
+    {
+      "epoch": 1.7810140237324703,
+      "grad_norm": 0.9731265306472778,
+      "learning_rate": 8.692221514841957e-05,
+      "loss": 0.5946303606033325,
+      "step": 4128
+    },
+    {
+      "epoch": 1.7818770226537217,
+      "grad_norm": 1.1613185405731201,
+      "learning_rate": 8.690565406263844e-05,
+      "loss": 0.687828779220581,
+      "step": 4130
+    },
+    {
+      "epoch": 1.782740021574973,
+      "grad_norm": 1.159293532371521,
+      "learning_rate": 8.68890840772252e-05,
+      "loss": 0.6315785646438599,
+      "step": 4132
+    },
+    {
+      "epoch": 1.7836030204962245,
+      "grad_norm": 1.2591674327850342,
+      "learning_rate": 8.687250519617564e-05,
+      "loss": 0.6761205196380615,
+      "step": 4134
+    },
+    {
+      "epoch": 1.7844660194174757,
+      "grad_norm": 1.082155466079712,
+      "learning_rate": 8.685591742348769e-05,
+      "loss": 0.6714316010475159,
+      "step": 4136
+    },
+    {
+      "epoch": 1.785329018338727,
+      "grad_norm": 1.3653912544250488,
+      "learning_rate": 8.68393207631614e-05,
+      "loss": 0.7500089406967163,
+      "step": 4138
+    },
+    {
+      "epoch": 1.7861920172599786,
+      "grad_norm": 1.124690055847168,
+      "learning_rate": 8.682271521919896e-05,
+      "loss": 0.6818236708641052,
+      "step": 4140
+    },
+    {
+      "epoch": 1.7870550161812297,
+      "grad_norm": 1.0788345336914062,
+      "learning_rate": 8.680610079560474e-05,
+      "loss": 0.5876757502555847,
+      "step": 4142
+    },
+    {
+      "epoch": 1.7879180151024812,
+      "grad_norm": 1.0923465490341187,
+      "learning_rate": 8.678947749638526e-05,
+      "loss": 0.7281823754310608,
+      "step": 4144
+    },
+    {
+      "epoch": 1.7887810140237326,
+      "grad_norm": 1.0655275583267212,
+      "learning_rate": 8.67728453255491e-05,
+      "loss": 0.70050048828125,
+      "step": 4146
+    },
+    {
+      "epoch": 1.7896440129449838,
+      "grad_norm": 1.2138395309448242,
+      "learning_rate": 8.67562042871071e-05,
+      "loss": 0.6276566386222839,
+      "step": 4148
+    },
+    {
+      "epoch": 1.7905070118662352,
+      "grad_norm": 1.1036893129348755,
+      "learning_rate": 8.673955438507211e-05,
+      "loss": 0.5959358811378479,
+      "step": 4150
+    },
+    {
+      "epoch": 1.7913700107874866,
+      "grad_norm": 1.0354493856430054,
+      "learning_rate": 8.67228956234592e-05,
+      "loss": 0.6574703454971313,
+      "step": 4152
+    },
+    {
+      "epoch": 1.7922330097087378,
+      "grad_norm": 0.985424280166626,
+      "learning_rate": 8.670622800628556e-05,
+      "loss": 0.6122542023658752,
+      "step": 4154
+    },
+    {
+      "epoch": 1.7930960086299892,
+      "grad_norm": 1.1953589916229248,
+      "learning_rate": 8.668955153757051e-05,
+      "loss": 0.6271086931228638,
+      "step": 4156
+    },
+    {
+      "epoch": 1.7939590075512406,
+      "grad_norm": 1.1208388805389404,
+      "learning_rate": 8.66728662213355e-05,
+      "loss": 0.6667966246604919,
+      "step": 4158
+    },
+    {
+      "epoch": 1.7948220064724918,
+      "grad_norm": 1.1992945671081543,
+      "learning_rate": 8.665617206160415e-05,
+      "loss": 0.6663333177566528,
+      "step": 4160
+    },
+    {
+      "epoch": 1.7956850053937432,
+      "grad_norm": 1.2541922330856323,
+      "learning_rate": 8.663946906240213e-05,
+      "loss": 0.7308130264282227,
+      "step": 4162
+    },
+    {
+      "epoch": 1.7965480043149946,
+      "grad_norm": 1.1797630786895752,
+      "learning_rate": 8.662275722775733e-05,
+      "loss": 0.5975028276443481,
+      "step": 4164
+    },
+    {
+      "epoch": 1.7974110032362458,
+      "grad_norm": 1.0160104036331177,
+      "learning_rate": 8.66060365616997e-05,
+      "loss": 0.6519654393196106,
+      "step": 4166
+    },
+    {
+      "epoch": 1.7982740021574974,
+      "grad_norm": 0.9538162350654602,
+      "learning_rate": 8.658930706826138e-05,
+      "loss": 0.5663926601409912,
+      "step": 4168
+    },
+    {
+      "epoch": 1.7991370010787486,
+      "grad_norm": 1.2381815910339355,
+      "learning_rate": 8.657256875147662e-05,
+      "loss": 0.6580140590667725,
+      "step": 4170
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1187360286712646,
+      "learning_rate": 8.655582161538177e-05,
+      "loss": 0.5924336314201355,
+      "step": 4172
+    },
+    {
+      "epoch": 1.8008629989212515,
+      "grad_norm": 1.1784483194351196,
+      "learning_rate": 8.653906566401533e-05,
+      "loss": 0.6451319456100464,
+      "step": 4174
+    },
+    {
+      "epoch": 1.8017259978425026,
+      "grad_norm": 1.1566145420074463,
+      "learning_rate": 8.652230090141791e-05,
+      "loss": 0.6331294775009155,
+      "step": 4176
+    },
+    {
+      "epoch": 1.802588996763754,
+      "grad_norm": 1.2551170587539673,
+      "learning_rate": 8.65055273316323e-05,
+      "loss": 0.6497499346733093,
+      "step": 4178
+    },
+    {
+      "epoch": 1.8034519956850055,
+      "grad_norm": 1.0211492776870728,
+      "learning_rate": 8.648874495870332e-05,
+      "loss": 0.6330156326293945,
+      "step": 4180
+    },
+    {
+      "epoch": 1.8043149946062567,
+      "grad_norm": 1.0939637422561646,
+      "learning_rate": 8.6471953786678e-05,
+      "loss": 0.6429944634437561,
+      "step": 4182
+    },
+    {
+      "epoch": 1.805177993527508,
+      "grad_norm": 1.2167190313339233,
+      "learning_rate": 8.645515381960543e-05,
+      "loss": 0.7175302505493164,
+      "step": 4184
+    },
+    {
+      "epoch": 1.8060409924487595,
+      "grad_norm": 1.2631546258926392,
+      "learning_rate": 8.643834506153685e-05,
+      "loss": 0.6182355284690857,
+      "step": 4186
+    },
+    {
+      "epoch": 1.8069039913700107,
+      "grad_norm": 1.2085437774658203,
+      "learning_rate": 8.642152751652564e-05,
+      "loss": 0.6001572012901306,
+      "step": 4188
+    },
+    {
+      "epoch": 1.807766990291262,
+      "grad_norm": 1.1557323932647705,
+      "learning_rate": 8.640470118862726e-05,
+      "loss": 0.6273662447929382,
+      "step": 4190
+    },
+    {
+      "epoch": 1.8086299892125135,
+      "grad_norm": 1.0053116083145142,
+      "learning_rate": 8.638786608189928e-05,
+      "loss": 0.5653483867645264,
+      "step": 4192
+    },
+    {
+      "epoch": 1.8094929881337647,
+      "grad_norm": 1.135331630706787,
+      "learning_rate": 8.637102220040145e-05,
+      "loss": 0.7450010776519775,
+      "step": 4194
+    },
+    {
+      "epoch": 1.8103559870550163,
+      "grad_norm": 1.1989952325820923,
+      "learning_rate": 8.635416954819557e-05,
+      "loss": 0.6311311721801758,
+      "step": 4196
+    },
+    {
+      "epoch": 1.8112189859762675,
+      "grad_norm": 1.3958007097244263,
+      "learning_rate": 8.633730812934559e-05,
+      "loss": 0.6941008567810059,
+      "step": 4198
+    },
+    {
+      "epoch": 1.812081984897519,
+      "grad_norm": 1.135804533958435,
+      "learning_rate": 8.632043794791758e-05,
+      "loss": 0.729922890663147,
+      "step": 4200
+    },
+    {
+      "epoch": 1.812081984897519,
+      "eval_loss": 0.6918642520904541,
+      "eval_runtime": 498.4981,
+      "eval_samples_per_second": 4.132,
+      "eval_steps_per_second": 4.132,
+      "step": 4200
+    },
+    {
+      "epoch": 1.8129449838187703,
+      "grad_norm": 1.082674503326416,
+      "learning_rate": 8.630355900797967e-05,
+      "loss": 0.5869379043579102,
+      "step": 4202
+    },
+    {
+      "epoch": 1.8138079827400215,
+      "grad_norm": 1.290963888168335,
+      "learning_rate": 8.628667131360218e-05,
+      "loss": 0.6634620428085327,
+      "step": 4204
+    },
+    {
+      "epoch": 1.814670981661273,
+      "grad_norm": 1.1635067462921143,
+      "learning_rate": 8.626977486885747e-05,
+      "loss": 0.6378920674324036,
+      "step": 4206
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 1.5235686302185059,
+      "learning_rate": 8.625286967782006e-05,
+      "loss": 0.6387537121772766,
+      "step": 4208
+    },
+    {
+      "epoch": 1.8163969795037755,
+      "grad_norm": 1.1585710048675537,
+      "learning_rate": 8.623595574456655e-05,
+      "loss": 0.7544008493423462,
+      "step": 4210
+    },
+    {
+      "epoch": 1.817259978425027,
+      "grad_norm": 1.0860902070999146,
+      "learning_rate": 8.621903307317566e-05,
+      "loss": 0.6456180810928345,
+      "step": 4212
+    },
+    {
+      "epoch": 1.8181229773462784,
+      "grad_norm": 1.0690677165985107,
+      "learning_rate": 8.620210166772824e-05,
+      "loss": 0.6784940361976624,
+      "step": 4214
+    },
+    {
+      "epoch": 1.8189859762675296,
+      "grad_norm": 1.3315730094909668,
+      "learning_rate": 8.61851615323072e-05,
+      "loss": 0.725097119808197,
+      "step": 4216
+    },
+    {
+      "epoch": 1.819848975188781,
+      "grad_norm": 1.3689441680908203,
+      "learning_rate": 8.616821267099757e-05,
+      "loss": 0.6593776345252991,
+      "step": 4218
+    },
+    {
+      "epoch": 1.8207119741100324,
+      "grad_norm": 1.0977379083633423,
+      "learning_rate": 8.61512550878865e-05,
+      "loss": 0.6334517598152161,
+      "step": 4220
+    },
+    {
+      "epoch": 1.8215749730312836,
+      "grad_norm": 1.2982810735702515,
+      "learning_rate": 8.613428878706326e-05,
+      "loss": 0.7029228210449219,
+      "step": 4222
+    },
+    {
+      "epoch": 1.8224379719525352,
+      "grad_norm": 1.117529273033142,
+      "learning_rate": 8.611731377261916e-05,
+      "loss": 0.6502435207366943,
+      "step": 4224
+    },
+    {
+      "epoch": 1.8233009708737864,
+      "grad_norm": 1.0791085958480835,
+      "learning_rate": 8.610033004864765e-05,
+      "loss": 0.6504095196723938,
+      "step": 4226
+    },
+    {
+      "epoch": 1.8241639697950378,
+      "grad_norm": 1.017614483833313,
+      "learning_rate": 8.60833376192443e-05,
+      "loss": 0.6614294052124023,
+      "step": 4228
+    },
+    {
+      "epoch": 1.8250269687162892,
+      "grad_norm": 1.0131317377090454,
+      "learning_rate": 8.606633648850673e-05,
+      "loss": 0.6149688959121704,
+      "step": 4230
+    },
+    {
+      "epoch": 1.8258899676375404,
+      "grad_norm": 1.0489189624786377,
+      "learning_rate": 8.604932666053474e-05,
+      "loss": 0.5616279244422913,
+      "step": 4232
+    },
+    {
+      "epoch": 1.8267529665587918,
+      "grad_norm": 1.210129976272583,
+      "learning_rate": 8.60323081394301e-05,
+      "loss": 0.6322590112686157,
+      "step": 4234
+    },
+    {
+      "epoch": 1.8276159654800432,
+      "grad_norm": 1.2635412216186523,
+      "learning_rate": 8.60152809292968e-05,
+      "loss": 0.6540767550468445,
+      "step": 4236
+    },
+    {
+      "epoch": 1.8284789644012944,
+      "grad_norm": 1.1350682973861694,
+      "learning_rate": 8.599824503424086e-05,
+      "loss": 0.6719615459442139,
+      "step": 4238
+    },
+    {
+      "epoch": 1.8293419633225458,
+      "grad_norm": 1.2475734949111938,
+      "learning_rate": 8.598120045837043e-05,
+      "loss": 0.737215518951416,
+      "step": 4240
+    },
+    {
+      "epoch": 1.8302049622437973,
+      "grad_norm": 1.109307885169983,
+      "learning_rate": 8.596414720579568e-05,
+      "loss": 0.6711077094078064,
+      "step": 4242
+    },
+    {
+      "epoch": 1.8310679611650484,
+      "grad_norm": 1.3011112213134766,
+      "learning_rate": 8.594708528062898e-05,
+      "loss": 0.6282169222831726,
+      "step": 4244
+    },
+    {
+      "epoch": 1.8319309600862999,
+      "grad_norm": 0.9670302271842957,
+      "learning_rate": 8.593001468698471e-05,
+      "loss": 0.6156499981880188,
+      "step": 4246
+    },
+    {
+      "epoch": 1.8327939590075513,
+      "grad_norm": 1.1828471422195435,
+      "learning_rate": 8.591293542897938e-05,
+      "loss": 0.6415660977363586,
+      "step": 4248
+    },
+    {
+      "epoch": 1.8336569579288025,
+      "grad_norm": 0.9576874375343323,
+      "learning_rate": 8.589584751073156e-05,
+      "loss": 0.551155149936676,
+      "step": 4250
+    },
+    {
+      "epoch": 1.834519956850054,
+      "grad_norm": 1.163041591644287,
+      "learning_rate": 8.587875093636193e-05,
+      "loss": 0.6177648305892944,
+      "step": 4252
+    },
+    {
+      "epoch": 1.8353829557713053,
+      "grad_norm": 0.9363775849342346,
+      "learning_rate": 8.586164570999327e-05,
+      "loss": 0.5701619386672974,
+      "step": 4254
+    },
+    {
+      "epoch": 1.8362459546925565,
+      "grad_norm": 1.1433947086334229,
+      "learning_rate": 8.584453183575038e-05,
+      "loss": 0.6038158535957336,
+      "step": 4256
+    },
+    {
+      "epoch": 1.837108953613808,
+      "grad_norm": 1.1908625364303589,
+      "learning_rate": 8.582740931776026e-05,
+      "loss": 0.6887767314910889,
+      "step": 4258
+    },
+    {
+      "epoch": 1.8379719525350593,
+      "grad_norm": 1.0207780599594116,
+      "learning_rate": 8.581027816015187e-05,
+      "loss": 0.6212680339813232,
+      "step": 4260
+    },
+    {
+      "epoch": 1.8388349514563107,
+      "grad_norm": 1.213992953300476,
+      "learning_rate": 8.579313836705635e-05,
+      "loss": 0.6180729866027832,
+      "step": 4262
+    },
+    {
+      "epoch": 1.8396979503775621,
+      "grad_norm": 1.1175587177276611,
+      "learning_rate": 8.577598994260688e-05,
+      "loss": 0.6168033480644226,
+      "step": 4264
+    },
+    {
+      "epoch": 1.8405609492988133,
+      "grad_norm": 0.9955618977546692,
+      "learning_rate": 8.575883289093869e-05,
+      "loss": 0.5986083149909973,
+      "step": 4266
+    },
+    {
+      "epoch": 1.8414239482200647,
+      "grad_norm": 1.0735031366348267,
+      "learning_rate": 8.574166721618915e-05,
+      "loss": 0.6276304721832275,
+      "step": 4268
+    },
+    {
+      "epoch": 1.8422869471413161,
+      "grad_norm": 1.1660799980163574,
+      "learning_rate": 8.57244929224977e-05,
+      "loss": 0.6733853220939636,
+      "step": 4270
+    },
+    {
+      "epoch": 1.8431499460625673,
+      "grad_norm": 1.149767279624939,
+      "learning_rate": 8.570731001400581e-05,
+      "loss": 0.6379335522651672,
+      "step": 4272
+    },
+    {
+      "epoch": 1.8440129449838187,
+      "grad_norm": 1.0226882696151733,
+      "learning_rate": 8.56901184948571e-05,
+      "loss": 0.6092600226402283,
+      "step": 4274
+    },
+    {
+      "epoch": 1.8448759439050701,
+      "grad_norm": 1.116834282875061,
+      "learning_rate": 8.567291836919719e-05,
+      "loss": 0.6186777949333191,
+      "step": 4276
+    },
+    {
+      "epoch": 1.8457389428263213,
+      "grad_norm": 1.219565987586975,
+      "learning_rate": 8.565570964117385e-05,
+      "loss": 0.6713102459907532,
+      "step": 4278
+    },
+    {
+      "epoch": 1.846601941747573,
+      "grad_norm": 1.1785911321640015,
+      "learning_rate": 8.563849231493683e-05,
+      "loss": 0.6963690519332886,
+      "step": 4280
+    },
+    {
+      "epoch": 1.8474649406688242,
+      "grad_norm": 0.9836017489433289,
+      "learning_rate": 8.562126639463808e-05,
+      "loss": 0.5925922989845276,
+      "step": 4282
+    },
+    {
+      "epoch": 1.8483279395900754,
+      "grad_norm": 1.027338981628418,
+      "learning_rate": 8.560403188443151e-05,
+      "loss": 0.6510034203529358,
+      "step": 4284
+    },
+    {
+      "epoch": 1.849190938511327,
+      "grad_norm": 1.0148735046386719,
+      "learning_rate": 8.558678878847317e-05,
+      "loss": 0.6744941473007202,
+      "step": 4286
+    },
+    {
+      "epoch": 1.8500539374325782,
+      "grad_norm": 1.0071876049041748,
+      "learning_rate": 8.556953711092112e-05,
+      "loss": 0.5461649298667908,
+      "step": 4288
+    },
+    {
+      "epoch": 1.8509169363538296,
+      "grad_norm": 1.1736774444580078,
+      "learning_rate": 8.555227685593555e-05,
+      "loss": 0.650245189666748,
+      "step": 4290
+    },
+    {
+      "epoch": 1.851779935275081,
+      "grad_norm": 1.1919746398925781,
+      "learning_rate": 8.553500802767869e-05,
+      "loss": 0.6824072599411011,
+      "step": 4292
+    },
+    {
+      "epoch": 1.8526429341963322,
+      "grad_norm": 1.1280466318130493,
+      "learning_rate": 8.551773063031484e-05,
+      "loss": 0.6167742609977722,
+      "step": 4294
+    },
+    {
+      "epoch": 1.8535059331175836,
+      "grad_norm": 1.32740318775177,
+      "learning_rate": 8.550044466801034e-05,
+      "loss": 0.6046481132507324,
+      "step": 4296
+    },
+    {
+      "epoch": 1.854368932038835,
+      "grad_norm": 1.1019376516342163,
+      "learning_rate": 8.548315014493367e-05,
+      "loss": 0.6526167988777161,
+      "step": 4298
+    },
+    {
+      "epoch": 1.8552319309600862,
+      "grad_norm": 1.1523452997207642,
+      "learning_rate": 8.54658470652553e-05,
+      "loss": 0.6063379645347595,
+      "step": 4300
+    },
+    {
+      "epoch": 1.8552319309600862,
+      "eval_loss": 0.6890321373939514,
+      "eval_runtime": 498.5289,
+      "eval_samples_per_second": 4.132,
+      "eval_steps_per_second": 4.132,
+      "step": 4300
+    },
+    {
+      "epoch": 1.8560949298813376,
+      "grad_norm": 1.1315364837646484,
+      "learning_rate": 8.544853543314778e-05,
+      "loss": 0.6220276951789856,
+      "step": 4302
+    },
+    {
+      "epoch": 1.856957928802589,
+      "grad_norm": 1.2688732147216797,
+      "learning_rate": 8.543121525278575e-05,
+      "loss": 0.6813952326774597,
+      "step": 4304
+    },
+    {
+      "epoch": 1.8578209277238402,
+      "grad_norm": 1.2689135074615479,
+      "learning_rate": 8.541388652834588e-05,
+      "loss": 0.7254486083984375,
+      "step": 4306
+    },
+    {
+      "epoch": 1.8586839266450919,
+      "grad_norm": 1.198376178741455,
+      "learning_rate": 8.539654926400692e-05,
+      "loss": 0.6455208659172058,
+      "step": 4308
+    },
+    {
+      "epoch": 1.859546925566343,
+      "grad_norm": 1.0240654945373535,
+      "learning_rate": 8.537920346394969e-05,
+      "loss": 0.6356139779090881,
+      "step": 4310
+    },
+    {
+      "epoch": 1.8604099244875942,
+      "grad_norm": 1.1928554773330688,
+      "learning_rate": 8.536184913235704e-05,
+      "loss": 0.6543931365013123,
+      "step": 4312
+    },
+    {
+      "epoch": 1.8612729234088459,
+      "grad_norm": 1.190274953842163,
+      "learning_rate": 8.534448627341384e-05,
+      "loss": 0.7181617021560669,
+      "step": 4314
+    },
+    {
+      "epoch": 1.862135922330097,
+      "grad_norm": 1.2967594861984253,
+      "learning_rate": 8.532711489130715e-05,
+      "loss": 0.661147952079773,
+      "step": 4316
+    },
+    {
+      "epoch": 1.8629989212513485,
+      "grad_norm": 1.226694941520691,
+      "learning_rate": 8.530973499022594e-05,
+      "loss": 0.6840251684188843,
+      "step": 4318
+    },
+    {
+      "epoch": 1.8638619201725999,
+      "grad_norm": 1.098263144493103,
+      "learning_rate": 8.529234657436133e-05,
+      "loss": 0.6289029121398926,
+      "step": 4320
+    },
+    {
+      "epoch": 1.864724919093851,
+      "grad_norm": 1.0724968910217285,
+      "learning_rate": 8.527494964790644e-05,
+      "loss": 0.6778033375740051,
+      "step": 4322
+    },
+    {
+      "epoch": 1.8655879180151025,
+      "grad_norm": 1.0636100769042969,
+      "learning_rate": 8.525754421505646e-05,
+      "loss": 0.6152138710021973,
+      "step": 4324
+    },
+    {
+      "epoch": 1.866450916936354,
+      "grad_norm": 1.1964589357376099,
+      "learning_rate": 8.524013028000861e-05,
+      "loss": 0.6195679306983948,
+      "step": 4326
+    },
+    {
+      "epoch": 1.867313915857605,
+      "grad_norm": 0.9388785362243652,
+      "learning_rate": 8.522270784696223e-05,
+      "loss": 0.5703424215316772,
+      "step": 4328
+    },
+    {
+      "epoch": 1.8681769147788565,
+      "grad_norm": 1.0190461874008179,
+      "learning_rate": 8.520527692011864e-05,
+      "loss": 0.6520850658416748,
+      "step": 4330
+    },
+    {
+      "epoch": 1.869039913700108,
+      "grad_norm": 1.2388478517532349,
+      "learning_rate": 8.518783750368123e-05,
+      "loss": 0.6146131157875061,
+      "step": 4332
+    },
+    {
+      "epoch": 1.869902912621359,
+      "grad_norm": 1.5210362672805786,
+      "learning_rate": 8.517038960185542e-05,
+      "loss": 0.663919985294342,
+      "step": 4334
+    },
+    {
+      "epoch": 1.8707659115426107,
+      "grad_norm": 1.0638145208358765,
+      "learning_rate": 8.51529332188487e-05,
+      "loss": 0.6086801886558533,
+      "step": 4336
+    },
+    {
+      "epoch": 1.871628910463862,
+      "grad_norm": 1.1796083450317383,
+      "learning_rate": 8.513546835887061e-05,
+      "loss": 0.7241880297660828,
+      "step": 4338
+    },
+    {
+      "epoch": 1.8724919093851131,
+      "grad_norm": 1.2609213590621948,
+      "learning_rate": 8.511799502613273e-05,
+      "loss": 0.6798967123031616,
+      "step": 4340
+    },
+    {
+      "epoch": 1.8733549083063648,
+      "grad_norm": 1.1063754558563232,
+      "learning_rate": 8.510051322484864e-05,
+      "loss": 0.6314971446990967,
+      "step": 4342
+    },
+    {
+      "epoch": 1.874217907227616,
+      "grad_norm": 1.2268476486206055,
+      "learning_rate": 8.508302295923405e-05,
+      "loss": 0.6779396533966064,
+      "step": 4344
+    },
+    {
+      "epoch": 1.8750809061488674,
+      "grad_norm": 1.254714846611023,
+      "learning_rate": 8.50655242335066e-05,
+      "loss": 0.6303188800811768,
+      "step": 4346
+    },
+    {
+      "epoch": 1.8759439050701188,
+      "grad_norm": 1.0341795682907104,
+      "learning_rate": 8.504801705188607e-05,
+      "loss": 0.626371443271637,
+      "step": 4348
+    },
+    {
+      "epoch": 1.87680690399137,
+      "grad_norm": 1.148277997970581,
+      "learning_rate": 8.503050141859423e-05,
+      "loss": 0.6452032327651978,
+      "step": 4350
+    },
+    {
+      "epoch": 1.8776699029126214,
+      "grad_norm": 1.1761208772659302,
+      "learning_rate": 8.50129773378549e-05,
+      "loss": 0.7556618452072144,
+      "step": 4352
+    },
+    {
+      "epoch": 1.8785329018338728,
+      "grad_norm": 1.2172788381576538,
+      "learning_rate": 8.49954448138939e-05,
+      "loss": 0.6422954201698303,
+      "step": 4354
+    },
+    {
+      "epoch": 1.879395900755124,
+      "grad_norm": 1.11391019821167,
+      "learning_rate": 8.497790385093916e-05,
+      "loss": 0.6151976585388184,
+      "step": 4356
+    },
+    {
+      "epoch": 1.8802588996763754,
+      "grad_norm": 1.060592532157898,
+      "learning_rate": 8.496035445322057e-05,
+      "loss": 0.5745062232017517,
+      "step": 4358
+    },
+    {
+      "epoch": 1.8811218985976268,
+      "grad_norm": 1.2631596326828003,
+      "learning_rate": 8.49427966249701e-05,
+      "loss": 0.6658170223236084,
+      "step": 4360
+    },
+    {
+      "epoch": 1.881984897518878,
+      "grad_norm": 1.096709966659546,
+      "learning_rate": 8.492523037042179e-05,
+      "loss": 0.6444084644317627,
+      "step": 4362
+    },
+    {
+      "epoch": 1.8828478964401294,
+      "grad_norm": 1.1675381660461426,
+      "learning_rate": 8.490765569381159e-05,
+      "loss": 0.6504629850387573,
+      "step": 4364
+    },
+    {
+      "epoch": 1.8837108953613808,
+      "grad_norm": 1.2067846059799194,
+      "learning_rate": 8.489007259937758e-05,
+      "loss": 0.6365054249763489,
+      "step": 4366
+    },
+    {
+      "epoch": 1.884573894282632,
+      "grad_norm": 1.0922157764434814,
+      "learning_rate": 8.487248109135986e-05,
+      "loss": 0.5638332366943359,
+      "step": 4368
+    },
+    {
+      "epoch": 1.8854368932038836,
+      "grad_norm": 1.1562784910202026,
+      "learning_rate": 8.485488117400052e-05,
+      "loss": 0.5690349340438843,
+      "step": 4370
+    },
+    {
+      "epoch": 1.8862998921251348,
+      "grad_norm": 1.2277979850769043,
+      "learning_rate": 8.483727285154373e-05,
+      "loss": 0.6626796126365662,
+      "step": 4372
+    },
+    {
+      "epoch": 1.8871628910463862,
+      "grad_norm": 1.1707464456558228,
+      "learning_rate": 8.481965612823562e-05,
+      "loss": 0.7027351260185242,
+      "step": 4374
+    },
+    {
+      "epoch": 1.8880258899676376,
+      "grad_norm": 1.0433859825134277,
+      "learning_rate": 8.480203100832441e-05,
+      "loss": 0.6483538150787354,
+      "step": 4376
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 1.226797342300415,
+      "learning_rate": 8.478439749606031e-05,
+      "loss": 0.6338735818862915,
+      "step": 4378
+    },
+    {
+      "epoch": 1.8897518878101403,
+      "grad_norm": 1.0683629512786865,
+      "learning_rate": 8.476675559569557e-05,
+      "loss": 0.5816352367401123,
+      "step": 4380
+    },
+    {
+      "epoch": 1.8906148867313917,
+      "grad_norm": 1.030871033668518,
+      "learning_rate": 8.474910531148444e-05,
+      "loss": 0.6338825821876526,
+      "step": 4382
+    },
+    {
+      "epoch": 1.8914778856526429,
+      "grad_norm": 1.13720703125,
+      "learning_rate": 8.473144664768321e-05,
+      "loss": 0.6636899709701538,
+      "step": 4384
+    },
+    {
+      "epoch": 1.8923408845738943,
+      "grad_norm": 1.1313227415084839,
+      "learning_rate": 8.47137796085502e-05,
+      "loss": 0.7455838322639465,
+      "step": 4386
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.9600776433944702,
+      "learning_rate": 8.469610419834572e-05,
+      "loss": 0.6590731739997864,
+      "step": 4388
+    },
+    {
+      "epoch": 1.8940668824163969,
+      "grad_norm": 1.1583033800125122,
+      "learning_rate": 8.467842042133213e-05,
+      "loss": 0.6724982261657715,
+      "step": 4390
+    },
+    {
+      "epoch": 1.8949298813376483,
+      "grad_norm": 1.2100845575332642,
+      "learning_rate": 8.466072828177379e-05,
+      "loss": 0.6483606100082397,
+      "step": 4392
+    },
+    {
+      "epoch": 1.8957928802588997,
+      "grad_norm": 1.1870096921920776,
+      "learning_rate": 8.464302778393707e-05,
+      "loss": 0.6018115282058716,
+      "step": 4394
+    },
+    {
+      "epoch": 1.8966558791801509,
+      "grad_norm": 1.2486129999160767,
+      "learning_rate": 8.462531893209036e-05,
+      "loss": 0.621696412563324,
+      "step": 4396
+    },
+    {
+      "epoch": 1.8975188781014025,
+      "grad_norm": 1.1300864219665527,
+      "learning_rate": 8.46076017305041e-05,
+      "loss": 0.6095483899116516,
+      "step": 4398
+    },
+    {
+      "epoch": 1.8983818770226537,
+      "grad_norm": 1.0515706539154053,
+      "learning_rate": 8.458987618345069e-05,
+      "loss": 0.5899648666381836,
+      "step": 4400
+    },
+    {
+      "epoch": 1.8983818770226537,
+      "eval_loss": 0.687366783618927,
+      "eval_runtime": 497.5319,
+      "eval_samples_per_second": 4.14,
+      "eval_steps_per_second": 4.14,
+      "step": 4400
+    },
+    {
+      "epoch": 1.8992448759439051,
+      "grad_norm": 1.194498896598816,
+      "learning_rate": 8.457214229520455e-05,
+      "loss": 0.6216431856155396,
+      "step": 4402
+    },
+    {
+      "epoch": 1.9001078748651565,
+      "grad_norm": 1.193523645401001,
+      "learning_rate": 8.455440007004219e-05,
+      "loss": 0.5427393317222595,
+      "step": 4404
+    },
+    {
+      "epoch": 1.9009708737864077,
+      "grad_norm": 1.299268126487732,
+      "learning_rate": 8.453664951224198e-05,
+      "loss": 0.5973649024963379,
+      "step": 4406
+    },
+    {
+      "epoch": 1.9018338727076591,
+      "grad_norm": 1.0188872814178467,
+      "learning_rate": 8.451889062608446e-05,
+      "loss": 0.6070631742477417,
+      "step": 4408
+    },
+    {
+      "epoch": 1.9026968716289105,
+      "grad_norm": 1.1419471502304077,
+      "learning_rate": 8.450112341585206e-05,
+      "loss": 0.6672911643981934,
+      "step": 4410
+    },
+    {
+      "epoch": 1.9035598705501617,
+      "grad_norm": 1.0402454137802124,
+      "learning_rate": 8.44833478858293e-05,
+      "loss": 0.7246735095977783,
+      "step": 4412
+    },
+    {
+      "epoch": 1.9044228694714131,
+      "grad_norm": 1.1744418144226074,
+      "learning_rate": 8.446556404030264e-05,
+      "loss": 0.6222015619277954,
+      "step": 4414
+    },
+    {
+      "epoch": 1.9052858683926646,
+      "grad_norm": 0.9495980739593506,
+      "learning_rate": 8.44477718835606e-05,
+      "loss": 0.5821858048439026,
+      "step": 4416
+    },
+    {
+      "epoch": 1.9061488673139158,
+      "grad_norm": 1.1915034055709839,
+      "learning_rate": 8.442997141989362e-05,
+      "loss": 0.6768892407417297,
+      "step": 4418
+    },
+    {
+      "epoch": 1.9070118662351672,
+      "grad_norm": 1.083964228630066,
+      "learning_rate": 8.441216265359428e-05,
+      "loss": 0.5882668495178223,
+      "step": 4420
+    },
+    {
+      "epoch": 1.9078748651564186,
+      "grad_norm": 1.1280747652053833,
+      "learning_rate": 8.439434558895702e-05,
+      "loss": 0.6347549557685852,
+      "step": 4422
+    },
+    {
+      "epoch": 1.9087378640776698,
+      "grad_norm": 1.2000610828399658,
+      "learning_rate": 8.43765202302784e-05,
+      "loss": 0.6966704726219177,
+      "step": 4424
+    },
+    {
+      "epoch": 1.9096008629989214,
+      "grad_norm": 1.0745738744735718,
+      "learning_rate": 8.435868658185687e-05,
+      "loss": 0.6997919082641602,
+      "step": 4426
+    },
+    {
+      "epoch": 1.9104638619201726,
+      "grad_norm": 0.9859429597854614,
+      "learning_rate": 8.434084464799298e-05,
+      "loss": 0.6183820366859436,
+      "step": 4428
+    },
+    {
+      "epoch": 1.911326860841424,
+      "grad_norm": 1.1770579814910889,
+      "learning_rate": 8.43229944329892e-05,
+      "loss": 0.6683753132820129,
+      "step": 4430
+    },
+    {
+      "epoch": 1.9121898597626754,
+      "grad_norm": 1.180786371231079,
+      "learning_rate": 8.430513594115006e-05,
+      "loss": 0.6190024614334106,
+      "step": 4432
+    },
+    {
+      "epoch": 1.9130528586839266,
+      "grad_norm": 1.098340630531311,
+      "learning_rate": 8.428726917678201e-05,
+      "loss": 0.6235488653182983,
+      "step": 4434
+    },
+    {
+      "epoch": 1.913915857605178,
+      "grad_norm": 1.0071780681610107,
+      "learning_rate": 8.426939414419359e-05,
+      "loss": 0.6297350525856018,
+      "step": 4436
+    },
+    {
+      "epoch": 1.9147788565264294,
+      "grad_norm": 1.2799793481826782,
+      "learning_rate": 8.425151084769524e-05,
+      "loss": 0.6700394749641418,
+      "step": 4438
+    },
+    {
+      "epoch": 1.9156418554476806,
+      "grad_norm": 1.0455679893493652,
+      "learning_rate": 8.423361929159948e-05,
+      "loss": 0.5961334109306335,
+      "step": 4440
+    },
+    {
+      "epoch": 1.916504854368932,
+      "grad_norm": 1.212123990058899,
+      "learning_rate": 8.421571948022073e-05,
+      "loss": 0.7021171450614929,
+      "step": 4442
+    },
+    {
+      "epoch": 1.9173678532901834,
+      "grad_norm": 1.2384458780288696,
+      "learning_rate": 8.419781141787548e-05,
+      "loss": 0.6828257441520691,
+      "step": 4444
+    },
+    {
+      "epoch": 1.9182308522114346,
+      "grad_norm": 1.100545048713684,
+      "learning_rate": 8.417989510888217e-05,
+      "loss": 0.6749114394187927,
+      "step": 4446
+    },
+    {
+      "epoch": 1.919093851132686,
+      "grad_norm": 1.0241851806640625,
+      "learning_rate": 8.416197055756124e-05,
+      "loss": 0.6202837228775024,
+      "step": 4448
+    },
+    {
+      "epoch": 1.9199568500539375,
+      "grad_norm": 1.1138529777526855,
+      "learning_rate": 8.414403776823511e-05,
+      "loss": 0.6312109231948853,
+      "step": 4450
+    },
+    {
+      "epoch": 1.9208198489751886,
+      "grad_norm": 1.1574071645736694,
+      "learning_rate": 8.41260967452282e-05,
+      "loss": 0.677727997303009,
+      "step": 4452
+    },
+    {
+      "epoch": 1.9216828478964403,
+      "grad_norm": 1.1757844686508179,
+      "learning_rate": 8.41081474928669e-05,
+      "loss": 0.625155508518219,
+      "step": 4454
+    },
+    {
+      "epoch": 1.9225458468176915,
+      "grad_norm": 1.2577253580093384,
+      "learning_rate": 8.409019001547959e-05,
+      "loss": 0.6271302700042725,
+      "step": 4456
+    },
+    {
+      "epoch": 1.9234088457389427,
+      "grad_norm": 1.0348252058029175,
+      "learning_rate": 8.407222431739663e-05,
+      "loss": 0.60232013463974,
+      "step": 4458
+    },
+    {
+      "epoch": 1.9242718446601943,
+      "grad_norm": 1.2921109199523926,
+      "learning_rate": 8.405425040295039e-05,
+      "loss": 0.6764297485351562,
+      "step": 4460
+    },
+    {
+      "epoch": 1.9251348435814455,
+      "grad_norm": 1.2089459896087646,
+      "learning_rate": 8.403626827647517e-05,
+      "loss": 0.5721253752708435,
+      "step": 4462
+    },
+    {
+      "epoch": 1.925997842502697,
+      "grad_norm": 1.1275287866592407,
+      "learning_rate": 8.401827794230729e-05,
+      "loss": 0.6360597014427185,
+      "step": 4464
+    },
+    {
+      "epoch": 1.9268608414239483,
+      "grad_norm": 1.162341594696045,
+      "learning_rate": 8.400027940478506e-05,
+      "loss": 0.654008150100708,
+      "step": 4466
+    },
+    {
+      "epoch": 1.9277238403451995,
+      "grad_norm": 1.1538556814193726,
+      "learning_rate": 8.398227266824871e-05,
+      "loss": 0.6744330525398254,
+      "step": 4468
+    },
+    {
+      "epoch": 1.928586839266451,
+      "grad_norm": 0.9633672833442688,
+      "learning_rate": 8.39642577370405e-05,
+      "loss": 0.6474409103393555,
+      "step": 4470
+    },
+    {
+      "epoch": 1.9294498381877023,
+      "grad_norm": 1.1969858407974243,
+      "learning_rate": 8.394623461550465e-05,
+      "loss": 0.6781411170959473,
+      "step": 4472
+    },
+    {
+      "epoch": 1.9303128371089535,
+      "grad_norm": 1.160700798034668,
+      "learning_rate": 8.392820330798733e-05,
+      "loss": 0.6781613826751709,
+      "step": 4474
+    },
+    {
+      "epoch": 1.931175836030205,
+      "grad_norm": 1.201220154762268,
+      "learning_rate": 8.391016381883678e-05,
+      "loss": 0.6875717043876648,
+      "step": 4476
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 1.090810775756836,
+      "learning_rate": 8.389211615240306e-05,
+      "loss": 0.6446496248245239,
+      "step": 4478
+    },
+    {
+      "epoch": 1.9329018338727075,
+      "grad_norm": 1.0396226644515991,
+      "learning_rate": 8.387406031303833e-05,
+      "loss": 0.6835271716117859,
+      "step": 4480
+    },
+    {
+      "epoch": 1.9337648327939592,
+      "grad_norm": 1.1818757057189941,
+      "learning_rate": 8.385599630509664e-05,
+      "loss": 0.587812066078186,
+      "step": 4482
+    },
+    {
+      "epoch": 1.9346278317152104,
+      "grad_norm": 1.0651488304138184,
+      "learning_rate": 8.383792413293408e-05,
+      "loss": 0.6302334666252136,
+      "step": 4484
+    },
+    {
+      "epoch": 1.9354908306364615,
+      "grad_norm": 1.0889242887496948,
+      "learning_rate": 8.381984380090865e-05,
+      "loss": 0.5748880505561829,
+      "step": 4486
+    },
+    {
+      "epoch": 1.9363538295577132,
+      "grad_norm": 1.0353775024414062,
+      "learning_rate": 8.380175531338037e-05,
+      "loss": 0.642596960067749,
+      "step": 4488
+    },
+    {
+      "epoch": 1.9372168284789644,
+      "grad_norm": 1.089475154876709,
+      "learning_rate": 8.378365867471115e-05,
+      "loss": 0.6552023887634277,
+      "step": 4490
+    },
+    {
+      "epoch": 1.9380798274002158,
+      "grad_norm": 1.0442018508911133,
+      "learning_rate": 8.376555388926495e-05,
+      "loss": 0.6077851057052612,
+      "step": 4492
+    },
+    {
+      "epoch": 1.9389428263214672,
+      "grad_norm": 1.0850406885147095,
+      "learning_rate": 8.374744096140764e-05,
+      "loss": 0.5917877554893494,
+      "step": 4494
+    },
+    {
+      "epoch": 1.9398058252427184,
+      "grad_norm": 1.0924079418182373,
+      "learning_rate": 8.372931989550708e-05,
+      "loss": 0.6818947196006775,
+      "step": 4496
+    },
+    {
+      "epoch": 1.9406688241639698,
+      "grad_norm": 1.2747828960418701,
+      "learning_rate": 8.371119069593308e-05,
+      "loss": 0.6909443736076355,
+      "step": 4498
+    },
+    {
+      "epoch": 1.9415318230852212,
+      "grad_norm": 1.0612404346466064,
+      "learning_rate": 8.369305336705742e-05,
+      "loss": 0.6680766344070435,
+      "step": 4500
+    },
+    {
+      "epoch": 1.9415318230852212,
+      "eval_loss": 0.6851214170455933,
+      "eval_runtime": 497.1141,
+      "eval_samples_per_second": 4.144,
+      "eval_steps_per_second": 4.144,
+      "step": 4500
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.690843183019114e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-4500/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-4500/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-4500/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/README.md b/sft_devstral_24B/checkpoints/checkpoint-500/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3a96a8bcb45ee112cf328fecbca168478e07ca85
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/README.md
@@ -0,0 +1,207 @@
+---
+base_model: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.0
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/adapter_config.json b/sft_devstral_24B/checkpoints/checkpoint-500/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fb650dec8a9177ca67ec2187f9199091ab52f5c
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/adapter_config.json
@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "CPT/runs/cpt_run_v1/merged_24b_cpt_lora",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/adapter_model.safetensors b/sft_devstral_24B/checkpoints/checkpoint-500/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..37231ea47bed45ba9ddd66c2ac17eb3dc7b1213f
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b257eba39acc9320d448c0df207650280dd4e0f9254eaaede9e3885d316c736
+size 45690960
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/optimizer.pt b/sft_devstral_24B/checkpoints/checkpoint-500/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..631245aad7b7ce9899e322fbe5cf627c1aca4d41
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf213f2460084d314c1c2e94bab21698a616b33366d7c3b27a324e7bc6a94eb9
+size 78912907
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/rng_state.pth b/sft_devstral_24B/checkpoints/checkpoint-500/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d280c22840842ebd620c89f61de8e8d7bad746f0
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb7b1a123877193fb4cca27cf6da17a2b0657192937cc6fbbf8532f9b10c938f
+size 14645
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/scheduler.pt b/sft_devstral_24B/checkpoints/checkpoint-500/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..810cb7bd9b29760fd684b9e22fb39893e9ae192d
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad5db93fb75f2909659c589582fdac02a88ddff299760a8161b204f75011b4da
+size 1465
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/trainer_state.json b/sft_devstral_24B/checkpoints/checkpoint-500/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..d9e53021829fd4c8884d24ab4769f368b35237c2
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/trainer_state.json
@@ -0,0 +1,1833 @@
+{
+  "best_global_step": 500,
+  "best_metric": 0.8945327997207642,
+  "best_model_checkpoint": "runs/instruct_run_24b/checkpoints/checkpoint-500",
+  "epoch": 0.21574973031283712,
+  "eval_steps": 100,
+  "global_step": 500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008629989212513484,
+      "grad_norm": 1.9118273258209229,
+      "learning_rate": 8.984725965858042e-08,
+      "loss": 1.398809790611267,
+      "step": 2
+    },
+    {
+      "epoch": 0.001725997842502697,
+      "grad_norm": 1.785147786140442,
+      "learning_rate": 2.6954177897574125e-07,
+      "loss": 1.3699476718902588,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025889967637540453,
+      "grad_norm": 1.6857414245605469,
+      "learning_rate": 4.492362982929021e-07,
+      "loss": 1.4487650394439697,
+      "step": 6
+    },
+    {
+      "epoch": 0.003451995685005394,
+      "grad_norm": 1.7351312637329102,
+      "learning_rate": 6.28930817610063e-07,
+      "loss": 1.3360817432403564,
+      "step": 8
+    },
+    {
+      "epoch": 0.004314994606256742,
+      "grad_norm": 1.7769269943237305,
+      "learning_rate": 8.086253369272238e-07,
+      "loss": 1.4294791221618652,
+      "step": 10
+    },
+    {
+      "epoch": 0.005177993527508091,
+      "grad_norm": 1.776668906211853,
+      "learning_rate": 9.883198562443846e-07,
+      "loss": 1.3927934169769287,
+      "step": 12
+    },
+    {
+      "epoch": 0.006040992448759439,
+      "grad_norm": 1.500367283821106,
+      "learning_rate": 1.1680143755615454e-06,
+      "loss": 1.4140764474868774,
+      "step": 14
+    },
+    {
+      "epoch": 0.006903991370010788,
+      "grad_norm": 1.7255949974060059,
+      "learning_rate": 1.3477088948787064e-06,
+      "loss": 1.3873623609542847,
+      "step": 16
+    },
+    {
+      "epoch": 0.007766990291262136,
+      "grad_norm": 1.9073798656463623,
+      "learning_rate": 1.5274034141958671e-06,
+      "loss": 1.4562759399414062,
+      "step": 18
+    },
+    {
+      "epoch": 0.008629989212513484,
+      "grad_norm": 1.7253859043121338,
+      "learning_rate": 1.707097933513028e-06,
+      "loss": 1.4264800548553467,
+      "step": 20
+    },
+    {
+      "epoch": 0.009492988133764833,
+      "grad_norm": 1.4858287572860718,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 1.3717138767242432,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355987055016181,
+      "grad_norm": 1.7516961097717285,
+      "learning_rate": 2.0664869721473496e-06,
+      "loss": 1.3605879545211792,
+      "step": 24
+    },
+    {
+      "epoch": 0.01121898597626753,
+      "grad_norm": 1.754793643951416,
+      "learning_rate": 2.2461814914645104e-06,
+      "loss": 1.465327262878418,
+      "step": 26
+    },
+    {
+      "epoch": 0.012081984897518877,
+      "grad_norm": 1.5204869508743286,
+      "learning_rate": 2.425876010781671e-06,
+      "loss": 1.3205785751342773,
+      "step": 28
+    },
+    {
+      "epoch": 0.012944983818770227,
+      "grad_norm": 2.03002667427063,
+      "learning_rate": 2.605570530098832e-06,
+      "loss": 1.4751381874084473,
+      "step": 30
+    },
+    {
+      "epoch": 0.013807982740021575,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.785265049415993e-06,
+      "loss": 1.3976212739944458,
+      "step": 32
+    },
+    {
+      "epoch": 0.014670981661272923,
+      "grad_norm": 1.7034062147140503,
+      "learning_rate": 2.964959568733154e-06,
+      "loss": 1.4062179327011108,
+      "step": 34
+    },
+    {
+      "epoch": 0.015533980582524271,
+      "grad_norm": 1.5724587440490723,
+      "learning_rate": 3.1446540880503146e-06,
+      "loss": 1.2940696477890015,
+      "step": 36
+    },
+    {
+      "epoch": 0.01639697950377562,
+      "grad_norm": 1.7072066068649292,
+      "learning_rate": 3.324348607367475e-06,
+      "loss": 1.4216598272323608,
+      "step": 38
+    },
+    {
+      "epoch": 0.017259978425026967,
+      "grad_norm": 1.6051130294799805,
+      "learning_rate": 3.504043126684636e-06,
+      "loss": 1.4204131364822388,
+      "step": 40
+    },
+    {
+      "epoch": 0.018122977346278317,
+      "grad_norm": 1.5114386081695557,
+      "learning_rate": 3.683737646001797e-06,
+      "loss": 1.3492953777313232,
+      "step": 42
+    },
+    {
+      "epoch": 0.018985976267529667,
+      "grad_norm": 1.5557284355163574,
+      "learning_rate": 3.863432165318958e-06,
+      "loss": 1.4241598844528198,
+      "step": 44
+    },
+    {
+      "epoch": 0.019848975188781013,
+      "grad_norm": 1.5814268589019775,
+      "learning_rate": 4.0431266846361185e-06,
+      "loss": 1.3903125524520874,
+      "step": 46
+    },
+    {
+      "epoch": 0.020711974110032363,
+      "grad_norm": 1.53256356716156,
+      "learning_rate": 4.22282120395328e-06,
+      "loss": 1.3726638555526733,
+      "step": 48
+    },
+    {
+      "epoch": 0.021574973031283712,
+      "grad_norm": 1.5670084953308105,
+      "learning_rate": 4.40251572327044e-06,
+      "loss": 1.2536468505859375,
+      "step": 50
+    },
+    {
+      "epoch": 0.02243797195253506,
+      "grad_norm": 1.3708246946334839,
+      "learning_rate": 4.582210242587602e-06,
+      "loss": 1.2309948205947876,
+      "step": 52
+    },
+    {
+      "epoch": 0.02330097087378641,
+      "grad_norm": 1.5670109987258911,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.3564441204071045,
+      "step": 54
+    },
+    {
+      "epoch": 0.024163969795037755,
+      "grad_norm": 1.3550655841827393,
+      "learning_rate": 4.941599281221923e-06,
+      "loss": 1.2251620292663574,
+      "step": 56
+    },
+    {
+      "epoch": 0.025026968716289105,
+      "grad_norm": 1.4416121244430542,
+      "learning_rate": 5.121293800539084e-06,
+      "loss": 1.2395249605178833,
+      "step": 58
+    },
+    {
+      "epoch": 0.025889967637540454,
+      "grad_norm": 1.4638670682907104,
+      "learning_rate": 5.300988319856245e-06,
+      "loss": 1.3229656219482422,
+      "step": 60
+    },
+    {
+      "epoch": 0.0267529665587918,
+      "grad_norm": 1.2204002141952515,
+      "learning_rate": 5.4806828391734054e-06,
+      "loss": 1.1076340675354004,
+      "step": 62
+    },
+    {
+      "epoch": 0.02761596548004315,
+      "grad_norm": 1.2641915082931519,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.237392544746399,
+      "step": 64
+    },
+    {
+      "epoch": 0.0284789644012945,
+      "grad_norm": 1.4635342359542847,
+      "learning_rate": 5.840071877807727e-06,
+      "loss": 1.3567079305648804,
+      "step": 66
+    },
+    {
+      "epoch": 0.029341963322545846,
+      "grad_norm": 1.5813465118408203,
+      "learning_rate": 6.019766397124888e-06,
+      "loss": 1.3595421314239502,
+      "step": 68
+    },
+    {
+      "epoch": 0.030204962243797196,
+      "grad_norm": 1.435356616973877,
+      "learning_rate": 6.1994609164420485e-06,
+      "loss": 1.2183864116668701,
+      "step": 70
+    },
+    {
+      "epoch": 0.031067961165048542,
+      "grad_norm": 1.4559556245803833,
+      "learning_rate": 6.37915543575921e-06,
+      "loss": 1.2084243297576904,
+      "step": 72
+    },
+    {
+      "epoch": 0.03193096008629989,
+      "grad_norm": 1.2870471477508545,
+      "learning_rate": 6.558849955076371e-06,
+      "loss": 1.2476426362991333,
+      "step": 74
+    },
+    {
+      "epoch": 0.03279395900755124,
+      "grad_norm": 1.3345798254013062,
+      "learning_rate": 6.738544474393531e-06,
+      "loss": 1.2200642824172974,
+      "step": 76
+    },
+    {
+      "epoch": 0.03365695792880259,
+      "grad_norm": 1.48223876953125,
+      "learning_rate": 6.918238993710692e-06,
+      "loss": 1.2497955560684204,
+      "step": 78
+    },
+    {
+      "epoch": 0.034519956850053934,
+      "grad_norm": 1.2421257495880127,
+      "learning_rate": 7.097933513027853e-06,
+      "loss": 1.1987318992614746,
+      "step": 80
+    },
+    {
+      "epoch": 0.035382955771305284,
+      "grad_norm": 1.5042678117752075,
+      "learning_rate": 7.277628032345015e-06,
+      "loss": 1.218438982963562,
+      "step": 82
+    },
+    {
+      "epoch": 0.036245954692556634,
+      "grad_norm": 1.3047534227371216,
+      "learning_rate": 7.457322551662174e-06,
+      "loss": 1.1181117296218872,
+      "step": 84
+    },
+    {
+      "epoch": 0.037108953613807984,
+      "grad_norm": 1.3297117948532104,
+      "learning_rate": 7.637017070979335e-06,
+      "loss": 1.047302484512329,
+      "step": 86
+    },
+    {
+      "epoch": 0.03797195253505933,
+      "grad_norm": 1.4515916109085083,
+      "learning_rate": 7.816711590296496e-06,
+      "loss": 1.1821871995925903,
+      "step": 88
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 1.5435426235198975,
+      "learning_rate": 7.996406109613658e-06,
+      "loss": 1.1171189546585083,
+      "step": 90
+    },
+    {
+      "epoch": 0.039697950377562026,
+      "grad_norm": 1.5270947217941284,
+      "learning_rate": 8.176100628930818e-06,
+      "loss": 1.067047357559204,
+      "step": 92
+    },
+    {
+      "epoch": 0.040560949298813376,
+      "grad_norm": 2.4402992725372314,
+      "learning_rate": 8.35579514824798e-06,
+      "loss": 1.0745081901550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.041423948220064725,
+      "grad_norm": 1.7703081369400024,
+      "learning_rate": 8.53548966756514e-06,
+      "loss": 1.252164602279663,
+      "step": 96
+    },
+    {
+      "epoch": 0.042286947141316075,
+      "grad_norm": 1.4324337244033813,
+      "learning_rate": 8.715184186882301e-06,
+      "loss": 1.1030420064926147,
+      "step": 98
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 8.89487870619946e-06,
+      "loss": 1.0817893743515015,
+      "step": 100
+    },
+    {
+      "epoch": 0.043149946062567425,
+      "eval_loss": 1.128923773765564,
+      "eval_runtime": 825.0166,
+      "eval_samples_per_second": 2.497,
+      "eval_steps_per_second": 2.497,
+      "step": 100
+    },
+    {
+      "epoch": 0.04401294498381877,
+      "grad_norm": 1.5833592414855957,
+      "learning_rate": 9.074573225516622e-06,
+      "loss": 1.0982940196990967,
+      "step": 102
+    },
+    {
+      "epoch": 0.04487594390507012,
+      "grad_norm": 1.5289201736450195,
+      "learning_rate": 9.254267744833784e-06,
+      "loss": 1.0529299974441528,
+      "step": 104
+    },
+    {
+      "epoch": 0.04573894282632147,
+      "grad_norm": 1.4507837295532227,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.0896434783935547,
+      "step": 106
+    },
+    {
+      "epoch": 0.04660194174757282,
+      "grad_norm": 1.3846485614776611,
+      "learning_rate": 9.613656783468104e-06,
+      "loss": 1.1196398735046387,
+      "step": 108
+    },
+    {
+      "epoch": 0.04746494066882417,
+      "grad_norm": 1.4094940423965454,
+      "learning_rate": 9.793351302785265e-06,
+      "loss": 1.1305382251739502,
+      "step": 110
+    },
+    {
+      "epoch": 0.04832793959007551,
+      "grad_norm": 1.7674189805984497,
+      "learning_rate": 9.973045822102427e-06,
+      "loss": 1.0138700008392334,
+      "step": 112
+    },
+    {
+      "epoch": 0.04919093851132686,
+      "grad_norm": 1.6083961725234985,
+      "learning_rate": 1.0152740341419587e-05,
+      "loss": 1.0583173036575317,
+      "step": 114
+    },
+    {
+      "epoch": 0.05005393743257821,
+      "grad_norm": 1.7483229637145996,
+      "learning_rate": 1.0332434860736747e-05,
+      "loss": 1.1116671562194824,
+      "step": 116
+    },
+    {
+      "epoch": 0.05091693635382956,
+      "grad_norm": 1.8256834745407104,
+      "learning_rate": 1.0512129380053909e-05,
+      "loss": 1.0832366943359375,
+      "step": 118
+    },
+    {
+      "epoch": 0.05177993527508091,
+      "grad_norm": 1.5875557661056519,
+      "learning_rate": 1.069182389937107e-05,
+      "loss": 1.1089236736297607,
+      "step": 120
+    },
+    {
+      "epoch": 0.05264293419633225,
+      "grad_norm": 1.6553549766540527,
+      "learning_rate": 1.087151841868823e-05,
+      "loss": 1.1191619634628296,
+      "step": 122
+    },
+    {
+      "epoch": 0.0535059331175836,
+      "grad_norm": 1.3675798177719116,
+      "learning_rate": 1.1051212938005392e-05,
+      "loss": 1.031973958015442,
+      "step": 124
+    },
+    {
+      "epoch": 0.05436893203883495,
+      "grad_norm": 1.5343760251998901,
+      "learning_rate": 1.1230907457322552e-05,
+      "loss": 0.9964010119438171,
+      "step": 126
+    },
+    {
+      "epoch": 0.0552319309600863,
+      "grad_norm": 1.5928982496261597,
+      "learning_rate": 1.1410601976639713e-05,
+      "loss": 1.1144609451293945,
+      "step": 128
+    },
+    {
+      "epoch": 0.05609492988133765,
+      "grad_norm": 1.6411585807800293,
+      "learning_rate": 1.1590296495956873e-05,
+      "loss": 0.9953935146331787,
+      "step": 130
+    },
+    {
+      "epoch": 0.056957928802589,
+      "grad_norm": 1.6077220439910889,
+      "learning_rate": 1.1769991015274035e-05,
+      "loss": 0.9940203428268433,
+      "step": 132
+    },
+    {
+      "epoch": 0.05782092772384034,
+      "grad_norm": 1.7327438592910767,
+      "learning_rate": 1.1949685534591196e-05,
+      "loss": 1.045546293258667,
+      "step": 134
+    },
+    {
+      "epoch": 0.05868392664509169,
+      "grad_norm": 1.607404112815857,
+      "learning_rate": 1.2129380053908356e-05,
+      "loss": 0.9932523369789124,
+      "step": 136
+    },
+    {
+      "epoch": 0.05954692556634304,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.2309074573225516e-05,
+      "loss": 0.9584172368049622,
+      "step": 138
+    },
+    {
+      "epoch": 0.06040992448759439,
+      "grad_norm": 1.5453966856002808,
+      "learning_rate": 1.2488769092542678e-05,
+      "loss": 0.9733421206474304,
+      "step": 140
+    },
+    {
+      "epoch": 0.06127292340884574,
+      "grad_norm": 1.546397089958191,
+      "learning_rate": 1.2668463611859838e-05,
+      "loss": 0.9971115589141846,
+      "step": 142
+    },
+    {
+      "epoch": 0.062135922330097085,
+      "grad_norm": 1.5259592533111572,
+      "learning_rate": 1.2848158131177e-05,
+      "loss": 1.0994833707809448,
+      "step": 144
+    },
+    {
+      "epoch": 0.06299892125134844,
+      "grad_norm": 1.56397545337677,
+      "learning_rate": 1.302785265049416e-05,
+      "loss": 1.0203801393508911,
+      "step": 146
+    },
+    {
+      "epoch": 0.06386192017259978,
+      "grad_norm": 1.618053674697876,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 1.0223872661590576,
+      "step": 148
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 1.9544042348861694,
+      "learning_rate": 1.3387241689128482e-05,
+      "loss": 1.0207403898239136,
+      "step": 150
+    },
+    {
+      "epoch": 0.06558791801510248,
+      "grad_norm": 1.7001299858093262,
+      "learning_rate": 1.3566936208445644e-05,
+      "loss": 1.025482177734375,
+      "step": 152
+    },
+    {
+      "epoch": 0.06645091693635383,
+      "grad_norm": 1.9833810329437256,
+      "learning_rate": 1.3746630727762804e-05,
+      "loss": 1.0077980756759644,
+      "step": 154
+    },
+    {
+      "epoch": 0.06731391585760518,
+      "grad_norm": 1.6378769874572754,
+      "learning_rate": 1.3926325247079966e-05,
+      "loss": 1.0046619176864624,
+      "step": 156
+    },
+    {
+      "epoch": 0.06817691477885653,
+      "grad_norm": 1.762683629989624,
+      "learning_rate": 1.4106019766397124e-05,
+      "loss": 1.0332108736038208,
+      "step": 158
+    },
+    {
+      "epoch": 0.06903991370010787,
+      "grad_norm": 1.5698739290237427,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.973042905330658,
+      "step": 160
+    },
+    {
+      "epoch": 0.06990291262135923,
+      "grad_norm": 1.6132292747497559,
+      "learning_rate": 1.4465408805031447e-05,
+      "loss": 0.9966566562652588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07076591154261057,
+      "grad_norm": 1.989534616470337,
+      "learning_rate": 1.4645103324348609e-05,
+      "loss": 1.078708529472351,
+      "step": 164
+    },
+    {
+      "epoch": 0.07162891046386193,
+      "grad_norm": 1.8199284076690674,
+      "learning_rate": 1.4824797843665769e-05,
+      "loss": 0.9075545072555542,
+      "step": 166
+    },
+    {
+      "epoch": 0.07249190938511327,
+      "grad_norm": 1.6908831596374512,
+      "learning_rate": 1.500449236298293e-05,
+      "loss": 0.9268680810928345,
+      "step": 168
+    },
+    {
+      "epoch": 0.07335490830636461,
+      "grad_norm": 1.5648242235183716,
+      "learning_rate": 1.5184186882300092e-05,
+      "loss": 1.0361101627349854,
+      "step": 170
+    },
+    {
+      "epoch": 0.07421790722761597,
+      "grad_norm": 1.8292360305786133,
+      "learning_rate": 1.5363881401617252e-05,
+      "loss": 0.9197717905044556,
+      "step": 172
+    },
+    {
+      "epoch": 0.07508090614886731,
+      "grad_norm": 1.6882517337799072,
+      "learning_rate": 1.554357592093441e-05,
+      "loss": 0.9661652445793152,
+      "step": 174
+    },
+    {
+      "epoch": 0.07594390507011867,
+      "grad_norm": 1.695921778678894,
+      "learning_rate": 1.572327044025157e-05,
+      "loss": 1.0006932020187378,
+      "step": 176
+    },
+    {
+      "epoch": 0.07680690399137001,
+      "grad_norm": 2.0128681659698486,
+      "learning_rate": 1.5902964959568733e-05,
+      "loss": 1.0096064805984497,
+      "step": 178
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.6833231449127197,
+      "learning_rate": 1.6082659478885895e-05,
+      "loss": 0.972679853439331,
+      "step": 180
+    },
+    {
+      "epoch": 0.07853290183387271,
+      "grad_norm": 1.6940670013427734,
+      "learning_rate": 1.6262353998203056e-05,
+      "loss": 0.9232968091964722,
+      "step": 182
+    },
+    {
+      "epoch": 0.07939590075512405,
+      "grad_norm": 2.0851218700408936,
+      "learning_rate": 1.6442048517520218e-05,
+      "loss": 0.981245219707489,
+      "step": 184
+    },
+    {
+      "epoch": 0.08025889967637541,
+      "grad_norm": 1.3769028186798096,
+      "learning_rate": 1.662174303683738e-05,
+      "loss": 0.9762678146362305,
+      "step": 186
+    },
+    {
+      "epoch": 0.08112189859762675,
+      "grad_norm": 1.85263192653656,
+      "learning_rate": 1.6801437556154538e-05,
+      "loss": 0.9905523061752319,
+      "step": 188
+    },
+    {
+      "epoch": 0.08198489751887811,
+      "grad_norm": 1.654221534729004,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.8457058072090149,
+      "step": 190
+    },
+    {
+      "epoch": 0.08284789644012945,
+      "grad_norm": 1.7715294361114502,
+      "learning_rate": 1.7160826594788858e-05,
+      "loss": 1.0171988010406494,
+      "step": 192
+    },
+    {
+      "epoch": 0.0837108953613808,
+      "grad_norm": 1.7162128686904907,
+      "learning_rate": 1.734052111410602e-05,
+      "loss": 0.9767695069313049,
+      "step": 194
+    },
+    {
+      "epoch": 0.08457389428263215,
+      "grad_norm": 1.7938538789749146,
+      "learning_rate": 1.752021563342318e-05,
+      "loss": 0.9786219596862793,
+      "step": 196
+    },
+    {
+      "epoch": 0.0854368932038835,
+      "grad_norm": 1.944226861000061,
+      "learning_rate": 1.7699910152740342e-05,
+      "loss": 0.9941093921661377,
+      "step": 198
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "grad_norm": 1.816835880279541,
+      "learning_rate": 1.7879604672057504e-05,
+      "loss": 0.8924376964569092,
+      "step": 200
+    },
+    {
+      "epoch": 0.08629989212513485,
+      "eval_loss": 0.981117308139801,
+      "eval_runtime": 826.4469,
+      "eval_samples_per_second": 2.493,
+      "eval_steps_per_second": 2.493,
+      "step": 200
+    },
+    {
+      "epoch": 0.08716289104638619,
+      "grad_norm": 1.8230189085006714,
+      "learning_rate": 1.8059299191374666e-05,
+      "loss": 0.9908216595649719,
+      "step": 202
+    },
+    {
+      "epoch": 0.08802588996763754,
+      "grad_norm": 1.8223077058792114,
+      "learning_rate": 1.8238993710691824e-05,
+      "loss": 0.9657744765281677,
+      "step": 204
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 1.811195731163025,
+      "learning_rate": 1.8418688230008986e-05,
+      "loss": 0.9397313594818115,
+      "step": 206
+    },
+    {
+      "epoch": 0.08975188781014024,
+      "grad_norm": 1.579998254776001,
+      "learning_rate": 1.8598382749326147e-05,
+      "loss": 0.8770087361335754,
+      "step": 208
+    },
+    {
+      "epoch": 0.09061488673139159,
+      "grad_norm": 1.592136263847351,
+      "learning_rate": 1.8778077268643305e-05,
+      "loss": 0.922799825668335,
+      "step": 210
+    },
+    {
+      "epoch": 0.09147788565264293,
+      "grad_norm": 1.9334869384765625,
+      "learning_rate": 1.8957771787960467e-05,
+      "loss": 0.9007947444915771,
+      "step": 212
+    },
+    {
+      "epoch": 0.09234088457389428,
+      "grad_norm": 1.8619475364685059,
+      "learning_rate": 1.913746630727763e-05,
+      "loss": 0.9261049628257751,
+      "step": 214
+    },
+    {
+      "epoch": 0.09320388349514563,
+      "grad_norm": 1.751584529876709,
+      "learning_rate": 1.931716082659479e-05,
+      "loss": 0.9319002628326416,
+      "step": 216
+    },
+    {
+      "epoch": 0.09406688241639698,
+      "grad_norm": 2.1829185485839844,
+      "learning_rate": 1.9496855345911952e-05,
+      "loss": 0.947629451751709,
+      "step": 218
+    },
+    {
+      "epoch": 0.09492988133764833,
+      "grad_norm": 1.8973567485809326,
+      "learning_rate": 1.967654986522911e-05,
+      "loss": 0.9213902354240417,
+      "step": 220
+    },
+    {
+      "epoch": 0.09579288025889968,
+      "grad_norm": 1.9995585680007935,
+      "learning_rate": 1.985624438454627e-05,
+      "loss": 0.9058182835578918,
+      "step": 222
+    },
+    {
+      "epoch": 0.09665587918015102,
+      "grad_norm": 1.9369088411331177,
+      "learning_rate": 2.0035938903863433e-05,
+      "loss": 0.8918466567993164,
+      "step": 224
+    },
+    {
+      "epoch": 0.09751887810140238,
+      "grad_norm": 2.1926183700561523,
+      "learning_rate": 2.0215633423180595e-05,
+      "loss": 1.009294033050537,
+      "step": 226
+    },
+    {
+      "epoch": 0.09838187702265372,
+      "grad_norm": 1.872399926185608,
+      "learning_rate": 2.0395327942497757e-05,
+      "loss": 0.9543544054031372,
+      "step": 228
+    },
+    {
+      "epoch": 0.09924487594390508,
+      "grad_norm": 2.011009454727173,
+      "learning_rate": 2.0575022461814915e-05,
+      "loss": 0.9188834428787231,
+      "step": 230
+    },
+    {
+      "epoch": 0.10010787486515642,
+      "grad_norm": 1.8422952890396118,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.84498131275177,
+      "step": 232
+    },
+    {
+      "epoch": 0.10097087378640776,
+      "grad_norm": 2.208738088607788,
+      "learning_rate": 2.0934411500449238e-05,
+      "loss": 0.9864624738693237,
+      "step": 234
+    },
+    {
+      "epoch": 0.10183387270765912,
+      "grad_norm": 1.6119916439056396,
+      "learning_rate": 2.1114106019766396e-05,
+      "loss": 0.928769588470459,
+      "step": 236
+    },
+    {
+      "epoch": 0.10269687162891046,
+      "grad_norm": 1.906336784362793,
+      "learning_rate": 2.1293800539083558e-05,
+      "loss": 0.885361909866333,
+      "step": 238
+    },
+    {
+      "epoch": 0.10355987055016182,
+      "grad_norm": 2.046064615249634,
+      "learning_rate": 2.147349505840072e-05,
+      "loss": 0.9497957229614258,
+      "step": 240
+    },
+    {
+      "epoch": 0.10442286947141316,
+      "grad_norm": 2.1612143516540527,
+      "learning_rate": 2.165318957771788e-05,
+      "loss": 0.958332896232605,
+      "step": 242
+    },
+    {
+      "epoch": 0.1052858683926645,
+      "grad_norm": 2.139829635620117,
+      "learning_rate": 2.1832884097035043e-05,
+      "loss": 1.0089390277862549,
+      "step": 244
+    },
+    {
+      "epoch": 0.10614886731391586,
+      "grad_norm": 1.9842045307159424,
+      "learning_rate": 2.2012578616352204e-05,
+      "loss": 0.9627355933189392,
+      "step": 246
+    },
+    {
+      "epoch": 0.1070118662351672,
+      "grad_norm": 1.968578577041626,
+      "learning_rate": 2.2192273135669366e-05,
+      "loss": 0.9260450601577759,
+      "step": 248
+    },
+    {
+      "epoch": 0.10787486515641856,
+      "grad_norm": 2.192275047302246,
+      "learning_rate": 2.2371967654986524e-05,
+      "loss": 1.0654462575912476,
+      "step": 250
+    },
+    {
+      "epoch": 0.1087378640776699,
+      "grad_norm": 2.0267577171325684,
+      "learning_rate": 2.2551662174303682e-05,
+      "loss": 0.9680672883987427,
+      "step": 252
+    },
+    {
+      "epoch": 0.10960086299892124,
+      "grad_norm": 1.9834461212158203,
+      "learning_rate": 2.2731356693620844e-05,
+      "loss": 1.0027307271957397,
+      "step": 254
+    },
+    {
+      "epoch": 0.1104638619201726,
+      "grad_norm": 2.0105156898498535,
+      "learning_rate": 2.2911051212938006e-05,
+      "loss": 1.0050861835479736,
+      "step": 256
+    },
+    {
+      "epoch": 0.11132686084142394,
+      "grad_norm": 1.9150307178497314,
+      "learning_rate": 2.3090745732255167e-05,
+      "loss": 0.8955062627792358,
+      "step": 258
+    },
+    {
+      "epoch": 0.1121898597626753,
+      "grad_norm": 2.100795269012451,
+      "learning_rate": 2.327044025157233e-05,
+      "loss": 0.941332221031189,
+      "step": 260
+    },
+    {
+      "epoch": 0.11305285868392664,
+      "grad_norm": 1.837814211845398,
+      "learning_rate": 2.345013477088949e-05,
+      "loss": 0.9292652606964111,
+      "step": 262
+    },
+    {
+      "epoch": 0.113915857605178,
+      "grad_norm": 1.9122955799102783,
+      "learning_rate": 2.3629829290206652e-05,
+      "loss": 0.9460650682449341,
+      "step": 264
+    },
+    {
+      "epoch": 0.11477885652642934,
+      "grad_norm": 1.9168611764907837,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.8730320334434509,
+      "step": 266
+    },
+    {
+      "epoch": 0.11564185544768069,
+      "grad_norm": 2.1184403896331787,
+      "learning_rate": 2.3989218328840972e-05,
+      "loss": 0.9194066524505615,
+      "step": 268
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.910470962524414,
+      "learning_rate": 2.4168912848158133e-05,
+      "loss": 0.8994337320327759,
+      "step": 270
+    },
+    {
+      "epoch": 0.11736785329018339,
+      "grad_norm": 1.9906529188156128,
+      "learning_rate": 2.434860736747529e-05,
+      "loss": 0.8509554266929626,
+      "step": 272
+    },
+    {
+      "epoch": 0.11823085221143474,
+      "grad_norm": 2.5982062816619873,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 1.0223658084869385,
+      "step": 274
+    },
+    {
+      "epoch": 0.11909385113268608,
+      "grad_norm": 1.8959418535232544,
+      "learning_rate": 2.4707996406109615e-05,
+      "loss": 0.8702670931816101,
+      "step": 276
+    },
+    {
+      "epoch": 0.11995685005393743,
+      "grad_norm": 2.115907669067383,
+      "learning_rate": 2.4887690925426776e-05,
+      "loss": 0.9471842050552368,
+      "step": 278
+    },
+    {
+      "epoch": 0.12081984897518878,
+      "grad_norm": 2.0255353450775146,
+      "learning_rate": 2.5067385444743935e-05,
+      "loss": 0.805242657661438,
+      "step": 280
+    },
+    {
+      "epoch": 0.12168284789644013,
+      "grad_norm": 1.9394105672836304,
+      "learning_rate": 2.52470799640611e-05,
+      "loss": 0.9008285999298096,
+      "step": 282
+    },
+    {
+      "epoch": 0.12254584681769148,
+      "grad_norm": 2.2017428874969482,
+      "learning_rate": 2.5426774483378258e-05,
+      "loss": 0.950639009475708,
+      "step": 284
+    },
+    {
+      "epoch": 0.12340884573894283,
+      "grad_norm": 4.048379421234131,
+      "learning_rate": 2.5606469002695423e-05,
+      "loss": 0.9673396944999695,
+      "step": 286
+    },
+    {
+      "epoch": 0.12427184466019417,
+      "grad_norm": 1.812450647354126,
+      "learning_rate": 2.578616352201258e-05,
+      "loss": 0.8251164555549622,
+      "step": 288
+    },
+    {
+      "epoch": 0.12513484358144553,
+      "grad_norm": 1.9298596382141113,
+      "learning_rate": 2.596585804132974e-05,
+      "loss": 0.9397821426391602,
+      "step": 290
+    },
+    {
+      "epoch": 0.12599784250269688,
+      "grad_norm": 2.0862202644348145,
+      "learning_rate": 2.61455525606469e-05,
+      "loss": 0.95896315574646,
+      "step": 292
+    },
+    {
+      "epoch": 0.1268608414239482,
+      "grad_norm": 2.113429307937622,
+      "learning_rate": 2.632524707996406e-05,
+      "loss": 0.9547653198242188,
+      "step": 294
+    },
+    {
+      "epoch": 0.12772384034519957,
+      "grad_norm": 2.0666873455047607,
+      "learning_rate": 2.6504941599281224e-05,
+      "loss": 0.9351654052734375,
+      "step": 296
+    },
+    {
+      "epoch": 0.12858683926645093,
+      "grad_norm": 2.0679547786712646,
+      "learning_rate": 2.6684636118598382e-05,
+      "loss": 0.9268350005149841,
+      "step": 298
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 2.028594493865967,
+      "learning_rate": 2.6864330637915547e-05,
+      "loss": 0.8522316813468933,
+      "step": 300
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "eval_loss": 0.9384957551956177,
+      "eval_runtime": 835.4727,
+      "eval_samples_per_second": 2.466,
+      "eval_steps_per_second": 2.466,
+      "step": 300
+    },
+    {
+      "epoch": 0.1303128371089536,
+      "grad_norm": 2.1932919025421143,
+      "learning_rate": 2.7044025157232706e-05,
+      "loss": 0.9460103511810303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13117583603020497,
+      "grad_norm": 2.275789737701416,
+      "learning_rate": 2.722371967654987e-05,
+      "loss": 0.9642201662063599,
+      "step": 304
+    },
+    {
+      "epoch": 0.13203883495145632,
+      "grad_norm": 1.9120908975601196,
+      "learning_rate": 2.740341419586703e-05,
+      "loss": 0.869883120059967,
+      "step": 306
+    },
+    {
+      "epoch": 0.13290183387270765,
+      "grad_norm": 1.9676507711410522,
+      "learning_rate": 2.7583108715184187e-05,
+      "loss": 0.8595444560050964,
+      "step": 308
+    },
+    {
+      "epoch": 0.133764832793959,
+      "grad_norm": 1.7901391983032227,
+      "learning_rate": 2.776280323450135e-05,
+      "loss": 0.8541103601455688,
+      "step": 310
+    },
+    {
+      "epoch": 0.13462783171521037,
+      "grad_norm": 2.160867691040039,
+      "learning_rate": 2.7942497753818507e-05,
+      "loss": 1.0146855115890503,
+      "step": 312
+    },
+    {
+      "epoch": 0.1354908306364617,
+      "grad_norm": 2.0809009075164795,
+      "learning_rate": 2.8122192273135672e-05,
+      "loss": 0.9228161573410034,
+      "step": 314
+    },
+    {
+      "epoch": 0.13635382955771305,
+      "grad_norm": 1.9626725912094116,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.8527485132217407,
+      "step": 316
+    },
+    {
+      "epoch": 0.1372168284789644,
+      "grad_norm": 1.9569967985153198,
+      "learning_rate": 2.8481581311769995e-05,
+      "loss": 0.8136687278747559,
+      "step": 318
+    },
+    {
+      "epoch": 0.13807982740021574,
+      "grad_norm": 2.301692247390747,
+      "learning_rate": 2.8661275831087153e-05,
+      "loss": 0.8999513387680054,
+      "step": 320
+    },
+    {
+      "epoch": 0.1389428263214671,
+      "grad_norm": 2.0247251987457275,
+      "learning_rate": 2.884097035040431e-05,
+      "loss": 0.9751577973365784,
+      "step": 322
+    },
+    {
+      "epoch": 0.13980582524271845,
+      "grad_norm": 2.019547462463379,
+      "learning_rate": 2.9020664869721477e-05,
+      "loss": 0.8741729259490967,
+      "step": 324
+    },
+    {
+      "epoch": 0.1406688241639698,
+      "grad_norm": 2.3503618240356445,
+      "learning_rate": 2.9200359389038635e-05,
+      "loss": 0.948066234588623,
+      "step": 326
+    },
+    {
+      "epoch": 0.14153182308522114,
+      "grad_norm": 2.544365406036377,
+      "learning_rate": 2.9380053908355796e-05,
+      "loss": 1.008673906326294,
+      "step": 328
+    },
+    {
+      "epoch": 0.1423948220064725,
+      "grad_norm": 3.1647145748138428,
+      "learning_rate": 2.9559748427672958e-05,
+      "loss": 0.9492161273956299,
+      "step": 330
+    },
+    {
+      "epoch": 0.14325782092772385,
+      "grad_norm": 2.0551259517669678,
+      "learning_rate": 2.973944294699012e-05,
+      "loss": 0.9077466130256653,
+      "step": 332
+    },
+    {
+      "epoch": 0.14412081984897518,
+      "grad_norm": 1.6792620420455933,
+      "learning_rate": 2.9919137466307278e-05,
+      "loss": 0.8839989304542542,
+      "step": 334
+    },
+    {
+      "epoch": 0.14498381877022654,
+      "grad_norm": 1.7771215438842773,
+      "learning_rate": 3.0098831985624436e-05,
+      "loss": 0.855834424495697,
+      "step": 336
+    },
+    {
+      "epoch": 0.1458468176914779,
+      "grad_norm": 2.030884027481079,
+      "learning_rate": 3.02785265049416e-05,
+      "loss": 0.8477111458778381,
+      "step": 338
+    },
+    {
+      "epoch": 0.14670981661272922,
+      "grad_norm": 2.4177005290985107,
+      "learning_rate": 3.045822102425876e-05,
+      "loss": 0.9139004945755005,
+      "step": 340
+    },
+    {
+      "epoch": 0.14757281553398058,
+      "grad_norm": 2.094775438308716,
+      "learning_rate": 3.063791554357592e-05,
+      "loss": 0.8836029767990112,
+      "step": 342
+    },
+    {
+      "epoch": 0.14843581445523193,
+      "grad_norm": 2.0046238899230957,
+      "learning_rate": 3.081761006289308e-05,
+      "loss": 0.9318807125091553,
+      "step": 344
+    },
+    {
+      "epoch": 0.1492988133764833,
+      "grad_norm": 2.40739369392395,
+      "learning_rate": 3.0997304582210244e-05,
+      "loss": 0.8916489481925964,
+      "step": 346
+    },
+    {
+      "epoch": 0.15016181229773462,
+      "grad_norm": 2.1190497875213623,
+      "learning_rate": 3.11769991015274e-05,
+      "loss": 0.90766441822052,
+      "step": 348
+    },
+    {
+      "epoch": 0.15102481121898598,
+      "grad_norm": 2.721255302429199,
+      "learning_rate": 3.135669362084457e-05,
+      "loss": 0.9210623502731323,
+      "step": 350
+    },
+    {
+      "epoch": 0.15188781014023733,
+      "grad_norm": 2.4988787174224854,
+      "learning_rate": 3.1536388140161726e-05,
+      "loss": 0.9134353399276733,
+      "step": 352
+    },
+    {
+      "epoch": 0.15275080906148866,
+      "grad_norm": 2.6075687408447266,
+      "learning_rate": 3.1716082659478884e-05,
+      "loss": 0.9127986431121826,
+      "step": 354
+    },
+    {
+      "epoch": 0.15361380798274002,
+      "grad_norm": 1.9883074760437012,
+      "learning_rate": 3.189577717879605e-05,
+      "loss": 0.8019667863845825,
+      "step": 356
+    },
+    {
+      "epoch": 0.15447680690399138,
+      "grad_norm": 2.058591365814209,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.8417298197746277,
+      "step": 358
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 2.0510308742523193,
+      "learning_rate": 3.225516621743037e-05,
+      "loss": 0.909717321395874,
+      "step": 360
+    },
+    {
+      "epoch": 0.15620280474649406,
+      "grad_norm": 2.3204104900360107,
+      "learning_rate": 3.243486073674753e-05,
+      "loss": 0.9414666295051575,
+      "step": 362
+    },
+    {
+      "epoch": 0.15706580366774542,
+      "grad_norm": 1.9658725261688232,
+      "learning_rate": 3.2614555256064695e-05,
+      "loss": 0.8631396889686584,
+      "step": 364
+    },
+    {
+      "epoch": 0.15792880258899678,
+      "grad_norm": 2.2223057746887207,
+      "learning_rate": 3.2794249775381853e-05,
+      "loss": 0.8996705412864685,
+      "step": 366
+    },
+    {
+      "epoch": 0.1587918015102481,
+      "grad_norm": 1.9815107583999634,
+      "learning_rate": 3.297394429469901e-05,
+      "loss": 0.8185732364654541,
+      "step": 368
+    },
+    {
+      "epoch": 0.15965480043149946,
+      "grad_norm": 2.319581985473633,
+      "learning_rate": 3.315363881401618e-05,
+      "loss": 0.9952898025512695,
+      "step": 370
+    },
+    {
+      "epoch": 0.16051779935275082,
+      "grad_norm": 1.7916827201843262,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7604310512542725,
+      "step": 372
+    },
+    {
+      "epoch": 0.16138079827400215,
+      "grad_norm": 1.955737590789795,
+      "learning_rate": 3.35130278526505e-05,
+      "loss": 0.7867834568023682,
+      "step": 374
+    },
+    {
+      "epoch": 0.1622437971952535,
+      "grad_norm": 2.409783124923706,
+      "learning_rate": 3.369272237196766e-05,
+      "loss": 0.9106539487838745,
+      "step": 376
+    },
+    {
+      "epoch": 0.16310679611650486,
+      "grad_norm": 2.2521097660064697,
+      "learning_rate": 3.3872416891284816e-05,
+      "loss": 0.8945732712745667,
+      "step": 378
+    },
+    {
+      "epoch": 0.16396979503775622,
+      "grad_norm": 2.357815742492676,
+      "learning_rate": 3.4052111410601975e-05,
+      "loss": 0.8950189352035522,
+      "step": 380
+    },
+    {
+      "epoch": 0.16483279395900755,
+      "grad_norm": 2.347301721572876,
+      "learning_rate": 3.423180592991914e-05,
+      "loss": 0.9965803623199463,
+      "step": 382
+    },
+    {
+      "epoch": 0.1656957928802589,
+      "grad_norm": 2.054586887359619,
+      "learning_rate": 3.44115004492363e-05,
+      "loss": 0.9081383347511292,
+      "step": 384
+    },
+    {
+      "epoch": 0.16655879180151026,
+      "grad_norm": 2.1550471782684326,
+      "learning_rate": 3.4591194968553456e-05,
+      "loss": 0.874001681804657,
+      "step": 386
+    },
+    {
+      "epoch": 0.1674217907227616,
+      "grad_norm": 1.6929951906204224,
+      "learning_rate": 3.477088948787062e-05,
+      "loss": 0.9111959934234619,
+      "step": 388
+    },
+    {
+      "epoch": 0.16828478964401294,
+      "grad_norm": 2.0102803707122803,
+      "learning_rate": 3.495058400718778e-05,
+      "loss": 0.8915985226631165,
+      "step": 390
+    },
+    {
+      "epoch": 0.1691477885652643,
+      "grad_norm": 4.215941429138184,
+      "learning_rate": 3.5130278526504944e-05,
+      "loss": 1.0158213376998901,
+      "step": 392
+    },
+    {
+      "epoch": 0.17001078748651563,
+      "grad_norm": 1.864119529724121,
+      "learning_rate": 3.53099730458221e-05,
+      "loss": 0.7811702489852905,
+      "step": 394
+    },
+    {
+      "epoch": 0.170873786407767,
+      "grad_norm": 1.9778105020523071,
+      "learning_rate": 3.548966756513927e-05,
+      "loss": 0.931502103805542,
+      "step": 396
+    },
+    {
+      "epoch": 0.17173678532901834,
+      "grad_norm": 2.0757675170898438,
+      "learning_rate": 3.5669362084456426e-05,
+      "loss": 0.9409074187278748,
+      "step": 398
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "grad_norm": 1.5847891569137573,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.8159269094467163,
+      "step": 400
+    },
+    {
+      "epoch": 0.1725997842502697,
+      "eval_loss": 0.9124138355255127,
+      "eval_runtime": 832.6331,
+      "eval_samples_per_second": 2.474,
+      "eval_steps_per_second": 2.474,
+      "step": 400
+    },
+    {
+      "epoch": 0.17346278317152103,
+      "grad_norm": 1.8158049583435059,
+      "learning_rate": 3.602875112309075e-05,
+      "loss": 0.8901475667953491,
+      "step": 402
+    },
+    {
+      "epoch": 0.17432578209277239,
+      "grad_norm": 1.9448504447937012,
+      "learning_rate": 3.620844564240791e-05,
+      "loss": 0.8330907821655273,
+      "step": 404
+    },
+    {
+      "epoch": 0.17518878101402374,
+      "grad_norm": 2.028280019760132,
+      "learning_rate": 3.638814016172507e-05,
+      "loss": 0.870513916015625,
+      "step": 406
+    },
+    {
+      "epoch": 0.17605177993527507,
+      "grad_norm": 2.2005157470703125,
+      "learning_rate": 3.656783468104223e-05,
+      "loss": 0.8064979910850525,
+      "step": 408
+    },
+    {
+      "epoch": 0.17691477885652643,
+      "grad_norm": 2.2756290435791016,
+      "learning_rate": 3.6747529200359395e-05,
+      "loss": 0.8457443714141846,
+      "step": 410
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.0970261096954346,
+      "learning_rate": 3.6927223719676554e-05,
+      "loss": 0.7478897571563721,
+      "step": 412
+    },
+    {
+      "epoch": 0.1786407766990291,
+      "grad_norm": 1.8917485475540161,
+      "learning_rate": 3.710691823899371e-05,
+      "loss": 0.863377034664154,
+      "step": 414
+    },
+    {
+      "epoch": 0.17950377562028047,
+      "grad_norm": 2.0429933071136475,
+      "learning_rate": 3.728661275831088e-05,
+      "loss": 0.8654620051383972,
+      "step": 416
+    },
+    {
+      "epoch": 0.18036677454153183,
+      "grad_norm": 2.005125045776367,
+      "learning_rate": 3.7466307277628035e-05,
+      "loss": 0.8295455574989319,
+      "step": 418
+    },
+    {
+      "epoch": 0.18122977346278318,
+      "grad_norm": 2.3852169513702393,
+      "learning_rate": 3.764600179694519e-05,
+      "loss": 0.8728633522987366,
+      "step": 420
+    },
+    {
+      "epoch": 0.1820927723840345,
+      "grad_norm": 1.8731592893600464,
+      "learning_rate": 3.782569631626235e-05,
+      "loss": 0.8217283487319946,
+      "step": 422
+    },
+    {
+      "epoch": 0.18295577130528587,
+      "grad_norm": 2.2818236351013184,
+      "learning_rate": 3.8005390835579516e-05,
+      "loss": 0.8869860172271729,
+      "step": 424
+    },
+    {
+      "epoch": 0.18381877022653723,
+      "grad_norm": 2.0646233558654785,
+      "learning_rate": 3.8185085354896675e-05,
+      "loss": 0.9732091426849365,
+      "step": 426
+    },
+    {
+      "epoch": 0.18468176914778855,
+      "grad_norm": 1.9991949796676636,
+      "learning_rate": 3.836477987421384e-05,
+      "loss": 0.9012333154678345,
+      "step": 428
+    },
+    {
+      "epoch": 0.1855447680690399,
+      "grad_norm": 2.209775447845459,
+      "learning_rate": 3.8544474393531e-05,
+      "loss": 0.867533802986145,
+      "step": 430
+    },
+    {
+      "epoch": 0.18640776699029127,
+      "grad_norm": 2.1242146492004395,
+      "learning_rate": 3.8724168912848156e-05,
+      "loss": 0.8697446584701538,
+      "step": 432
+    },
+    {
+      "epoch": 0.1872707659115426,
+      "grad_norm": 1.771337628364563,
+      "learning_rate": 3.890386343216532e-05,
+      "loss": 0.8712933659553528,
+      "step": 434
+    },
+    {
+      "epoch": 0.18813376483279395,
+      "grad_norm": 2.133652925491333,
+      "learning_rate": 3.908355795148248e-05,
+      "loss": 0.8638070821762085,
+      "step": 436
+    },
+    {
+      "epoch": 0.1889967637540453,
+      "grad_norm": 1.9354501962661743,
+      "learning_rate": 3.9263252470799644e-05,
+      "loss": 0.8255205750465393,
+      "step": 438
+    },
+    {
+      "epoch": 0.18985976267529667,
+      "grad_norm": 1.9564658403396606,
+      "learning_rate": 3.94429469901168e-05,
+      "loss": 0.9652490019798279,
+      "step": 440
+    },
+    {
+      "epoch": 0.190722761596548,
+      "grad_norm": 1.9733822345733643,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.8960518836975098,
+      "step": 442
+    },
+    {
+      "epoch": 0.19158576051779935,
+      "grad_norm": 2.1670451164245605,
+      "learning_rate": 3.9802336028751126e-05,
+      "loss": 0.9555241465568542,
+      "step": 444
+    },
+    {
+      "epoch": 0.1924487594390507,
+      "grad_norm": 2.0059614181518555,
+      "learning_rate": 3.9982030548068284e-05,
+      "loss": 0.8238654732704163,
+      "step": 446
+    },
+    {
+      "epoch": 0.19331175836030204,
+      "grad_norm": 1.9803026914596558,
+      "learning_rate": 4.016172506738545e-05,
+      "loss": 0.8701026439666748,
+      "step": 448
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.8800172805786133,
+      "learning_rate": 4.034141958670261e-05,
+      "loss": 0.8927714228630066,
+      "step": 450
+    },
+    {
+      "epoch": 0.19503775620280475,
+      "grad_norm": 1.8555818796157837,
+      "learning_rate": 4.052111410601977e-05,
+      "loss": 0.8635751008987427,
+      "step": 452
+    },
+    {
+      "epoch": 0.1959007551240561,
+      "grad_norm": 2.5119166374206543,
+      "learning_rate": 4.070080862533693e-05,
+      "loss": 0.925403892993927,
+      "step": 454
+    },
+    {
+      "epoch": 0.19676375404530744,
+      "grad_norm": 2.0825788974761963,
+      "learning_rate": 4.088050314465409e-05,
+      "loss": 0.8917028307914734,
+      "step": 456
+    },
+    {
+      "epoch": 0.1976267529665588,
+      "grad_norm": 2.4413394927978516,
+      "learning_rate": 4.1060197663971254e-05,
+      "loss": 0.9097893834114075,
+      "step": 458
+    },
+    {
+      "epoch": 0.19848975188781015,
+      "grad_norm": 2.503617286682129,
+      "learning_rate": 4.123989218328841e-05,
+      "loss": 0.8284633755683899,
+      "step": 460
+    },
+    {
+      "epoch": 0.19935275080906148,
+      "grad_norm": 1.9360398054122925,
+      "learning_rate": 4.141958670260557e-05,
+      "loss": 0.8743083477020264,
+      "step": 462
+    },
+    {
+      "epoch": 0.20021574973031284,
+      "grad_norm": 1.9228297472000122,
+      "learning_rate": 4.159928122192273e-05,
+      "loss": 0.837934136390686,
+      "step": 464
+    },
+    {
+      "epoch": 0.2010787486515642,
+      "grad_norm": 2.3240439891815186,
+      "learning_rate": 4.1778975741239893e-05,
+      "loss": 0.831945538520813,
+      "step": 466
+    },
+    {
+      "epoch": 0.20194174757281552,
+      "grad_norm": 1.879684329032898,
+      "learning_rate": 4.195867026055705e-05,
+      "loss": 0.8012097477912903,
+      "step": 468
+    },
+    {
+      "epoch": 0.20280474649406688,
+      "grad_norm": 2.2995200157165527,
+      "learning_rate": 4.213836477987422e-05,
+      "loss": 0.8670937418937683,
+      "step": 470
+    },
+    {
+      "epoch": 0.20366774541531824,
+      "grad_norm": 2.098020553588867,
+      "learning_rate": 4.2318059299191375e-05,
+      "loss": 0.88299161195755,
+      "step": 472
+    },
+    {
+      "epoch": 0.2045307443365696,
+      "grad_norm": 2.0588274002075195,
+      "learning_rate": 4.249775381850854e-05,
+      "loss": 0.9045401811599731,
+      "step": 474
+    },
+    {
+      "epoch": 0.20539374325782092,
+      "grad_norm": 1.8832436800003052,
+      "learning_rate": 4.26774483378257e-05,
+      "loss": 0.9787635803222656,
+      "step": 476
+    },
+    {
+      "epoch": 0.20625674217907228,
+      "grad_norm": 2.3451809883117676,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.8874871730804443,
+      "step": 478
+    },
+    {
+      "epoch": 0.20711974110032363,
+      "grad_norm": 1.8093878030776978,
+      "learning_rate": 4.303683737646002e-05,
+      "loss": 0.794157087802887,
+      "step": 480
+    },
+    {
+      "epoch": 0.20798274002157496,
+      "grad_norm": 1.8396174907684326,
+      "learning_rate": 4.321653189577718e-05,
+      "loss": 0.792723536491394,
+      "step": 482
+    },
+    {
+      "epoch": 0.20884573894282632,
+      "grad_norm": 2.3939783573150635,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.9221814274787903,
+      "step": 484
+    },
+    {
+      "epoch": 0.20970873786407768,
+      "grad_norm": 2.2652194499969482,
+      "learning_rate": 4.35759209344115e-05,
+      "loss": 0.8716039061546326,
+      "step": 486
+    },
+    {
+      "epoch": 0.210571736785329,
+      "grad_norm": 2.068258762359619,
+      "learning_rate": 4.375561545372867e-05,
+      "loss": 0.8517807126045227,
+      "step": 488
+    },
+    {
+      "epoch": 0.21143473570658036,
+      "grad_norm": 2.264620542526245,
+      "learning_rate": 4.3935309973045826e-05,
+      "loss": 0.8699264526367188,
+      "step": 490
+    },
+    {
+      "epoch": 0.21229773462783172,
+      "grad_norm": 2.0821611881256104,
+      "learning_rate": 4.4115004492362984e-05,
+      "loss": 0.8760254383087158,
+      "step": 492
+    },
+    {
+      "epoch": 0.21316073354908308,
+      "grad_norm": 1.9486546516418457,
+      "learning_rate": 4.429469901168015e-05,
+      "loss": 0.8957411050796509,
+      "step": 494
+    },
+    {
+      "epoch": 0.2140237324703344,
+      "grad_norm": 2.092957019805908,
+      "learning_rate": 4.447439353099731e-05,
+      "loss": 0.981200635433197,
+      "step": 496
+    },
+    {
+      "epoch": 0.21488673139158576,
+      "grad_norm": 2.325758218765259,
+      "learning_rate": 4.4654088050314466e-05,
+      "loss": 0.8295405507087708,
+      "step": 498
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "grad_norm": 2.251295566558838,
+      "learning_rate": 4.4833782569631624e-05,
+      "loss": 0.8463165163993835,
+      "step": 500
+    },
+    {
+      "epoch": 0.21574973031283712,
+      "eval_loss": 0.8945327997207642,
+      "eval_runtime": 822.9776,
+      "eval_samples_per_second": 2.503,
+      "eval_steps_per_second": 2.503,
+      "step": 500
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 13908,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.2025749846072934e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/sft_devstral_24B/checkpoints/checkpoint-500/training_args.bin b/sft_devstral_24B/checkpoints/checkpoint-500/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c82ad9bf1473ea8d701f8ee9e7d69dc1ce34184b
--- /dev/null
+++ b/sft_devstral_24B/checkpoints/checkpoint-500/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67cec3db40d9b4bcc81f7a80aeef5218639d2e42b1c9e2f551945abddeb1ff4d
+size 5201
diff --git a/sft_devstral_24B/config_instruct.yaml b/sft_devstral_24B/config_instruct.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..351713220a38459b826e0992877249f3a921d71e
--- /dev/null
+++ b/sft_devstral_24B/config_instruct.yaml
@@ -0,0 +1,144 @@
+run:
+  run_dir: "./runs/instruct_run_24b"
+  seed: 42
+
+# WandB integration for experiment tracking
+wandb:
+  enabled: true  # Set to true to enable wandb logging
+  project: "sft-training"  # WandB project name
+  entity: null  # WandB entity/team (optional)
+  name: null  # Run name (optional, will auto-generate if null)
+  tags: ["sft-lora", "24b-Devstral"]  # List of tags for the run (e.g., ["lora", "qlora", "experiment-1"])
+  notes: null  # Run description/notes (optional)
+
+model:
+  # Use local Qwen2.5-Coder-14B model
+  repo_id: "./CPT/runs/cpt_run_v1/merged_24b_cpt_lora"
+  revision: null
+
+  # Used only when repo_id is a HF repo (not a local path)
+  base_local_dir: "base_model"
+
+  trust_remote_code: true
+  tokenizer_use_fast: true
+  device_map: "auto"
+
+  torch_dtype: "bfloat16"  # "float16" | "bfloat16" | "float32"
+
+  # QLoRA
+  use_4bit: false
+  bnb_4bit_quant_type: "nf4"
+  bnb_4bit_use_double_quant: false
+  bnb_4bit_compute_dtype: "bfloat16"
+
+  # optional: "flash_attention_2" | "sdpa" | null
+  attn_implementation: null
+
+data:
+  train_jsonl: "../sft_dataset.jsonl"
+  eval_jsonl: null
+  eval_split_ratio: 0.1
+  
+  # Field names in your JSONL data
+  instruction_field: "instruction"  # This will be the system prompt
+  input_field: "input"             # This is the task description
+  output_field: "output"           # This is the analysis + selection
+  
+  # Formatting options
+  format_type: "custom"  # "chatml" | "alpaca" | "custom"
+  
+  # For chatml format
+  system_prompt: |
+    You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.
+
+    ## Output Format
+
+    ##OUTPUT
+    Explain the data flow and why each component must change:
+    - Flow: [Input → Processing → Output with arrows]
+    - For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"
+    - Explain coupling between components
+
+    ##SELECT
+    modify::crates/path/to/file.rs::impl::ComponentName
+    add::crates/another/file.rs::function::AnotherComponent
+    <EOS>
+
+    ## Rules
+
+    1. Use full paths: `remove::crates/folder/file.rs::Type::Name`
+    2. Use `::` for nested items: `status::StructName::Type::Name`
+    3. Always explain "must change because" and "without this"
+    3. Types of components: function, struct, enum, impl, trait
+    4. If there is extra information (e.g., enum variants), include that too.
+    5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>
+
+    ## Example
+
+    ##TASK
+    Add webhook subscription support
+
+    ##OUTPUT
+    The webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don't trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.
+
+    ##SELECT
+    crates/common_enums/src/enums.rs::EventClass
+    crates/common_enums/src/transformers.rs::SubscriptionStatus
+    <EOS>
+  
+  # For custom format (only used when format_type="custom")
+  custom_template: "##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>"
+  
+  max_length: 2048
+  shuffle: true
+  num_proc: 4
+
+peft:
+  enabled: true
+  r: 8
+  lora_alpha: 16
+  lora_dropout: 0.05
+  bias: "none"
+  target_modules: "auto"
+
+train:
+  # max_steps: 10
+  num_train_epochs: 6
+
+  per_device_train_batch_size: 1
+  per_device_eval_batch_size: 1
+  gradient_accumulation_steps: 8
+
+  learning_rate: 1e-4
+  weight_decay: 0.0
+  warmup_ratio: 0.08
+  lr_scheduler_type: "cosine"
+
+  optim: "adamw_torch"  # ✅ Changed from paged_adamw_8bit (requires use_4bit=true)
+  max_grad_norm: 0.8
+  gradient_checkpointing: true
+
+  logging_steps: 2
+  save_strategy: "steps"
+  save_steps: 500
+  save_total_limit: 20
+
+  evaluation_strategy: "steps"
+  eval_steps: 100
+  load_best_model_at_end: true
+
+  # Early stopping
+  early_stopping:
+    enabled: true
+    patience: 3  # Number of evaluations with no improvement before stopping
+    min_delta: 0.001  # Minimum change to qualify as improvement
+    metric: "eval_loss"  # Metric to monitor
+    mode: "min"  # "min" for loss, "max" for accuracy/etc.
+
+  resume_from_checkpoint: "auto"
+
+merge:
+  enabled: true
+  merged_dtype: "float16"
+  max_shard_size: "2GB"
+  output_dir: "./merged_24b_instruct_lora"
\ No newline at end of file
diff --git a/sft_devstral_24B/config_resolved.yaml b/sft_devstral_24B/config_resolved.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..b933dbb3047a3ee5aabc27debe642b12804f77eb
--- /dev/null
+++ b/sft_devstral_24B/config_resolved.yaml
@@ -0,0 +1,106 @@
+run:
+  run_dir: ./runs/instruct_run_24b
+  seed: 42
+wandb:
+  enabled: true
+  project: sft-training
+  entity: null
+  name: null
+  tags:
+  - sft-lora
+  - 24b-Devstral
+  notes: null
+model:
+  repo_id: ./CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+  revision: null
+  base_local_dir: base_model
+  trust_remote_code: true
+  tokenizer_use_fast: true
+  device_map: auto
+  torch_dtype: bfloat16
+  use_4bit: false
+  bnb_4bit_quant_type: nf4
+  bnb_4bit_use_double_quant: false
+  bnb_4bit_compute_dtype: bfloat16
+  attn_implementation: null
+data:
+  train_jsonl: ../sft_dataset.jsonl
+  eval_jsonl: null
+  eval_split_ratio: 0.1
+  instruction_field: instruction
+  input_field: input
+  output_field: output
+  format_type: custom
+  system_prompt: "You are a Hyperswitch Rust code analyzer. Identify functions/structs\
+    \ that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain\
+    \ the data flow and why each component must change:\n- Flow: [Input \u2192 Processing\
+    \ \u2192 Output with arrows]\n- For each component: \"The [ComponentName] ([path])\
+    \ must [action] because [reason]\u2014without this, [consequence]\"\n- Explain\
+    \ coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\n\
+    add::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n\
+    1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for\
+    \ nested items: `status::StructName::Type::Name`\n3. Always explain \"must change\
+    \ because\" and \"without this\"\n3. Types of components: function, struct, enum,\
+    \ impl, trait\n4. If there is extra information (e.g., enum variants), include\
+    \ that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\
+    \n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook\
+    \ system routes events via EventClass enum. Flow: webhook \u2192 EventClass \u2192\
+    \ handler \u2192 processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass)\
+    \ must add Subscriptions variant because it defines event routing\u2014without\
+    \ this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus)\
+    \ must map to EventType because it converts status to events\u2014without this,\
+    \ status changes don't trigger webhooks. These are coupled: EventClass routes\
+    \ to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\n\
+    crates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n"
+  custom_template: '##INSTRUCTION
+
+    {instruction}<|im_end|>
+
+    ##TASK
+
+    {input}<|im_end|>
+
+    ##OUTPUT
+
+    {output}<|im_end|>'
+  max_length: 2048
+  shuffle: true
+  num_proc: 4
+peft:
+  enabled: true
+  r: 8
+  lora_alpha: 16
+  lora_dropout: 0.05
+  bias: none
+  target_modules: auto
+train:
+  num_train_epochs: 6
+  per_device_train_batch_size: 1
+  per_device_eval_batch_size: 1
+  gradient_accumulation_steps: 8
+  learning_rate: 1e-4
+  weight_decay: 0.0
+  warmup_ratio: 0.08
+  lr_scheduler_type: cosine
+  optim: adamw_torch
+  max_grad_norm: 0.8
+  gradient_checkpointing: true
+  logging_steps: 2
+  save_strategy: steps
+  save_steps: 500
+  save_total_limit: 20
+  evaluation_strategy: steps
+  eval_steps: 100
+  load_best_model_at_end: true
+  early_stopping:
+    enabled: true
+    patience: 3
+    min_delta: 0.001
+    metric: eval_loss
+    mode: min
+  resume_from_checkpoint: auto
+merge:
+  enabled: true
+  merged_dtype: float16
+  max_shard_size: 2GB
+  output_dir: ./merged_24b_instruct_lora
diff --git a/sft_devstral_24B/eval_final.json b/sft_devstral_24B/eval_final.json
new file mode 100644
index 0000000000000000000000000000000000000000..c0f4111c93acdc4cbd2240c9f2d90507cfb6e92b
--- /dev/null
+++ b/sft_devstral_24B/eval_final.json
@@ -0,0 +1,8 @@
+{
+  "eval_loss": 0.6851214170455933,
+  "eval_runtime": 497.0682,
+  "eval_samples_per_second": 4.144,
+  "eval_steps_per_second": 4.144,
+  "epoch": 2.0707659115426105,
+  "perplexity": 1.9840127138756343
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/logs/eval.jsonl b/sft_devstral_24B/logs/eval.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..5cde06a62f8f6a2993ab80e1a0bf7a8b0827e390
--- /dev/null
+++ b/sft_devstral_24B/logs/eval.jsonl
@@ -0,0 +1,50 @@
+{"ts": "2025-12-23T14:16:36", "event": "eval", "step": 100, "epoch": 0.043149946062567425, "eval_loss": 0.9360476136207581, "eval_runtime": 834.9, "eval_samples_per_second": 2.467, "eval_steps_per_second": 2.467, "perplexity": 2.5498833515166743}
+{"ts": "2025-12-23T14:51:03", "event": "eval", "step": 100, "epoch": 0.043149946062567425, "eval_loss": 1.128923773765564, "eval_runtime": 825.0166, "eval_samples_per_second": 2.497, "eval_steps_per_second": 2.497, "perplexity": 3.0923266655355475}
+{"ts": "2025-12-23T15:20:59", "event": "eval", "step": 200, "epoch": 0.08629989212513485, "eval_loss": 0.981117308139801, "eval_runtime": 826.4469, "eval_samples_per_second": 2.493, "eval_steps_per_second": 2.493, "perplexity": 2.6674349243198323}
+{"ts": "2025-12-23T15:50:45", "event": "eval", "step": 300, "epoch": 0.12944983818770225, "eval_loss": 0.9384957551956177, "eval_runtime": 835.4727, "eval_samples_per_second": 2.466, "eval_steps_per_second": 2.466, "perplexity": 2.556133474431821}
+{"ts": "2025-12-23T16:20:43", "event": "eval", "step": 400, "epoch": 0.1725997842502697, "eval_loss": 0.9124138355255127, "eval_runtime": 832.6331, "eval_samples_per_second": 2.474, "eval_steps_per_second": 2.474, "perplexity": 2.490326522778942}
+{"ts": "2025-12-23T16:50:41", "event": "eval", "step": 500, "epoch": 0.21574973031283712, "eval_loss": 0.8945327997207642, "eval_runtime": 822.9776, "eval_samples_per_second": 2.503, "eval_steps_per_second": 2.503, "perplexity": 2.4461926605539563}
+{"ts": "2025-12-23T17:20:42", "event": "eval", "step": 600, "epoch": 0.2588996763754045, "eval_loss": 0.8770867586135864, "eval_runtime": 827.8893, "eval_samples_per_second": 2.488, "eval_steps_per_second": 2.488, "perplexity": 2.4038863943096764}
+{"ts": "2025-12-23T17:50:37", "event": "eval", "step": 700, "epoch": 0.30204962243797195, "eval_loss": 0.8631294369697571, "eval_runtime": 839.6269, "eval_samples_per_second": 2.453, "eval_steps_per_second": 2.453, "perplexity": 2.3705676398868443}
+{"ts": "2025-12-23T18:20:39", "event": "eval", "step": 800, "epoch": 0.3451995685005394, "eval_loss": 0.8476243615150452, "eval_runtime": 831.6735, "eval_samples_per_second": 2.477, "eval_steps_per_second": 2.477, "perplexity": 2.3340952936819845}
+{"ts": "2025-12-23T18:50:33", "event": "eval", "step": 900, "epoch": 0.3883495145631068, "eval_loss": 0.8351938128471375, "eval_runtime": 823.5649, "eval_samples_per_second": 2.501, "eval_steps_per_second": 2.501, "perplexity": 2.305260794150979}
+{"ts": "2025-12-23T19:20:44", "event": "eval", "step": 1000, "epoch": 0.43149946062567424, "eval_loss": 0.8266019821166992, "eval_runtime": 835.0044, "eval_samples_per_second": 2.467, "eval_steps_per_second": 2.467, "perplexity": 2.2855392271205766}
+{"ts": "2025-12-23T19:50:37", "event": "eval", "step": 1100, "epoch": 0.4746494066882416, "eval_loss": 0.8152187466621399, "eval_runtime": 835.0394, "eval_samples_per_second": 2.467, "eval_steps_per_second": 2.467, "perplexity": 2.2596699135420506}
+{"ts": "2025-12-23T20:20:32", "event": "eval", "step": 1200, "epoch": 0.517799352750809, "eval_loss": 0.8070361018180847, "eval_runtime": 823.9296, "eval_samples_per_second": 2.5, "eval_steps_per_second": 2.5, "perplexity": 2.24125528012424}
+{"ts": "2025-12-23T20:50:45", "event": "eval", "step": 1300, "epoch": 0.5609492988133765, "eval_loss": 0.7962777018547058, "eval_runtime": 827.8798, "eval_samples_per_second": 2.488, "eval_steps_per_second": 2.488, "perplexity": 2.2172722005047705}
+{"ts": "2025-12-23T21:20:37", "event": "eval", "step": 1400, "epoch": 0.6040992448759439, "eval_loss": 0.7866972088813782, "eval_runtime": 836.0759, "eval_samples_per_second": 2.464, "eval_steps_per_second": 2.464, "perplexity": 2.1961310726855836}
+{"ts": "2025-12-23T21:50:35", "event": "eval", "step": 1500, "epoch": 0.6472491909385113, "eval_loss": 0.778505802154541, "eval_runtime": 836.2317, "eval_samples_per_second": 2.463, "eval_steps_per_second": 2.463, "perplexity": 2.1782151483338836}
+{"ts": "2025-12-23T22:20:28", "event": "eval", "step": 1600, "epoch": 0.6903991370010788, "eval_loss": 0.7706249952316284, "eval_runtime": 821.4894, "eval_samples_per_second": 2.508, "eval_steps_per_second": 2.508, "perplexity": 2.1611165193057653}
+{"ts": "2025-12-23T22:50:30", "event": "eval", "step": 1700, "epoch": 0.7335490830636462, "eval_loss": 0.7649155259132385, "eval_runtime": 828.3542, "eval_samples_per_second": 2.487, "eval_steps_per_second": 2.487, "perplexity": 2.1488128479852233}
+{"ts": "2025-12-23T23:20:32", "event": "eval", "step": 1800, "epoch": 0.7766990291262136, "eval_loss": 0.7598537802696228, "eval_runtime": 838.9645, "eval_samples_per_second": 2.455, "eval_steps_per_second": 2.455, "perplexity": 2.1379635851816685}
+{"ts": "2025-12-23T23:50:29", "event": "eval", "step": 1900, "epoch": 0.819848975188781, "eval_loss": 0.7533769607543945, "eval_runtime": 830.2894, "eval_samples_per_second": 2.481, "eval_steps_per_second": 2.481, "perplexity": 2.124161127175132}
+{"ts": "2025-12-24T00:20:24", "event": "eval", "step": 2000, "epoch": 0.8629989212513485, "eval_loss": 0.7487027645111084, "eval_runtime": 825.141, "eval_samples_per_second": 2.497, "eval_steps_per_second": 2.497, "perplexity": 2.114255549556043}
+{"ts": "2025-12-24T00:50:25", "event": "eval", "step": 2100, "epoch": 0.9061488673139159, "eval_loss": 0.7449111342430115, "eval_runtime": 832.5135, "eval_samples_per_second": 2.474, "eval_steps_per_second": 2.474, "perplexity": 2.1062542527852224}
+{"ts": "2025-12-24T01:20:20", "event": "eval", "step": 2200, "epoch": 0.9492988133764833, "eval_loss": 0.7400083541870117, "eval_runtime": 838.7633, "eval_samples_per_second": 2.456, "eval_steps_per_second": 2.456, "perplexity": 2.0959530244047575}
+{"ts": "2025-12-24T01:50:21", "event": "eval", "step": 2300, "epoch": 0.9924487594390508, "eval_loss": 0.7360134720802307, "eval_runtime": 825.8559, "eval_samples_per_second": 2.494, "eval_steps_per_second": 2.494, "perplexity": 2.087596641666197}
+{"ts": "2025-12-24T02:20:14", "event": "eval", "step": 2400, "epoch": 1.0353829557713052, "eval_loss": 0.7346783876419067, "eval_runtime": 825.8356, "eval_samples_per_second": 2.494, "eval_steps_per_second": 2.494, "perplexity": 2.0848113835674984}
+{"ts": "2025-12-24T02:50:23", "event": "eval", "step": 2500, "epoch": 1.0785329018338727, "eval_loss": 0.7294915914535522, "eval_runtime": 836.4643, "eval_samples_per_second": 2.463, "eval_steps_per_second": 2.463, "perplexity": 2.07402588709581}
+{"ts": "2025-12-24T03:20:08", "event": "eval", "step": 2600, "epoch": 1.12168284789644, "eval_loss": 0.7287447452545166, "eval_runtime": 831.8138, "eval_samples_per_second": 2.477, "eval_steps_per_second": 2.477, "perplexity": 2.0724774870256564}
+{"ts": "2025-12-24T03:49:59", "event": "eval", "step": 2700, "epoch": 1.1648327939590075, "eval_loss": 0.72404944896698, "eval_runtime": 824.2877, "eval_samples_per_second": 2.499, "eval_steps_per_second": 2.499, "perplexity": 2.0627694001820904}
+{"ts": "2025-12-24T04:20:03", "event": "eval", "step": 2800, "epoch": 1.207982740021575, "eval_loss": 0.7220398187637329, "eval_runtime": 831.5812, "eval_samples_per_second": 2.477, "eval_steps_per_second": 2.477, "perplexity": 2.0586281590685003}
+{"ts": "2025-12-24T04:50:12", "event": "eval", "step": 2900, "epoch": 1.2511326860841425, "eval_loss": 0.7182289958000183, "eval_runtime": 840.8156, "eval_samples_per_second": 2.45, "eval_steps_per_second": 2.45, "perplexity": 2.0507980207177035}
+{"ts": "2025-12-24T05:20:02", "event": "eval", "step": 3000, "epoch": 1.2942826321467098, "eval_loss": 0.7160727977752686, "eval_runtime": 830.8186, "eval_samples_per_second": 2.479, "eval_steps_per_second": 2.479, "perplexity": 2.046380857926466}
+{"ts": "2025-12-24T05:49:53", "event": "eval", "step": 3100, "epoch": 1.3374325782092773, "eval_loss": 0.7127581834793091, "eval_runtime": 821.1346, "eval_samples_per_second": 2.509, "eval_steps_per_second": 2.509, "perplexity": 2.0396091237232032}
+{"ts": "2025-12-24T06:19:57", "event": "eval", "step": 3200, "epoch": 1.3805825242718446, "eval_loss": 0.7103806734085083, "eval_runtime": 831.4047, "eval_samples_per_second": 2.478, "eval_steps_per_second": 2.478, "perplexity": 2.034765692425877}
+{"ts": "2025-12-24T06:49:54", "event": "eval", "step": 3300, "epoch": 1.423732470334412, "eval_loss": 0.7086145281791687, "eval_runtime": 838.887, "eval_samples_per_second": 2.456, "eval_steps_per_second": 2.456, "perplexity": 2.0311751723286617}
+{"ts": "2025-12-24T07:19:48", "event": "eval", "step": 3400, "epoch": 1.4668824163969796, "eval_loss": 0.7067069411277771, "eval_runtime": 829.9411, "eval_samples_per_second": 2.482, "eval_steps_per_second": 2.482, "perplexity": 2.027304222131885}
+{"ts": "2025-12-24T07:49:43", "event": "eval", "step": 3500, "epoch": 1.510032362459547, "eval_loss": 0.7047909498214722, "eval_runtime": 821.8811, "eval_samples_per_second": 2.506, "eval_steps_per_second": 2.506, "perplexity": 2.023423643631521}
+{"ts": "2025-12-24T08:19:44", "event": "eval", "step": 3600, "epoch": 1.5531823085221144, "eval_loss": 0.701831042766571, "eval_runtime": 836.0199, "eval_samples_per_second": 2.464, "eval_steps_per_second": 2.464, "perplexity": 2.0174433526325766}
+{"ts": "2025-12-24T08:49:46", "event": "eval", "step": 3700, "epoch": 1.5963322545846816, "eval_loss": 0.6999035477638245, "eval_runtime": 836.4944, "eval_samples_per_second": 2.463, "eval_steps_per_second": 2.463, "perplexity": 2.01355848588544}
+{"ts": "2025-12-24T09:19:47", "event": "eval", "step": 3800, "epoch": 1.6394822006472491, "eval_loss": 0.6974969506263733, "eval_runtime": 823.7017, "eval_samples_per_second": 2.501, "eval_steps_per_second": 2.501, "perplexity": 2.008718488095619}
+{"ts": "2025-12-24T09:49:37", "event": "eval", "step": 3900, "epoch": 1.6826321467098166, "eval_loss": 0.6949095726013184, "eval_runtime": 817.4497, "eval_samples_per_second": 2.52, "eval_steps_per_second": 2.52, "perplexity": 2.003527891933936}
+{"ts": "2025-12-24T10:10:24", "event": "eval", "step": 4000, "epoch": 1.7257820927723841, "eval_loss": 0.6941319704055786, "eval_runtime": 498.1423, "eval_samples_per_second": 4.135, "eval_steps_per_second": 4.135, "perplexity": 2.0019705498207383}
+{"ts": "2025-12-24T10:29:00", "event": "eval", "step": 4100, "epoch": 1.7689320388349514, "eval_loss": 0.6929343938827515, "eval_runtime": 497.805, "eval_samples_per_second": 4.138, "eval_steps_per_second": 4.138, "perplexity": 1.999574471920571}
+{"ts": "2025-12-24T10:47:38", "event": "eval", "step": 4200, "epoch": 1.812081984897519, "eval_loss": 0.6918642520904541, "eval_runtime": 498.4981, "eval_samples_per_second": 4.132, "eval_steps_per_second": 4.132, "perplexity": 1.9974357882628415}
+{"ts": "2025-12-24T11:06:16", "event": "eval", "step": 4300, "epoch": 1.8552319309600862, "eval_loss": 0.6890321373939514, "eval_runtime": 498.5289, "eval_samples_per_second": 4.132, "eval_steps_per_second": 4.132, "perplexity": 1.9917868240446748}
+{"ts": "2025-12-24T11:24:55", "event": "eval", "step": 4400, "epoch": 1.8983818770226537, "eval_loss": 0.687366783618927, "eval_runtime": 497.5319, "eval_samples_per_second": 4.14, "eval_steps_per_second": 4.14, "perplexity": 1.9884725548195745}
+{"ts": "2025-12-24T11:43:29", "event": "eval", "step": 4500, "epoch": 1.9415318230852212, "eval_loss": 0.6851214170455933, "eval_runtime": 497.1141, "eval_samples_per_second": 4.144, "eval_steps_per_second": 4.144, "perplexity": 1.9840127138756343}
+{"ts": "2025-12-24T12:02:11", "event": "eval", "step": 4600, "epoch": 1.9846817691477887, "eval_loss": 0.6843361258506775, "eval_runtime": 498.3201, "eval_samples_per_second": 4.134, "eval_steps_per_second": 4.134, "perplexity": 1.9824552977534464}
+{"ts": "2025-12-24T12:20:51", "event": "eval", "step": 4700, "epoch": 2.027615965480043, "eval_loss": 0.6861101984977722, "eval_runtime": 498.0682, "eval_samples_per_second": 4.136, "eval_steps_per_second": 4.136, "perplexity": 1.98597543904121}
+{"ts": "2025-12-24T12:39:30", "event": "eval", "step": 4800, "epoch": 2.0707659115426105, "eval_loss": 0.6874316930770874, "eval_runtime": 497.7977, "eval_samples_per_second": 4.138, "eval_steps_per_second": 4.138, "perplexity": 1.988601629684719}
+{"ts": "2025-12-24T12:47:48", "event": "eval", "step": 4800, "epoch": 2.0707659115426105, "eval_loss": 0.6851214170455933, "eval_runtime": 497.0682, "eval_samples_per_second": 4.144, "eval_steps_per_second": 4.144, "perplexity": 1.9840127138756343}
diff --git a/sft_devstral_24B/logs/train.jsonl b/sft_devstral_24B/logs/train.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..a43ddcb8dac3634485daf1c9ee3c5d8539f6c022
--- /dev/null
+++ b/sft_devstral_24B/logs/train.jsonl
@@ -0,0 +1,2508 @@
+{"ts": "2025-12-23T13:45:15", "event": "train_log", "step": 2, "epoch": 0.0008629989212513484, "progress_pct": 0.01, "epoch_pct": 0.01, "eta": "37:47:48", "max_grad_norm": 1.0, "loss": 1.398809790611267, "grad_norm": 5.481492042541504, "learning_rate": 3.5714285714285716e-07}
+{"ts": "2025-12-23T13:45:35", "event": "train_log", "step": 4, "epoch": 0.001725997842502697, "progress_pct": 0.03, "epoch_pct": 0.03, "eta": "38:09:15", "max_grad_norm": 1.0, "loss": 1.3696162700653076, "grad_norm": 5.111456871032715, "learning_rate": 1.0714285714285716e-06}
+{"ts": "2025-12-23T13:47:02", "event": "train_log", "step": 2, "epoch": 0.0008629989212513484, "progress_pct": 0.01, "epoch_pct": 0.01, "eta": "37:41:55", "max_grad_norm": 1.0, "loss": 1.398809790611267, "grad_norm": 5.482326507568359, "learning_rate": 3.5714285714285716e-07}
+{"ts": "2025-12-23T13:47:20", "event": "train_log", "step": 4, "epoch": 0.001725997842502697, "progress_pct": 0.03, "epoch_pct": 0.03, "eta": "36:45:00", "max_grad_norm": 1.0, "loss": 1.369497299194336, "grad_norm": 5.093585968017578, "learning_rate": 1.0714285714285716e-06}
+{"ts": "2025-12-23T13:47:38", "event": "train_log", "step": 6, "epoch": 0.0025889967637540453, "progress_pct": 0.04, "epoch_pct": 0.04, "eta": "35:36:28", "max_grad_norm": 1.0, "loss": 1.4427783489227295, "grad_norm": 4.639692306518555, "learning_rate": 1.7857142857142857e-06}
+{"ts": "2025-12-23T13:47:58", "event": "train_log", "step": 8, "epoch": 0.003451995685005394, "progress_pct": 0.06, "epoch_pct": 0.06, "eta": "36:26:45", "max_grad_norm": 1.0, "loss": 1.317796230316162, "grad_norm": 4.058807373046875, "learning_rate": 2.5e-06}
+{"ts": "2025-12-23T13:48:16", "event": "train_log", "step": 10, "epoch": 0.004314994606256742, "progress_pct": 0.07, "epoch_pct": 0.07, "eta": "36:16:11", "max_grad_norm": 1.0, "loss": 1.3889520168304443, "grad_norm": 4.042872428894043, "learning_rate": 3.2142857142857143e-06}
+{"ts": "2025-12-23T13:48:36", "event": "train_log", "step": 12, "epoch": 0.005177993527508091, "progress_pct": 0.09, "epoch_pct": 0.09, "eta": "36:32:55", "max_grad_norm": 1.0, "loss": 1.328926682472229, "grad_norm": 3.2458009719848633, "learning_rate": 3.928571428571429e-06}
+{"ts": "2025-12-23T13:48:55", "event": "train_log", "step": 14, "epoch": 0.006040992448759439, "progress_pct": 0.1, "epoch_pct": 0.1, "eta": "36:41:33", "max_grad_norm": 1.0, "loss": 1.3180789947509766, "grad_norm": 2.7812764644622803, "learning_rate": 4.642857142857143e-06}
+{"ts": "2025-12-23T13:49:14", "event": "train_log", "step": 16, "epoch": 0.006903991370010788, "progress_pct": 0.12, "epoch_pct": 0.12, "eta": "36:31:02", "max_grad_norm": 1.0, "loss": 1.2523515224456787, "grad_norm": 3.072478771209717, "learning_rate": 5.357142857142857e-06}
+{"ts": "2025-12-23T13:49:34", "event": "train_log", "step": 18, "epoch": 0.007766990291262136, "progress_pct": 0.13, "epoch_pct": 0.13, "eta": "36:47:46", "max_grad_norm": 1.0, "loss": 1.2902865409851074, "grad_norm": 3.046903133392334, "learning_rate": 6.071428571428572e-06}
+{"ts": "2025-12-23T13:49:53", "event": "train_log", "step": 20, "epoch": 0.008629989212513484, "progress_pct": 0.14, "epoch_pct": 0.14, "eta": "36:50:15", "max_grad_norm": 1.0, "loss": 1.2371950149536133, "grad_norm": 2.7814486026763916, "learning_rate": 6.785714285714285e-06}
+{"ts": "2025-12-23T13:50:12", "event": "train_log", "step": 22, "epoch": 0.009492988133764833, "progress_pct": 0.16, "epoch_pct": 0.16, "eta": "36:48:55", "max_grad_norm": 1.0, "loss": 1.1684906482696533, "grad_norm": 2.3911187648773193, "learning_rate": 7.5e-06}
+{"ts": "2025-12-23T13:50:31", "event": "train_log", "step": 24, "epoch": 0.010355987055016181, "progress_pct": 0.17, "epoch_pct": 0.17, "eta": "36:44:38", "max_grad_norm": 1.0, "loss": 1.1299467086791992, "grad_norm": 3.6106483936309814, "learning_rate": 8.214285714285714e-06}
+{"ts": "2025-12-23T13:50:50", "event": "train_log", "step": 26, "epoch": 0.01121898597626753, "progress_pct": 0.19, "epoch_pct": 0.19, "eta": "36:40:51", "max_grad_norm": 1.0, "loss": 1.2047266960144043, "grad_norm": 2.8624985218048096, "learning_rate": 8.92857142857143e-06}
+{"ts": "2025-12-23T13:51:10", "event": "train_log", "step": 28, "epoch": 0.012081984897518877, "progress_pct": 0.2, "epoch_pct": 0.2, "eta": "36:51:26", "max_grad_norm": 1.0, "loss": 1.069214105606079, "grad_norm": 2.4541709423065186, "learning_rate": 9.642857142857144e-06}
+{"ts": "2025-12-23T13:51:28", "event": "train_log", "step": 30, "epoch": 0.012944983818770227, "progress_pct": 0.22, "epoch_pct": 0.22, "eta": "36:42:06", "max_grad_norm": 1.0, "loss": 1.1609973907470703, "grad_norm": 3.123530149459839, "learning_rate": 1.0357142857142859e-05}
+{"ts": "2025-12-23T13:51:47", "event": "train_log", "step": 32, "epoch": 0.013807982740021575, "progress_pct": 0.23, "epoch_pct": 0.23, "eta": "36:40:48", "max_grad_norm": 1.0, "loss": 1.1104941368103027, "grad_norm": 2.6204144954681396, "learning_rate": 1.107142857142857e-05}
+{"ts": "2025-12-23T13:52:06", "event": "train_log", "step": 34, "epoch": 0.014670981661272923, "progress_pct": 0.24, "epoch_pct": 0.24, "eta": "36:44:32", "max_grad_norm": 1.0, "loss": 1.1072092056274414, "grad_norm": 2.8375027179718018, "learning_rate": 1.1785714285714286e-05}
+{"ts": "2025-12-23T13:52:26", "event": "train_log", "step": 36, "epoch": 0.015533980582524271, "progress_pct": 0.26, "epoch_pct": 0.26, "eta": "36:50:31", "max_grad_norm": 1.0, "loss": 1.0041414499282837, "grad_norm": 2.09930419921875, "learning_rate": 1.25e-05}
+{"ts": "2025-12-23T13:52:44", "event": "train_log", "step": 38, "epoch": 0.01639697950377562, "progress_pct": 0.27, "epoch_pct": 0.27, "eta": "36:43:47", "max_grad_norm": 1.0, "loss": 1.0711103677749634, "grad_norm": 2.638849973678589, "learning_rate": 1.3214285714285716e-05}
+{"ts": "2025-12-23T13:53:02", "event": "train_log", "step": 40, "epoch": 0.017259978425026967, "progress_pct": 0.29, "epoch_pct": 0.29, "eta": "36:36:21", "max_grad_norm": 1.0, "loss": 1.0649389028549194, "grad_norm": 2.677057981491089, "learning_rate": 1.392857142857143e-05}
+{"ts": "2025-12-23T13:53:22", "event": "train_log", "step": 42, "epoch": 0.018122977346278317, "progress_pct": 0.3, "epoch_pct": 0.3, "eta": "36:39:31", "max_grad_norm": 1.0, "loss": 1.0159238576889038, "grad_norm": 3.0247950553894043, "learning_rate": 1.4642857142857144e-05}
+{"ts": "2025-12-23T13:53:40", "event": "train_log", "step": 44, "epoch": 0.018985976267529667, "progress_pct": 0.32, "epoch_pct": 0.32, "eta": "36:35:38", "max_grad_norm": 1.0, "loss": 1.0905426740646362, "grad_norm": 2.372150182723999, "learning_rate": 1.535714285714286e-05}
+{"ts": "2025-12-23T13:53:59", "event": "train_log", "step": 46, "epoch": 0.019848975188781013, "progress_pct": 0.33, "epoch_pct": 0.33, "eta": "36:32:36", "max_grad_norm": 1.0, "loss": 1.0296133756637573, "grad_norm": 5.351362228393555, "learning_rate": 1.6071428571428572e-05}
+{"ts": "2025-12-23T13:54:18", "event": "train_log", "step": 48, "epoch": 0.020711974110032363, "progress_pct": 0.35, "epoch_pct": 0.35, "eta": "36:31:37", "max_grad_norm": 1.0, "loss": 1.0082216262817383, "grad_norm": 2.344869375228882, "learning_rate": 1.6785714285714285e-05}
+{"ts": "2025-12-23T13:54:37", "event": "train_log", "step": 50, "epoch": 0.021574973031283712, "progress_pct": 0.36, "epoch_pct": 0.36, "eta": "36:33:18", "max_grad_norm": 1.0, "loss": 0.9324379563331604, "grad_norm": 3.2270748615264893, "learning_rate": 1.75e-05}
+{"ts": "2025-12-23T13:54:57", "event": "train_log", "step": 52, "epoch": 0.02243797195253506, "progress_pct": 0.37, "epoch_pct": 0.37, "eta": "36:35:11", "max_grad_norm": 1.0, "loss": 0.9127687215805054, "grad_norm": 2.3037588596343994, "learning_rate": 1.8214285714285715e-05}
+{"ts": "2025-12-23T13:55:15", "event": "train_log", "step": 54, "epoch": 0.02330097087378641, "progress_pct": 0.39, "epoch_pct": 0.39, "eta": "36:33:58", "max_grad_norm": 1.0, "loss": 0.9793843030929565, "grad_norm": 2.3500611782073975, "learning_rate": 1.892857142857143e-05}
+{"ts": "2025-12-23T13:55:34", "event": "train_log", "step": 56, "epoch": 0.024163969795037755, "progress_pct": 0.4, "epoch_pct": 0.4, "eta": "36:34:02", "max_grad_norm": 1.0, "loss": 0.9212470054626465, "grad_norm": 2.2497150897979736, "learning_rate": 1.9642857142857145e-05}
+{"ts": "2025-12-23T13:55:53", "event": "train_log", "step": 58, "epoch": 0.025026968716289105, "progress_pct": 0.42, "epoch_pct": 0.42, "eta": "36:30:57", "max_grad_norm": 1.0, "loss": 0.929749608039856, "grad_norm": 2.16688871383667, "learning_rate": 2.0357142857142858e-05}
+{"ts": "2025-12-23T13:56:11", "event": "train_log", "step": 60, "epoch": 0.025889967637540454, "progress_pct": 0.43, "epoch_pct": 0.43, "eta": "36:28:19", "max_grad_norm": 1.0, "loss": 0.9784646034240723, "grad_norm": 2.5682408809661865, "learning_rate": 2.107142857142857e-05}
+{"ts": "2025-12-23T13:56:32", "event": "train_log", "step": 62, "epoch": 0.0267529665587918, "progress_pct": 0.45, "epoch_pct": 0.45, "eta": "36:36:06", "max_grad_norm": 1.0, "loss": 0.8337579965591431, "grad_norm": 2.0420022010803223, "learning_rate": 2.1785714285714288e-05}
+{"ts": "2025-12-23T13:56:52", "event": "train_log", "step": 64, "epoch": 0.02761596548004315, "progress_pct": 0.46, "epoch_pct": 0.46, "eta": "36:39:06", "max_grad_norm": 1.0, "loss": 0.9178240299224854, "grad_norm": 2.075817823410034, "learning_rate": 2.25e-05}
+{"ts": "2025-12-23T13:57:10", "event": "train_log", "step": 66, "epoch": 0.0284789644012945, "progress_pct": 0.47, "epoch_pct": 0.47, "eta": "36:34:45", "max_grad_norm": 1.0, "loss": 1.0382288694381714, "grad_norm": 2.457765579223633, "learning_rate": 2.3214285714285715e-05}
+{"ts": "2025-12-23T13:57:28", "event": "train_log", "step": 68, "epoch": 0.029341963322545846, "progress_pct": 0.49, "epoch_pct": 0.49, "eta": "36:29:08", "max_grad_norm": 1.0, "loss": 1.0000046491622925, "grad_norm": 2.5692148208618164, "learning_rate": 2.392857142857143e-05}
+{"ts": "2025-12-23T13:57:47", "event": "train_log", "step": 70, "epoch": 0.030204962243797196, "progress_pct": 0.5, "epoch_pct": 0.5, "eta": "36:29:20", "max_grad_norm": 1.0, "loss": 0.9485794305801392, "grad_norm": 2.3122708797454834, "learning_rate": 2.4642857142857145e-05}
+{"ts": "2025-12-23T13:58:06", "event": "train_log", "step": 72, "epoch": 0.031067961165048542, "progress_pct": 0.52, "epoch_pct": 0.52, "eta": "36:30:11", "max_grad_norm": 1.0, "loss": 0.9413180947303772, "grad_norm": 2.129300832748413, "learning_rate": 2.5357142857142858e-05}
+{"ts": "2025-12-23T13:58:26", "event": "train_log", "step": 74, "epoch": 0.03193096008629989, "progress_pct": 0.53, "epoch_pct": 0.53, "eta": "36:32:17", "max_grad_norm": 1.0, "loss": 0.9600517749786377, "grad_norm": 2.1712887287139893, "learning_rate": 2.6071428571428574e-05}
+{"ts": "2025-12-23T13:58:45", "event": "train_log", "step": 76, "epoch": 0.03279395900755124, "progress_pct": 0.55, "epoch_pct": 0.55, "eta": "36:33:22", "max_grad_norm": 1.0, "loss": 0.9564223289489746, "grad_norm": 1.9636520147323608, "learning_rate": 2.6785714285714288e-05}
+{"ts": "2025-12-23T13:59:04", "event": "train_log", "step": 78, "epoch": 0.03365695792880259, "progress_pct": 0.56, "epoch_pct": 0.56, "eta": "36:32:22", "max_grad_norm": 1.0, "loss": 0.9691935777664185, "grad_norm": 2.7914485931396484, "learning_rate": 2.7500000000000004e-05}
+{"ts": "2025-12-23T13:59:23", "event": "train_log", "step": 80, "epoch": 0.034519956850053934, "progress_pct": 0.58, "epoch_pct": 0.58, "eta": "36:32:28", "max_grad_norm": 1.0, "loss": 0.9473975300788879, "grad_norm": 2.1383204460144043, "learning_rate": 2.8214285714285714e-05}
+{"ts": "2025-12-23T13:59:42", "event": "train_log", "step": 82, "epoch": 0.035382955771305284, "progress_pct": 0.59, "epoch_pct": 0.59, "eta": "36:30:26", "max_grad_norm": 1.0, "loss": 0.9599665403366089, "grad_norm": 2.583268165588379, "learning_rate": 2.8928571428571434e-05}
+{"ts": "2025-12-23T14:00:01", "event": "train_log", "step": 84, "epoch": 0.036245954692556634, "progress_pct": 0.6, "epoch_pct": 0.6, "eta": "36:31:38", "max_grad_norm": 1.0, "loss": 0.8810346722602844, "grad_norm": 2.1355762481689453, "learning_rate": 2.9642857142857144e-05}
+{"ts": "2025-12-23T14:00:22", "event": "train_log", "step": 86, "epoch": 0.037108953613807984, "progress_pct": 0.62, "epoch_pct": 0.62, "eta": "36:36:30", "max_grad_norm": 1.0, "loss": 0.8347419500350952, "grad_norm": 1.890364408493042, "learning_rate": 3.0357142857142857e-05}
+{"ts": "2025-12-23T14:00:42", "event": "train_log", "step": 88, "epoch": 0.03797195253505933, "progress_pct": 0.63, "epoch_pct": 0.63, "eta": "36:37:33", "max_grad_norm": 1.0, "loss": 0.9488616585731506, "grad_norm": 2.685027837753296, "learning_rate": 3.107142857142857e-05}
+{"ts": "2025-12-23T14:01:00", "event": "train_log", "step": 90, "epoch": 0.038834951456310676, "progress_pct": 0.65, "epoch_pct": 0.65, "eta": "36:33:51", "max_grad_norm": 1.0, "loss": 0.898134171962738, "grad_norm": 2.3440029621124268, "learning_rate": 3.1785714285714284e-05}
+{"ts": "2025-12-23T14:01:21", "event": "train_log", "step": 92, "epoch": 0.039697950377562026, "progress_pct": 0.66, "epoch_pct": 0.66, "eta": "36:38:47", "max_grad_norm": 1.0, "loss": 0.8588011860847473, "grad_norm": 2.402268171310425, "learning_rate": 3.2500000000000004e-05}
+{"ts": "2025-12-23T14:01:42", "event": "train_log", "step": 94, "epoch": 0.040560949298813376, "progress_pct": 0.68, "epoch_pct": 0.68, "eta": "36:43:48", "max_grad_norm": 1.0, "loss": 0.8564016222953796, "grad_norm": 1.965386152267456, "learning_rate": 3.321428571428572e-05}
+{"ts": "2025-12-23T14:02:01", "event": "train_log", "step": 96, "epoch": 0.041423948220064725, "progress_pct": 0.69, "epoch_pct": 0.69, "eta": "36:43:36", "max_grad_norm": 1.0, "loss": 1.0142979621887207, "grad_norm": 2.4201598167419434, "learning_rate": 3.392857142857143e-05}
+{"ts": "2025-12-23T14:02:21", "event": "train_log", "step": 98, "epoch": 0.042286947141316075, "progress_pct": 0.7, "epoch_pct": 0.7, "eta": "36:45:41", "max_grad_norm": 1.0, "loss": 0.8932055234909058, "grad_norm": 3.0488028526306152, "learning_rate": 3.4642857142857144e-05}
+{"ts": "2025-12-23T14:02:41", "event": "train_log", "step": 100, "epoch": 0.043149946062567425, "progress_pct": 0.72, "epoch_pct": 0.72, "eta": "36:46:38", "max_grad_norm": 1.0, "loss": 0.9055630564689636, "grad_norm": 1.8199278116226196, "learning_rate": 3.5357142857142864e-05}
+{"ts": "2025-12-23T14:16:36", "event": "train_log", "step": 100, "epoch": 0.043149946062567425, "progress_pct": 0.72, "epoch_pct": 0.72, "eta": "68:48:01", "max_grad_norm": 1.0, "eval_loss": 0.9360476136207581, "eval_runtime": 834.9, "eval_samples_per_second": 2.467, "eval_steps_per_second": 2.467}
+{"ts": "2025-12-23T14:16:54", "event": "train_log", "step": 102, "epoch": 0.04401294498381877, "progress_pct": 0.73, "epoch_pct": 0.73, "eta": "68:08:07", "max_grad_norm": 1.0, "loss": 0.9160765409469604, "grad_norm": 2.019763946533203, "learning_rate": 3.607142857142857e-05}
+{"ts": "2025-12-23T14:17:14", "event": "train_log", "step": 104, "epoch": 0.04487594390507012, "progress_pct": 0.75, "epoch_pct": 0.75, "eta": "67:32:47", "max_grad_norm": 1.0, "loss": 0.8526360988616943, "grad_norm": 2.063302993774414, "learning_rate": 3.678571428571429e-05}
+{"ts": "2025-12-23T14:17:33", "event": "train_log", "step": 106, "epoch": 0.04573894282632147, "progress_pct": 0.76, "epoch_pct": 0.76, "eta": "66:57:28", "max_grad_norm": 1.0, "loss": 0.906327486038208, "grad_norm": 2.216637372970581, "learning_rate": 3.7500000000000003e-05}
+{"ts": "2025-12-23T14:17:53", "event": "train_log", "step": 108, "epoch": 0.04660194174757282, "progress_pct": 0.78, "epoch_pct": 0.78, "eta": "66:23:30", "max_grad_norm": 1.0, "loss": 0.932016909122467, "grad_norm": 1.9437347650527954, "learning_rate": 3.821428571428572e-05}
+{"ts": "2025-12-23T14:18:13", "event": "train_log", "step": 110, "epoch": 0.04746494066882417, "progress_pct": 0.79, "epoch_pct": 0.79, "eta": "65:52:32", "max_grad_norm": 1.0, "loss": 0.9619514346122742, "grad_norm": 1.9441817998886108, "learning_rate": 3.892857142857143e-05}
+{"ts": "2025-12-23T14:23:42", "event": "train_log", "step": 2, "epoch": 0.0008629989212513484, "progress_pct": 0.01, "epoch_pct": 0.01, "eta": "37:21:50", "max_grad_norm": 0.8, "loss": 1.398809790611267, "grad_norm": 1.9118273258209229, "learning_rate": 8.984725965858042e-08}
+{"ts": "2025-12-23T14:24:02", "event": "train_log", "step": 4, "epoch": 0.001725997842502697, "progress_pct": 0.03, "epoch_pct": 0.03, "eta": "38:13:08", "max_grad_norm": 0.8, "loss": 1.3699476718902588, "grad_norm": 1.785147786140442, "learning_rate": 2.6954177897574125e-07}
+{"ts": "2025-12-23T14:24:21", "event": "train_log", "step": 6, "epoch": 0.0025889967637540453, "progress_pct": 0.04, "epoch_pct": 0.04, "eta": "37:30:21", "max_grad_norm": 0.8, "loss": 1.4487650394439697, "grad_norm": 1.6857414245605469, "learning_rate": 4.492362982929021e-07}
+{"ts": "2025-12-23T14:24:33", "event": "train_log", "step": 8, "epoch": 0.003451995685005394, "progress_pct": 0.06, "epoch_pct": 0.06, "eta": "34:19:58", "max_grad_norm": 0.8, "loss": 1.3360817432403564, "grad_norm": 1.7351312637329102, "learning_rate": 6.28930817610063e-07}
+{"ts": "2025-12-23T14:24:45", "event": "train_log", "step": 10, "epoch": 0.004314994606256742, "progress_pct": 0.07, "epoch_pct": 0.07, "eta": "32:02:38", "max_grad_norm": 0.8, "loss": 1.4294791221618652, "grad_norm": 1.7769269943237305, "learning_rate": 8.086253369272238e-07}
+{"ts": "2025-12-23T14:24:58", "event": "train_log", "step": 12, "epoch": 0.005177993527508091, "progress_pct": 0.09, "epoch_pct": 0.09, "eta": "30:43:21", "max_grad_norm": 0.8, "loss": 1.3927934169769287, "grad_norm": 1.776668906211853, "learning_rate": 9.883198562443846e-07}
+{"ts": "2025-12-23T14:25:10", "event": "train_log", "step": 14, "epoch": 0.006040992448759439, "progress_pct": 0.1, "epoch_pct": 0.1, "eta": "29:45:59", "max_grad_norm": 0.8, "loss": 1.4140764474868774, "grad_norm": 1.500367283821106, "learning_rate": 1.1680143755615454e-06}
+{"ts": "2025-12-23T14:25:22", "event": "train_log", "step": 16, "epoch": 0.006903991370010788, "progress_pct": 0.12, "epoch_pct": 0.12, "eta": "28:55:39", "max_grad_norm": 0.8, "loss": 1.3873623609542847, "grad_norm": 1.7255949974060059, "learning_rate": 1.3477088948787064e-06}
+{"ts": "2025-12-23T14:25:35", "event": "train_log", "step": 18, "epoch": 0.007766990291262136, "progress_pct": 0.13, "epoch_pct": 0.13, "eta": "28:22:14", "max_grad_norm": 0.8, "loss": 1.4562759399414062, "grad_norm": 1.9073798656463623, "learning_rate": 1.5274034141958671e-06}
+{"ts": "2025-12-23T14:25:47", "event": "train_log", "step": 20, "epoch": 0.008629989212513484, "progress_pct": 0.14, "epoch_pct": 0.14, "eta": "27:57:42", "max_grad_norm": 0.8, "loss": 1.4264800548553467, "grad_norm": 1.7253859043121338, "learning_rate": 1.707097933513028e-06}
+{"ts": "2025-12-23T14:26:00", "event": "train_log", "step": 22, "epoch": 0.009492988133764833, "progress_pct": 0.16, "epoch_pct": 0.16, "eta": "27:35:51", "max_grad_norm": 0.8, "loss": 1.3717138767242432, "grad_norm": 1.4858287572860718, "learning_rate": 1.8867924528301887e-06}
+{"ts": "2025-12-23T14:26:12", "event": "train_log", "step": 24, "epoch": 0.010355987055016181, "progress_pct": 0.17, "epoch_pct": 0.17, "eta": "27:13:58", "max_grad_norm": 0.8, "loss": 1.3605879545211792, "grad_norm": 1.7516961097717285, "learning_rate": 2.0664869721473496e-06}
+{"ts": "2025-12-23T14:26:24", "event": "train_log", "step": 26, "epoch": 0.01121898597626753, "progress_pct": 0.19, "epoch_pct": 0.19, "eta": "26:55:38", "max_grad_norm": 0.8, "loss": 1.465327262878418, "grad_norm": 1.754793643951416, "learning_rate": 2.2461814914645104e-06}
+{"ts": "2025-12-23T14:26:37", "event": "train_log", "step": 28, "epoch": 0.012081984897518877, "progress_pct": 0.2, "epoch_pct": 0.2, "eta": "26:48:50", "max_grad_norm": 0.8, "loss": 1.3205785751342773, "grad_norm": 1.5204869508743286, "learning_rate": 2.425876010781671e-06}
+{"ts": "2025-12-23T14:26:49", "event": "train_log", "step": 30, "epoch": 0.012944983818770227, "progress_pct": 0.22, "epoch_pct": 0.22, "eta": "26:31:19", "max_grad_norm": 0.8, "loss": 1.4751381874084473, "grad_norm": 2.03002667427063, "learning_rate": 2.605570530098832e-06}
+{"ts": "2025-12-23T14:27:01", "event": "train_log", "step": 32, "epoch": 0.013807982740021575, "progress_pct": 0.23, "epoch_pct": 0.23, "eta": "26:19:34", "max_grad_norm": 0.8, "loss": 1.3976212739944458, "grad_norm": 1.6153842210769653, "learning_rate": 2.785265049415993e-06}
+{"ts": "2025-12-23T14:27:14", "event": "train_log", "step": 34, "epoch": 0.014670981661272923, "progress_pct": 0.24, "epoch_pct": 0.24, "eta": "26:14:29", "max_grad_norm": 0.8, "loss": 1.4062179327011108, "grad_norm": 1.7034062147140503, "learning_rate": 2.964959568733154e-06}
+{"ts": "2025-12-23T14:27:27", "event": "train_log", "step": 36, "epoch": 0.015533980582524271, "progress_pct": 0.26, "epoch_pct": 0.26, "eta": "26:10:43", "max_grad_norm": 0.8, "loss": 1.2940696477890015, "grad_norm": 1.5724587440490723, "learning_rate": 3.1446540880503146e-06}
+{"ts": "2025-12-23T14:27:38", "event": "train_log", "step": 38, "epoch": 0.01639697950377562, "progress_pct": 0.27, "epoch_pct": 0.27, "eta": "25:57:31", "max_grad_norm": 0.8, "loss": 1.4216598272323608, "grad_norm": 1.7072066068649292, "learning_rate": 3.324348607367475e-06}
+{"ts": "2025-12-23T14:27:50", "event": "train_log", "step": 40, "epoch": 0.017259978425026967, "progress_pct": 0.29, "epoch_pct": 0.29, "eta": "25:45:30", "max_grad_norm": 0.8, "loss": 1.4204131364822388, "grad_norm": 1.6051130294799805, "learning_rate": 3.504043126684636e-06}
+{"ts": "2025-12-23T14:28:03", "event": "train_log", "step": 42, "epoch": 0.018122977346278317, "progress_pct": 0.3, "epoch_pct": 0.3, "eta": "25:42:59", "max_grad_norm": 0.8, "loss": 1.3492953777313232, "grad_norm": 1.5114386081695557, "learning_rate": 3.683737646001797e-06}
+{"ts": "2025-12-23T14:28:15", "event": "train_log", "step": 44, "epoch": 0.018985976267529667, "progress_pct": 0.32, "epoch_pct": 0.32, "eta": "25:35:42", "max_grad_norm": 0.8, "loss": 1.4241598844528198, "grad_norm": 1.5557284355163574, "learning_rate": 3.863432165318958e-06}
+{"ts": "2025-12-23T14:28:27", "event": "train_log", "step": 46, "epoch": 0.019848975188781013, "progress_pct": 0.33, "epoch_pct": 0.33, "eta": "25:28:53", "max_grad_norm": 0.8, "loss": 1.3903125524520874, "grad_norm": 1.5814268589019775, "learning_rate": 4.0431266846361185e-06}
+{"ts": "2025-12-23T14:28:45", "event": "train_log", "step": 48, "epoch": 0.020711974110032363, "progress_pct": 0.35, "epoch_pct": 0.35, "eta": "25:53:49", "max_grad_norm": 0.8, "loss": 1.3726638555526733, "grad_norm": 1.53256356716156, "learning_rate": 4.22282120395328e-06}
+{"ts": "2025-12-23T14:29:06", "event": "train_log", "step": 50, "epoch": 0.021574973031283712, "progress_pct": 0.36, "epoch_pct": 0.36, "eta": "26:26:03", "max_grad_norm": 0.8, "loss": 1.2536468505859375, "grad_norm": 1.5670084953308105, "learning_rate": 4.40251572327044e-06}
+{"ts": "2025-12-23T14:29:26", "event": "train_log", "step": 52, "epoch": 0.02243797195253506, "progress_pct": 0.37, "epoch_pct": 0.37, "eta": "26:56:16", "max_grad_norm": 0.8, "loss": 1.2309948205947876, "grad_norm": 1.3708246946334839, "learning_rate": 4.582210242587602e-06}
+{"ts": "2025-12-23T14:29:45", "event": "train_log", "step": 54, "epoch": 0.02330097087378641, "progress_pct": 0.39, "epoch_pct": 0.39, "eta": "27:15:48", "max_grad_norm": 0.8, "loss": 1.3564441204071045, "grad_norm": 1.5670109987258911, "learning_rate": 4.7619047619047615e-06}
+{"ts": "2025-12-23T14:30:05", "event": "train_log", "step": 56, "epoch": 0.024163969795037755, "progress_pct": 0.4, "epoch_pct": 0.4, "eta": "27:41:26", "max_grad_norm": 0.8, "loss": 1.2251620292663574, "grad_norm": 1.3550655841827393, "learning_rate": 4.941599281221923e-06}
+{"ts": "2025-12-23T14:30:24", "event": "train_log", "step": 58, "epoch": 0.025026968716289105, "progress_pct": 0.42, "epoch_pct": 0.42, "eta": "27:57:13", "max_grad_norm": 0.8, "loss": 1.2395249605178833, "grad_norm": 1.4416121244430542, "learning_rate": 5.121293800539084e-06}
+{"ts": "2025-12-23T14:30:42", "event": "train_log", "step": 60, "epoch": 0.025889967637540454, "progress_pct": 0.43, "epoch_pct": 0.43, "eta": "28:12:25", "max_grad_norm": 0.8, "loss": 1.3229656219482422, "grad_norm": 1.4638670682907104, "learning_rate": 5.300988319856245e-06}
+{"ts": "2025-12-23T14:31:04", "event": "train_log", "step": 62, "epoch": 0.0267529665587918, "progress_pct": 0.45, "epoch_pct": 0.45, "eta": "28:37:17", "max_grad_norm": 0.8, "loss": 1.1076340675354004, "grad_norm": 1.2204002141952515, "learning_rate": 5.4806828391734054e-06}
+{"ts": "2025-12-23T14:31:23", "event": "train_log", "step": 64, "epoch": 0.02761596548004315, "progress_pct": 0.46, "epoch_pct": 0.46, "eta": "28:51:21", "max_grad_norm": 0.8, "loss": 1.237392544746399, "grad_norm": 1.2641915082931519, "learning_rate": 5.660377358490566e-06}
+{"ts": "2025-12-23T14:31:42", "event": "train_log", "step": 66, "epoch": 0.0284789644012945, "progress_pct": 0.47, "epoch_pct": 0.47, "eta": "29:06:01", "max_grad_norm": 0.8, "loss": 1.3567079305648804, "grad_norm": 1.4635342359542847, "learning_rate": 5.840071877807727e-06}
+{"ts": "2025-12-23T14:31:59", "event": "train_log", "step": 68, "epoch": 0.029341963322545846, "progress_pct": 0.49, "epoch_pct": 0.49, "eta": "29:13:26", "max_grad_norm": 0.8, "loss": 1.3595421314239502, "grad_norm": 1.5813465118408203, "learning_rate": 6.019766397124888e-06}
+{"ts": "2025-12-23T14:32:19", "event": "train_log", "step": 70, "epoch": 0.030204962243797196, "progress_pct": 0.5, "epoch_pct": 0.5, "eta": "29:26:59", "max_grad_norm": 0.8, "loss": 1.2183864116668701, "grad_norm": 1.435356616973877, "learning_rate": 6.1994609164420485e-06}
+{"ts": "2025-12-23T14:32:38", "event": "train_log", "step": 72, "epoch": 0.031067961165048542, "progress_pct": 0.52, "epoch_pct": 0.52, "eta": "29:40:41", "max_grad_norm": 0.8, "loss": 1.2084243297576904, "grad_norm": 1.4559556245803833, "learning_rate": 6.37915543575921e-06}
+{"ts": "2025-12-23T14:32:58", "event": "train_log", "step": 74, "epoch": 0.03193096008629989, "progress_pct": 0.53, "epoch_pct": 0.53, "eta": "29:54:41", "max_grad_norm": 0.8, "loss": 1.2476426362991333, "grad_norm": 1.2870471477508545, "learning_rate": 6.558849955076371e-06}
+{"ts": "2025-12-23T14:33:18", "event": "train_log", "step": 76, "epoch": 0.03279395900755124, "progress_pct": 0.55, "epoch_pct": 0.55, "eta": "30:07:32", "max_grad_norm": 0.8, "loss": 1.2200642824172974, "grad_norm": 1.3345798254013062, "learning_rate": 6.738544474393531e-06}
+{"ts": "2025-12-23T14:33:38", "event": "train_log", "step": 78, "epoch": 0.03365695792880259, "progress_pct": 0.56, "epoch_pct": 0.56, "eta": "30:19:09", "max_grad_norm": 0.8, "loss": 1.2497955560684204, "grad_norm": 1.48223876953125, "learning_rate": 6.918238993710692e-06}
+{"ts": "2025-12-23T14:33:57", "event": "train_log", "step": 80, "epoch": 0.034519956850053934, "progress_pct": 0.58, "epoch_pct": 0.58, "eta": "30:28:58", "max_grad_norm": 0.8, "loss": 1.1987318992614746, "grad_norm": 1.2421257495880127, "learning_rate": 7.097933513027853e-06}
+{"ts": "2025-12-23T14:34:16", "event": "train_log", "step": 82, "epoch": 0.035382955771305284, "progress_pct": 0.59, "epoch_pct": 0.59, "eta": "30:36:01", "max_grad_norm": 0.8, "loss": 1.218438982963562, "grad_norm": 1.5042678117752075, "learning_rate": 7.277628032345015e-06}
+{"ts": "2025-12-23T14:34:37", "event": "train_log", "step": 84, "epoch": 0.036245954692556634, "progress_pct": 0.6, "epoch_pct": 0.6, "eta": "30:49:56", "max_grad_norm": 0.8, "loss": 1.1181117296218872, "grad_norm": 1.3047534227371216, "learning_rate": 7.457322551662174e-06}
+{"ts": "2025-12-23T14:34:58", "event": "train_log", "step": 86, "epoch": 0.037108953613807984, "progress_pct": 0.62, "epoch_pct": 0.62, "eta": "31:04:55", "max_grad_norm": 0.8, "loss": 1.047302484512329, "grad_norm": 1.3297117948532104, "learning_rate": 7.637017070979335e-06}
+{"ts": "2025-12-23T14:35:19", "event": "train_log", "step": 88, "epoch": 0.03797195253505933, "progress_pct": 0.63, "epoch_pct": 0.63, "eta": "31:16:58", "max_grad_norm": 0.8, "loss": 1.1821871995925903, "grad_norm": 1.4515916109085083, "learning_rate": 7.816711590296496e-06}
+{"ts": "2025-12-23T14:35:38", "event": "train_log", "step": 90, "epoch": 0.038834951456310676, "progress_pct": 0.65, "epoch_pct": 0.65, "eta": "31:23:32", "max_grad_norm": 0.8, "loss": 1.1171189546585083, "grad_norm": 1.5435426235198975, "learning_rate": 7.996406109613658e-06}
+{"ts": "2025-12-23T14:35:58", "event": "train_log", "step": 92, "epoch": 0.039697950377562026, "progress_pct": 0.66, "epoch_pct": 0.66, "eta": "31:32:19", "max_grad_norm": 0.8, "loss": 1.067047357559204, "grad_norm": 1.5270947217941284, "learning_rate": 8.176100628930818e-06}
+{"ts": "2025-12-23T14:36:18", "event": "train_log", "step": 94, "epoch": 0.040560949298813376, "progress_pct": 0.68, "epoch_pct": 0.68, "eta": "31:40:11", "max_grad_norm": 0.8, "loss": 1.0745081901550293, "grad_norm": 2.4402992725372314, "learning_rate": 8.35579514824798e-06}
+{"ts": "2025-12-23T14:36:38", "event": "train_log", "step": 96, "epoch": 0.041423948220064725, "progress_pct": 0.69, "epoch_pct": 0.69, "eta": "31:48:39", "max_grad_norm": 0.8, "loss": 1.252164602279663, "grad_norm": 1.7703081369400024, "learning_rate": 8.53548966756514e-06}
+{"ts": "2025-12-23T14:36:59", "event": "train_log", "step": 98, "epoch": 0.042286947141316075, "progress_pct": 0.7, "epoch_pct": 0.7, "eta": "31:57:18", "max_grad_norm": 0.8, "loss": 1.1030420064926147, "grad_norm": 1.4324337244033813, "learning_rate": 8.715184186882301e-06}
+{"ts": "2025-12-23T14:37:18", "event": "train_log", "step": 100, "epoch": 0.043149946062567425, "progress_pct": 0.72, "epoch_pct": 0.72, "eta": "32:02:51", "max_grad_norm": 0.8, "loss": 1.0817893743515015, "grad_norm": 1.3258590698242188, "learning_rate": 8.89487870619946e-06}
+{"ts": "2025-12-23T14:51:03", "event": "train_log", "step": 100, "epoch": 0.043149946062567425, "progress_pct": 0.72, "epoch_pct": 0.72, "eta": "63:41:30", "max_grad_norm": 0.8, "eval_loss": 1.128923773765564, "eval_runtime": 825.0166, "eval_samples_per_second": 2.497, "eval_steps_per_second": 2.497}
+{"ts": "2025-12-23T14:51:22", "event": "train_log", "step": 102, "epoch": 0.04401294498381877, "progress_pct": 0.73, "epoch_pct": 0.73, "eta": "63:08:08", "max_grad_norm": 0.8, "loss": 1.0982940196990967, "grad_norm": 1.5833592414855957, "learning_rate": 9.074573225516622e-06}
+{"ts": "2025-12-23T14:51:41", "event": "train_log", "step": 104, "epoch": 0.04487594390507012, "progress_pct": 0.75, "epoch_pct": 0.75, "eta": "62:38:52", "max_grad_norm": 0.8, "loss": 1.0529299974441528, "grad_norm": 1.5289201736450195, "learning_rate": 9.254267744833784e-06}
+{"ts": "2025-12-23T14:52:01", "event": "train_log", "step": 106, "epoch": 0.04573894282632147, "progress_pct": 0.76, "epoch_pct": 0.76, "eta": "62:08:48", "max_grad_norm": 0.8, "loss": 1.0896434783935547, "grad_norm": 1.4507837295532227, "learning_rate": 9.433962264150944e-06}
+{"ts": "2025-12-23T14:52:21", "event": "train_log", "step": 108, "epoch": 0.04660194174757282, "progress_pct": 0.78, "epoch_pct": 0.78, "eta": "61:43:24", "max_grad_norm": 0.8, "loss": 1.1196398735046387, "grad_norm": 1.3846485614776611, "learning_rate": 9.613656783468104e-06}
+{"ts": "2025-12-23T14:52:42", "event": "train_log", "step": 110, "epoch": 0.04746494066882417, "progress_pct": 0.79, "epoch_pct": 0.79, "eta": "61:19:03", "max_grad_norm": 0.8, "loss": 1.1305382251739502, "grad_norm": 1.4094940423965454, "learning_rate": 9.793351302785265e-06}
+{"ts": "2025-12-23T14:53:03", "event": "train_log", "step": 112, "epoch": 0.04832793959007551, "progress_pct": 0.81, "epoch_pct": 0.81, "eta": "60:55:42", "max_grad_norm": 0.8, "loss": 1.0138700008392334, "grad_norm": 1.7674189805984497, "learning_rate": 9.973045822102427e-06}
+{"ts": "2025-12-23T14:53:23", "event": "train_log", "step": 114, "epoch": 0.04919093851132686, "progress_pct": 0.82, "epoch_pct": 0.82, "eta": "60:32:08", "max_grad_norm": 0.8, "loss": 1.0583173036575317, "grad_norm": 1.6083961725234985, "learning_rate": 1.0152740341419587e-05}
+{"ts": "2025-12-23T14:53:43", "event": "train_log", "step": 116, "epoch": 0.05005393743257821, "progress_pct": 0.83, "epoch_pct": 0.83, "eta": "60:07:21", "max_grad_norm": 0.8, "loss": 1.1116671562194824, "grad_norm": 1.7483229637145996, "learning_rate": 1.0332434860736747e-05}
+{"ts": "2025-12-23T14:54:00", "event": "train_log", "step": 118, "epoch": 0.05091693635382956, "progress_pct": 0.85, "epoch_pct": 0.85, "eta": "59:39:44", "max_grad_norm": 0.8, "loss": 1.0832366943359375, "grad_norm": 1.8256834745407104, "learning_rate": 1.0512129380053909e-05}
+{"ts": "2025-12-23T14:54:19", "event": "train_log", "step": 120, "epoch": 0.05177993527508091, "progress_pct": 0.86, "epoch_pct": 0.86, "eta": "59:15:46", "max_grad_norm": 0.8, "loss": 1.1089236736297607, "grad_norm": 1.5875557661056519, "learning_rate": 1.069182389937107e-05}
+{"ts": "2025-12-23T14:54:38", "event": "train_log", "step": 122, "epoch": 0.05264293419633225, "progress_pct": 0.88, "epoch_pct": 0.88, "eta": "58:53:08", "max_grad_norm": 0.8, "loss": 1.1191619634628296, "grad_norm": 1.6553549766540527, "learning_rate": 1.087151841868823e-05}
+{"ts": "2025-12-23T14:54:57", "event": "train_log", "step": 124, "epoch": 0.0535059331175836, "progress_pct": 0.89, "epoch_pct": 0.89, "eta": "58:30:58", "max_grad_norm": 0.8, "loss": 1.031973958015442, "grad_norm": 1.3675798177719116, "learning_rate": 1.1051212938005392e-05}
+{"ts": "2025-12-23T14:55:17", "event": "train_log", "step": 126, "epoch": 0.05436893203883495, "progress_pct": 0.91, "epoch_pct": 0.91, "eta": "58:10:20", "max_grad_norm": 0.8, "loss": 0.9964010119438171, "grad_norm": 1.5343760251998901, "learning_rate": 1.1230907457322552e-05}
+{"ts": "2025-12-23T14:55:35", "event": "train_log", "step": 128, "epoch": 0.0552319309600863, "progress_pct": 0.92, "epoch_pct": 0.92, "eta": "57:48:04", "max_grad_norm": 0.8, "loss": 1.1144609451293945, "grad_norm": 1.5928982496261597, "learning_rate": 1.1410601976639713e-05}
+{"ts": "2025-12-23T14:55:54", "event": "train_log", "step": 130, "epoch": 0.05609492988133765, "progress_pct": 0.93, "epoch_pct": 0.93, "eta": "57:27:16", "max_grad_norm": 0.8, "loss": 0.9953935146331787, "grad_norm": 1.6411585807800293, "learning_rate": 1.1590296495956873e-05}
+{"ts": "2025-12-23T14:56:14", "event": "train_log", "step": 132, "epoch": 0.056957928802589, "progress_pct": 0.95, "epoch_pct": 0.95, "eta": "57:09:32", "max_grad_norm": 0.8, "loss": 0.9940203428268433, "grad_norm": 1.6077220439910889, "learning_rate": 1.1769991015274035e-05}
+{"ts": "2025-12-23T14:56:33", "event": "train_log", "step": 134, "epoch": 0.05782092772384034, "progress_pct": 0.96, "epoch_pct": 0.96, "eta": "56:49:43", "max_grad_norm": 0.8, "loss": 1.045546293258667, "grad_norm": 1.7327438592910767, "learning_rate": 1.1949685534591196e-05}
+{"ts": "2025-12-23T14:56:52", "event": "train_log", "step": 136, "epoch": 0.05868392664509169, "progress_pct": 0.98, "epoch_pct": 0.98, "eta": "56:32:24", "max_grad_norm": 0.8, "loss": 0.9932523369789124, "grad_norm": 1.607404112815857, "learning_rate": 1.2129380053908356e-05}
+{"ts": "2025-12-23T14:57:13", "event": "train_log", "step": 138, "epoch": 0.05954692556634304, "progress_pct": 0.99, "epoch_pct": 0.99, "eta": "56:16:47", "max_grad_norm": 0.8, "loss": 0.9584172368049622, "grad_norm": 1.5056802034378052, "learning_rate": 1.2309074573225516e-05}
+{"ts": "2025-12-23T14:57:33", "event": "train_log", "step": 140, "epoch": 0.06040992448759439, "progress_pct": 1.01, "epoch_pct": 1.01, "eta": "56:01:36", "max_grad_norm": 0.8, "loss": 0.9733421206474304, "grad_norm": 1.5453966856002808, "learning_rate": 1.2488769092542678e-05}
+{"ts": "2025-12-23T14:57:53", "event": "train_log", "step": 142, "epoch": 0.06127292340884574, "progress_pct": 1.02, "epoch_pct": 1.02, "eta": "55:45:34", "max_grad_norm": 0.8, "loss": 0.9971115589141846, "grad_norm": 1.546397089958191, "learning_rate": 1.2668463611859838e-05}
+{"ts": "2025-12-23T14:58:13", "event": "train_log", "step": 144, "epoch": 0.062135922330097085, "progress_pct": 1.04, "epoch_pct": 1.04, "eta": "55:30:17", "max_grad_norm": 0.8, "loss": 1.0994833707809448, "grad_norm": 1.5259592533111572, "learning_rate": 1.2848158131177e-05}
+{"ts": "2025-12-23T14:58:33", "event": "train_log", "step": 146, "epoch": 0.06299892125134844, "progress_pct": 1.05, "epoch_pct": 1.05, "eta": "55:15:20", "max_grad_norm": 0.8, "loss": 1.0203801393508911, "grad_norm": 1.56397545337677, "learning_rate": 1.302785265049416e-05}
+{"ts": "2025-12-23T14:58:52", "event": "train_log", "step": 148, "epoch": 0.06386192017259978, "progress_pct": 1.06, "epoch_pct": 1.06, "eta": "55:00:03", "max_grad_norm": 0.8, "loss": 1.0223872661590576, "grad_norm": 1.618053674697876, "learning_rate": 1.320754716981132e-05}
+{"ts": "2025-12-23T14:59:11", "event": "train_log", "step": 150, "epoch": 0.06472491909385113, "progress_pct": 1.08, "epoch_pct": 1.08, "eta": "54:44:50", "max_grad_norm": 0.8, "loss": 1.0207403898239136, "grad_norm": 1.9544042348861694, "learning_rate": 1.3387241689128482e-05}
+{"ts": "2025-12-23T14:59:30", "event": "train_log", "step": 152, "epoch": 0.06558791801510248, "progress_pct": 1.09, "epoch_pct": 1.09, "eta": "54:29:02", "max_grad_norm": 0.8, "loss": 1.025482177734375, "grad_norm": 1.7001299858093262, "learning_rate": 1.3566936208445644e-05}
+{"ts": "2025-12-23T14:59:48", "event": "train_log", "step": 154, "epoch": 0.06645091693635383, "progress_pct": 1.11, "epoch_pct": 1.11, "eta": "54:13:07", "max_grad_norm": 0.8, "loss": 1.0077980756759644, "grad_norm": 1.9833810329437256, "learning_rate": 1.3746630727762804e-05}
+{"ts": "2025-12-23T15:00:07", "event": "train_log", "step": 156, "epoch": 0.06731391585760518, "progress_pct": 1.12, "epoch_pct": 1.12, "eta": "53:59:15", "max_grad_norm": 0.8, "loss": 1.0046619176864624, "grad_norm": 1.6378769874572754, "learning_rate": 1.3926325247079966e-05}
+{"ts": "2025-12-23T15:00:27", "event": "train_log", "step": 158, "epoch": 0.06817691477885653, "progress_pct": 1.14, "epoch_pct": 1.14, "eta": "53:46:18", "max_grad_norm": 0.8, "loss": 1.0332108736038208, "grad_norm": 1.762683629989624, "learning_rate": 1.4106019766397124e-05}
+{"ts": "2025-12-23T15:00:46", "event": "train_log", "step": 160, "epoch": 0.06903991370010787, "progress_pct": 1.15, "epoch_pct": 1.15, "eta": "53:33:11", "max_grad_norm": 0.8, "loss": 0.973042905330658, "grad_norm": 1.5698739290237427, "learning_rate": 1.4285714285714285e-05}
+{"ts": "2025-12-23T15:01:05", "event": "train_log", "step": 162, "epoch": 0.06990291262135923, "progress_pct": 1.16, "epoch_pct": 1.17, "eta": "53:19:45", "max_grad_norm": 0.8, "loss": 0.9966566562652588, "grad_norm": 1.6132292747497559, "learning_rate": 1.4465408805031447e-05}
+{"ts": "2025-12-23T15:01:24", "event": "train_log", "step": 164, "epoch": 0.07076591154261057, "progress_pct": 1.18, "epoch_pct": 1.18, "eta": "53:06:56", "max_grad_norm": 0.8, "loss": 1.078708529472351, "grad_norm": 1.989534616470337, "learning_rate": 1.4645103324348609e-05}
+{"ts": "2025-12-23T15:01:45", "event": "train_log", "step": 166, "epoch": 0.07162891046386193, "progress_pct": 1.19, "epoch_pct": 1.19, "eta": "52:57:03", "max_grad_norm": 0.8, "loss": 0.9075545072555542, "grad_norm": 1.8199284076690674, "learning_rate": 1.4824797843665769e-05}
+{"ts": "2025-12-23T15:02:05", "event": "train_log", "step": 168, "epoch": 0.07249190938511327, "progress_pct": 1.21, "epoch_pct": 1.21, "eta": "52:45:47", "max_grad_norm": 0.8, "loss": 0.9268680810928345, "grad_norm": 1.6908831596374512, "learning_rate": 1.500449236298293e-05}
+{"ts": "2025-12-23T15:02:24", "event": "train_log", "step": 170, "epoch": 0.07335490830636461, "progress_pct": 1.22, "epoch_pct": 1.22, "eta": "52:34:04", "max_grad_norm": 0.8, "loss": 1.0361101627349854, "grad_norm": 1.5648242235183716, "learning_rate": 1.5184186882300092e-05}
+{"ts": "2025-12-23T15:02:44", "event": "train_log", "step": 172, "epoch": 0.07421790722761597, "progress_pct": 1.24, "epoch_pct": 1.24, "eta": "52:23:48", "max_grad_norm": 0.8, "loss": 0.9197717905044556, "grad_norm": 1.8292360305786133, "learning_rate": 1.5363881401617252e-05}
+{"ts": "2025-12-23T15:03:03", "event": "train_log", "step": 174, "epoch": 0.07508090614886731, "progress_pct": 1.25, "epoch_pct": 1.25, "eta": "52:12:28", "max_grad_norm": 0.8, "loss": 0.9661652445793152, "grad_norm": 1.6882517337799072, "learning_rate": 1.554357592093441e-05}
+{"ts": "2025-12-23T15:03:22", "event": "train_log", "step": 176, "epoch": 0.07594390507011867, "progress_pct": 1.27, "epoch_pct": 1.27, "eta": "52:00:50", "max_grad_norm": 0.8, "loss": 1.0006932020187378, "grad_norm": 1.695921778678894, "learning_rate": 1.572327044025157e-05}
+{"ts": "2025-12-23T15:03:40", "event": "train_log", "step": 178, "epoch": 0.07680690399137001, "progress_pct": 1.28, "epoch_pct": 1.28, "eta": "51:47:41", "max_grad_norm": 0.8, "loss": 1.0096064805984497, "grad_norm": 2.0128681659698486, "learning_rate": 1.5902964959568733e-05}
+{"ts": "2025-12-23T15:03:59", "event": "train_log", "step": 180, "epoch": 0.07766990291262135, "progress_pct": 1.29, "epoch_pct": 1.29, "eta": "51:37:43", "max_grad_norm": 0.8, "loss": 0.972679853439331, "grad_norm": 1.6833231449127197, "learning_rate": 1.6082659478885895e-05}
+{"ts": "2025-12-23T15:04:20", "event": "train_log", "step": 182, "epoch": 0.07853290183387271, "progress_pct": 1.31, "epoch_pct": 1.31, "eta": "51:28:40", "max_grad_norm": 0.8, "loss": 0.9232968091964722, "grad_norm": 1.6940670013427734, "learning_rate": 1.6262353998203056e-05}
+{"ts": "2025-12-23T15:04:38", "event": "train_log", "step": 184, "epoch": 0.07939590075512405, "progress_pct": 1.32, "epoch_pct": 1.32, "eta": "51:17:13", "max_grad_norm": 0.8, "loss": 0.981245219707489, "grad_norm": 2.0851218700408936, "learning_rate": 1.6442048517520218e-05}
+{"ts": "2025-12-23T15:04:59", "event": "train_log", "step": 186, "epoch": 0.08025889967637541, "progress_pct": 1.34, "epoch_pct": 1.34, "eta": "51:09:22", "max_grad_norm": 0.8, "loss": 0.9762678146362305, "grad_norm": 1.3769028186798096, "learning_rate": 1.662174303683738e-05}
+{"ts": "2025-12-23T15:05:18", "event": "train_log", "step": 188, "epoch": 0.08112189859762675, "progress_pct": 1.35, "epoch_pct": 1.35, "eta": "50:59:20", "max_grad_norm": 0.8, "loss": 0.9905523061752319, "grad_norm": 1.85263192653656, "learning_rate": 1.6801437556154538e-05}
+{"ts": "2025-12-23T15:05:38", "event": "train_log", "step": 190, "epoch": 0.08198489751887811, "progress_pct": 1.37, "epoch_pct": 1.37, "eta": "50:51:08", "max_grad_norm": 0.8, "loss": 0.8457058072090149, "grad_norm": 1.654221534729004, "learning_rate": 1.69811320754717e-05}
+{"ts": "2025-12-23T15:05:56", "event": "train_log", "step": 192, "epoch": 0.08284789644012945, "progress_pct": 1.38, "epoch_pct": 1.38, "eta": "50:40:52", "max_grad_norm": 0.8, "loss": 1.0171988010406494, "grad_norm": 1.7715294361114502, "learning_rate": 1.7160826594788858e-05}
+{"ts": "2025-12-23T15:06:15", "event": "train_log", "step": 194, "epoch": 0.0837108953613808, "progress_pct": 1.39, "epoch_pct": 1.4, "eta": "50:31:28", "max_grad_norm": 0.8, "loss": 0.9767695069313049, "grad_norm": 1.7162128686904907, "learning_rate": 1.734052111410602e-05}
+{"ts": "2025-12-23T15:06:34", "event": "train_log", "step": 196, "epoch": 0.08457389428263215, "progress_pct": 1.41, "epoch_pct": 1.41, "eta": "50:22:26", "max_grad_norm": 0.8, "loss": 0.9786219596862793, "grad_norm": 1.7938538789749146, "learning_rate": 1.752021563342318e-05}
+{"ts": "2025-12-23T15:06:53", "event": "train_log", "step": 198, "epoch": 0.0854368932038835, "progress_pct": 1.42, "epoch_pct": 1.42, "eta": "50:12:56", "max_grad_norm": 0.8, "loss": 0.9941093921661377, "grad_norm": 1.944226861000061, "learning_rate": 1.7699910152740342e-05}
+{"ts": "2025-12-23T15:07:12", "event": "train_log", "step": 200, "epoch": 0.08629989212513485, "progress_pct": 1.44, "epoch_pct": 1.44, "eta": "50:04:17", "max_grad_norm": 0.8, "loss": 0.8924376964569092, "grad_norm": 1.816835880279541, "learning_rate": 1.7879604672057504e-05}
+{"ts": "2025-12-23T15:20:59", "event": "train_log", "step": 200, "epoch": 0.08629989212513485, "progress_pct": 1.44, "epoch_pct": 1.44, "eta": "65:48:22", "max_grad_norm": 0.8, "eval_loss": 0.981117308139801, "eval_runtime": 826.4469, "eval_samples_per_second": 2.493, "eval_steps_per_second": 2.493}
+{"ts": "2025-12-23T15:21:18", "event": "train_log", "step": 202, "epoch": 0.08716289104638619, "progress_pct": 1.45, "epoch_pct": 1.45, "eta": "65:30:29", "max_grad_norm": 0.8, "loss": 0.9908216595649719, "grad_norm": 1.8230189085006714, "learning_rate": 1.8059299191374666e-05}
+{"ts": "2025-12-23T15:21:37", "event": "train_log", "step": 204, "epoch": 0.08802588996763754, "progress_pct": 1.47, "epoch_pct": 1.47, "eta": "65:13:12", "max_grad_norm": 0.8, "loss": 0.9657744765281677, "grad_norm": 1.8223077058792114, "learning_rate": 1.8238993710691824e-05}
+{"ts": "2025-12-23T15:21:57", "event": "train_log", "step": 206, "epoch": 0.08888888888888889, "progress_pct": 1.48, "epoch_pct": 1.48, "eta": "64:55:57", "max_grad_norm": 0.8, "loss": 0.9397313594818115, "grad_norm": 1.811195731163025, "learning_rate": 1.8418688230008986e-05}
+{"ts": "2025-12-23T15:22:18", "event": "train_log", "step": 208, "epoch": 0.08975188781014024, "progress_pct": 1.5, "epoch_pct": 1.5, "eta": "64:41:06", "max_grad_norm": 0.8, "loss": 0.8770087361335754, "grad_norm": 1.579998254776001, "learning_rate": 1.8598382749326147e-05}
+{"ts": "2025-12-23T15:22:37", "event": "train_log", "step": 210, "epoch": 0.09061488673139159, "progress_pct": 1.51, "epoch_pct": 1.51, "eta": "64:24:05", "max_grad_norm": 0.8, "loss": 0.922799825668335, "grad_norm": 1.592136263847351, "learning_rate": 1.8778077268643305e-05}
+{"ts": "2025-12-23T15:22:56", "event": "train_log", "step": 212, "epoch": 0.09147788565264293, "progress_pct": 1.52, "epoch_pct": 1.52, "eta": "64:07:44", "max_grad_norm": 0.8, "loss": 0.9007947444915771, "grad_norm": 1.9334869384765625, "learning_rate": 1.8957771787960467e-05}
+{"ts": "2025-12-23T15:23:15", "event": "train_log", "step": 214, "epoch": 0.09234088457389428, "progress_pct": 1.54, "epoch_pct": 1.54, "eta": "63:52:11", "max_grad_norm": 0.8, "loss": 0.9261049628257751, "grad_norm": 1.8619475364685059, "learning_rate": 1.913746630727763e-05}
+{"ts": "2025-12-23T15:23:35", "event": "train_log", "step": 216, "epoch": 0.09320388349514563, "progress_pct": 1.55, "epoch_pct": 1.55, "eta": "63:36:34", "max_grad_norm": 0.8, "loss": 0.9319002628326416, "grad_norm": 1.751584529876709, "learning_rate": 1.931716082659479e-05}
+{"ts": "2025-12-23T15:23:53", "event": "train_log", "step": 218, "epoch": 0.09406688241639698, "progress_pct": 1.57, "epoch_pct": 1.57, "eta": "63:20:05", "max_grad_norm": 0.8, "loss": 0.947629451751709, "grad_norm": 2.1829185485839844, "learning_rate": 1.9496855345911952e-05}
+{"ts": "2025-12-23T15:24:11", "event": "train_log", "step": 220, "epoch": 0.09492988133764833, "progress_pct": 1.58, "epoch_pct": 1.58, "eta": "63:03:35", "max_grad_norm": 0.8, "loss": 0.9213902354240417, "grad_norm": 1.8973567485809326, "learning_rate": 1.967654986522911e-05}
+{"ts": "2025-12-23T15:24:30", "event": "train_log", "step": 222, "epoch": 0.09579288025889968, "progress_pct": 1.6, "epoch_pct": 1.6, "eta": "62:48:33", "max_grad_norm": 0.8, "loss": 0.9058182835578918, "grad_norm": 1.9995585680007935, "learning_rate": 1.985624438454627e-05}
+{"ts": "2025-12-23T15:24:49", "event": "train_log", "step": 224, "epoch": 0.09665587918015102, "progress_pct": 1.61, "epoch_pct": 1.61, "eta": "62:34:07", "max_grad_norm": 0.8, "loss": 0.8918466567993164, "grad_norm": 1.9369088411331177, "learning_rate": 2.0035938903863433e-05}
+{"ts": "2025-12-23T15:25:09", "event": "train_log", "step": 226, "epoch": 0.09751887810140238, "progress_pct": 1.62, "epoch_pct": 1.63, "eta": "62:20:33", "max_grad_norm": 0.8, "loss": 1.009294033050537, "grad_norm": 2.1926183700561523, "learning_rate": 2.0215633423180595e-05}
+{"ts": "2025-12-23T15:25:28", "event": "train_log", "step": 228, "epoch": 0.09838187702265372, "progress_pct": 1.64, "epoch_pct": 1.64, "eta": "62:05:44", "max_grad_norm": 0.8, "loss": 0.9543544054031372, "grad_norm": 1.872399926185608, "learning_rate": 2.0395327942497757e-05}
+{"ts": "2025-12-23T15:25:48", "event": "train_log", "step": 230, "epoch": 0.09924487594390508, "progress_pct": 1.65, "epoch_pct": 1.65, "eta": "61:53:03", "max_grad_norm": 0.8, "loss": 0.9188834428787231, "grad_norm": 2.011009454727173, "learning_rate": 2.0575022461814915e-05}
+{"ts": "2025-12-23T15:26:08", "event": "train_log", "step": 232, "epoch": 0.10010787486515642, "progress_pct": 1.67, "epoch_pct": 1.67, "eta": "61:40:07", "max_grad_norm": 0.8, "loss": 0.84498131275177, "grad_norm": 1.8422952890396118, "learning_rate": 2.0754716981132076e-05}
+{"ts": "2025-12-23T15:26:27", "event": "train_log", "step": 234, "epoch": 0.10097087378640776, "progress_pct": 1.68, "epoch_pct": 1.68, "eta": "61:26:16", "max_grad_norm": 0.8, "loss": 0.9864624738693237, "grad_norm": 2.208738088607788, "learning_rate": 2.0934411500449238e-05}
+{"ts": "2025-12-23T15:26:47", "event": "train_log", "step": 236, "epoch": 0.10183387270765912, "progress_pct": 1.7, "epoch_pct": 1.7, "eta": "61:13:55", "max_grad_norm": 0.8, "loss": 0.928769588470459, "grad_norm": 1.6119916439056396, "learning_rate": 2.1114106019766396e-05}
+{"ts": "2025-12-23T15:27:07", "event": "train_log", "step": 238, "epoch": 0.10269687162891046, "progress_pct": 1.71, "epoch_pct": 1.71, "eta": "61:01:41", "max_grad_norm": 0.8, "loss": 0.885361909866333, "grad_norm": 1.906336784362793, "learning_rate": 2.1293800539083558e-05}
+{"ts": "2025-12-23T15:27:25", "event": "train_log", "step": 240, "epoch": 0.10355987055016182, "progress_pct": 1.73, "epoch_pct": 1.73, "eta": "60:47:20", "max_grad_norm": 0.8, "loss": 0.9497957229614258, "grad_norm": 2.046064615249634, "learning_rate": 2.147349505840072e-05}
+{"ts": "2025-12-23T15:27:44", "event": "train_log", "step": 242, "epoch": 0.10442286947141316, "progress_pct": 1.74, "epoch_pct": 1.74, "eta": "60:34:34", "max_grad_norm": 0.8, "loss": 0.958332896232605, "grad_norm": 2.1612143516540527, "learning_rate": 2.165318957771788e-05}
+{"ts": "2025-12-23T15:28:02", "event": "train_log", "step": 244, "epoch": 0.1052858683926645, "progress_pct": 1.75, "epoch_pct": 1.75, "eta": "60:21:21", "max_grad_norm": 0.8, "loss": 1.0089390277862549, "grad_norm": 2.139829635620117, "learning_rate": 2.1832884097035043e-05}
+{"ts": "2025-12-23T15:28:21", "event": "train_log", "step": 246, "epoch": 0.10614886731391586, "progress_pct": 1.77, "epoch_pct": 1.77, "eta": "60:08:29", "max_grad_norm": 0.8, "loss": 0.9627355933189392, "grad_norm": 1.9842045307159424, "learning_rate": 2.2012578616352204e-05}
+{"ts": "2025-12-23T15:28:40", "event": "train_log", "step": 248, "epoch": 0.1070118662351672, "progress_pct": 1.78, "epoch_pct": 1.78, "eta": "59:56:57", "max_grad_norm": 0.8, "loss": 0.9260450601577759, "grad_norm": 1.968578577041626, "learning_rate": 2.2192273135669366e-05}
+{"ts": "2025-12-23T15:28:58", "event": "train_log", "step": 250, "epoch": 0.10787486515641856, "progress_pct": 1.8, "epoch_pct": 1.8, "eta": "59:44:03", "max_grad_norm": 0.8, "loss": 1.0654462575912476, "grad_norm": 2.192275047302246, "learning_rate": 2.2371967654986524e-05}
+{"ts": "2025-12-23T15:29:17", "event": "train_log", "step": 252, "epoch": 0.1087378640776699, "progress_pct": 1.81, "epoch_pct": 1.81, "eta": "59:32:15", "max_grad_norm": 0.8, "loss": 0.9680672883987427, "grad_norm": 2.0267577171325684, "learning_rate": 2.2551662174303682e-05}
+{"ts": "2025-12-23T15:29:36", "event": "train_log", "step": 254, "epoch": 0.10960086299892124, "progress_pct": 1.83, "epoch_pct": 1.83, "eta": "59:20:22", "max_grad_norm": 0.8, "loss": 1.0027307271957397, "grad_norm": 1.9834461212158203, "learning_rate": 2.2731356693620844e-05}
+{"ts": "2025-12-23T15:29:55", "event": "train_log", "step": 256, "epoch": 0.1104638619201726, "progress_pct": 1.84, "epoch_pct": 1.84, "eta": "59:08:21", "max_grad_norm": 0.8, "loss": 1.0050861835479736, "grad_norm": 2.0105156898498535, "learning_rate": 2.2911051212938006e-05}
+{"ts": "2025-12-23T15:30:13", "event": "train_log", "step": 258, "epoch": 0.11132686084142394, "progress_pct": 1.86, "epoch_pct": 1.86, "eta": "58:57:00", "max_grad_norm": 0.8, "loss": 0.8955062627792358, "grad_norm": 1.9150307178497314, "learning_rate": 2.3090745732255167e-05}
+{"ts": "2025-12-23T15:30:32", "event": "train_log", "step": 260, "epoch": 0.1121898597626753, "progress_pct": 1.87, "epoch_pct": 1.87, "eta": "58:45:42", "max_grad_norm": 0.8, "loss": 0.941332221031189, "grad_norm": 2.100795269012451, "learning_rate": 2.327044025157233e-05}
+{"ts": "2025-12-23T15:30:51", "event": "train_log", "step": 262, "epoch": 0.11305285868392664, "progress_pct": 1.88, "epoch_pct": 1.88, "eta": "58:34:40", "max_grad_norm": 0.8, "loss": 0.9292652606964111, "grad_norm": 1.837814211845398, "learning_rate": 2.345013477088949e-05}
+{"ts": "2025-12-23T15:31:10", "event": "train_log", "step": 264, "epoch": 0.113915857605178, "progress_pct": 1.9, "epoch_pct": 1.9, "eta": "58:23:46", "max_grad_norm": 0.8, "loss": 0.9460650682449341, "grad_norm": 1.9122955799102783, "learning_rate": 2.3629829290206652e-05}
+{"ts": "2025-12-23T15:31:28", "event": "train_log", "step": 266, "epoch": 0.11477885652642934, "progress_pct": 1.91, "epoch_pct": 1.91, "eta": "58:12:18", "max_grad_norm": 0.8, "loss": 0.8730320334434509, "grad_norm": 1.9168611764907837, "learning_rate": 2.380952380952381e-05}
+{"ts": "2025-12-23T15:31:46", "event": "train_log", "step": 268, "epoch": 0.11564185544768069, "progress_pct": 1.93, "epoch_pct": 1.93, "eta": "58:01:05", "max_grad_norm": 0.8, "loss": 0.9194066524505615, "grad_norm": 2.1184403896331787, "learning_rate": 2.3989218328840972e-05}
+{"ts": "2025-12-23T15:32:05", "event": "train_log", "step": 270, "epoch": 0.11650485436893204, "progress_pct": 1.94, "epoch_pct": 1.94, "eta": "57:51:06", "max_grad_norm": 0.8, "loss": 0.8994337320327759, "grad_norm": 1.910470962524414, "learning_rate": 2.4168912848158133e-05}
+{"ts": "2025-12-23T15:32:25", "event": "train_log", "step": 272, "epoch": 0.11736785329018339, "progress_pct": 1.96, "epoch_pct": 1.96, "eta": "57:41:01", "max_grad_norm": 0.8, "loss": 0.8509554266929626, "grad_norm": 1.9906529188156128, "learning_rate": 2.434860736747529e-05}
+{"ts": "2025-12-23T15:32:42", "event": "train_log", "step": 274, "epoch": 0.11823085221143474, "progress_pct": 1.97, "epoch_pct": 1.97, "eta": "57:29:59", "max_grad_norm": 0.8, "loss": 1.0223658084869385, "grad_norm": 2.5982062816619873, "learning_rate": 2.4528301886792453e-05}
+{"ts": "2025-12-23T15:33:03", "event": "train_log", "step": 276, "epoch": 0.11909385113268608, "progress_pct": 1.98, "epoch_pct": 1.98, "eta": "57:21:15", "max_grad_norm": 0.8, "loss": 0.8702670931816101, "grad_norm": 1.8959418535232544, "learning_rate": 2.4707996406109615e-05}
+{"ts": "2025-12-23T15:33:21", "event": "train_log", "step": 278, "epoch": 0.11995685005393743, "progress_pct": 2.0, "epoch_pct": 2.0, "eta": "57:11:19", "max_grad_norm": 0.8, "loss": 0.9471842050552368, "grad_norm": 2.115907669067383, "learning_rate": 2.4887690925426776e-05}
+{"ts": "2025-12-23T15:33:41", "event": "train_log", "step": 280, "epoch": 0.12081984897518878, "progress_pct": 2.01, "epoch_pct": 2.01, "eta": "57:02:12", "max_grad_norm": 0.8, "loss": 0.805242657661438, "grad_norm": 2.0255353450775146, "learning_rate": 2.5067385444743935e-05}
+{"ts": "2025-12-23T15:34:00", "event": "train_log", "step": 282, "epoch": 0.12168284789644013, "progress_pct": 2.03, "epoch_pct": 2.03, "eta": "56:52:32", "max_grad_norm": 0.8, "loss": 0.9008285999298096, "grad_norm": 1.9394105672836304, "learning_rate": 2.52470799640611e-05}
+{"ts": "2025-12-23T15:34:17", "event": "train_log", "step": 284, "epoch": 0.12254584681769148, "progress_pct": 2.04, "epoch_pct": 2.04, "eta": "56:42:10", "max_grad_norm": 0.8, "loss": 0.950639009475708, "grad_norm": 2.2017428874969482, "learning_rate": 2.5426774483378258e-05}
+{"ts": "2025-12-23T15:34:36", "event": "train_log", "step": 286, "epoch": 0.12340884573894283, "progress_pct": 2.06, "epoch_pct": 2.06, "eta": "56:32:17", "max_grad_norm": 0.8, "loss": 0.9673396944999695, "grad_norm": 4.048379421234131, "learning_rate": 2.5606469002695423e-05}
+{"ts": "2025-12-23T15:34:56", "event": "train_log", "step": 288, "epoch": 0.12427184466019417, "progress_pct": 2.07, "epoch_pct": 2.07, "eta": "56:24:23", "max_grad_norm": 0.8, "loss": 0.8251164555549622, "grad_norm": 1.812450647354126, "learning_rate": 2.578616352201258e-05}
+{"ts": "2025-12-23T15:35:15", "event": "train_log", "step": 290, "epoch": 0.12513484358144553, "progress_pct": 2.09, "epoch_pct": 2.09, "eta": "56:15:07", "max_grad_norm": 0.8, "loss": 0.9397821426391602, "grad_norm": 1.9298596382141113, "learning_rate": 2.596585804132974e-05}
+{"ts": "2025-12-23T15:35:33", "event": "train_log", "step": 292, "epoch": 0.12599784250269688, "progress_pct": 2.1, "epoch_pct": 2.1, "eta": "56:05:51", "max_grad_norm": 0.8, "loss": 0.95896315574646, "grad_norm": 2.0862202644348145, "learning_rate": 2.61455525606469e-05}
+{"ts": "2025-12-23T15:35:52", "event": "train_log", "step": 294, "epoch": 0.1268608414239482, "progress_pct": 2.11, "epoch_pct": 2.11, "eta": "55:56:55", "max_grad_norm": 0.8, "loss": 0.9547653198242188, "grad_norm": 2.113429307937622, "learning_rate": 2.632524707996406e-05}
+{"ts": "2025-12-23T15:36:11", "event": "train_log", "step": 296, "epoch": 0.12772384034519957, "progress_pct": 2.13, "epoch_pct": 2.13, "eta": "55:48:19", "max_grad_norm": 0.8, "loss": 0.9351654052734375, "grad_norm": 2.0666873455047607, "learning_rate": 2.6504941599281224e-05}
+{"ts": "2025-12-23T15:36:29", "event": "train_log", "step": 298, "epoch": 0.12858683926645093, "progress_pct": 2.14, "epoch_pct": 2.14, "eta": "55:39:24", "max_grad_norm": 0.8, "loss": 0.9268350005149841, "grad_norm": 2.0679547786712646, "learning_rate": 2.6684636118598382e-05}
+{"ts": "2025-12-23T15:36:50", "event": "train_log", "step": 300, "epoch": 0.12944983818770225, "progress_pct": 2.16, "epoch_pct": 2.16, "eta": "55:31:53", "max_grad_norm": 0.8, "loss": 0.8522316813468933, "grad_norm": 2.028594493865967, "learning_rate": 2.6864330637915547e-05}
+{"ts": "2025-12-23T15:50:45", "event": "train_log", "step": 300, "epoch": 0.12944983818770225, "progress_pct": 2.16, "epoch_pct": 2.16, "eta": "66:03:30", "max_grad_norm": 0.8, "eval_loss": 0.9384957551956177, "eval_runtime": 835.4727, "eval_samples_per_second": 2.466, "eval_steps_per_second": 2.466}
+{"ts": "2025-12-23T15:51:03", "event": "train_log", "step": 302, "epoch": 0.1303128371089536, "progress_pct": 2.17, "epoch_pct": 2.17, "eta": "65:49:48", "max_grad_norm": 0.8, "loss": 0.9460103511810303, "grad_norm": 2.1932919025421143, "learning_rate": 2.7044025157232706e-05}
+{"ts": "2025-12-23T15:51:21", "event": "train_log", "step": 304, "epoch": 0.13117583603020497, "progress_pct": 2.19, "epoch_pct": 2.19, "eta": "65:37:08", "max_grad_norm": 0.8, "loss": 0.9642201662063599, "grad_norm": 2.275789737701416, "learning_rate": 2.722371967654987e-05}
+{"ts": "2025-12-23T15:51:42", "event": "train_log", "step": 306, "epoch": 0.13203883495145632, "progress_pct": 2.2, "epoch_pct": 2.2, "eta": "65:26:27", "max_grad_norm": 0.8, "loss": 0.869883120059967, "grad_norm": 1.9120908975601196, "learning_rate": 2.740341419586703e-05}
+{"ts": "2025-12-23T15:52:02", "event": "train_log", "step": 308, "epoch": 0.13290183387270765, "progress_pct": 2.21, "epoch_pct": 2.22, "eta": "65:15:02", "max_grad_norm": 0.8, "loss": 0.8595444560050964, "grad_norm": 1.9676507711410522, "learning_rate": 2.7583108715184187e-05}
+{"ts": "2025-12-23T15:52:21", "event": "train_log", "step": 310, "epoch": 0.133764832793959, "progress_pct": 2.23, "epoch_pct": 2.23, "eta": "65:03:20", "max_grad_norm": 0.8, "loss": 0.8541103601455688, "grad_norm": 1.7901391983032227, "learning_rate": 2.776280323450135e-05}
+{"ts": "2025-12-23T15:52:40", "event": "train_log", "step": 312, "epoch": 0.13462783171521037, "progress_pct": 2.24, "epoch_pct": 2.24, "eta": "64:51:19", "max_grad_norm": 0.8, "loss": 1.0146855115890503, "grad_norm": 2.160867691040039, "learning_rate": 2.7942497753818507e-05}
+{"ts": "2025-12-23T15:52:58", "event": "train_log", "step": 314, "epoch": 0.1354908306364617, "progress_pct": 2.26, "epoch_pct": 2.26, "eta": "64:39:11", "max_grad_norm": 0.8, "loss": 0.9228161573410034, "grad_norm": 2.0809009075164795, "learning_rate": 2.8122192273135672e-05}
+{"ts": "2025-12-23T15:53:17", "event": "train_log", "step": 316, "epoch": 0.13635382955771305, "progress_pct": 2.27, "epoch_pct": 2.27, "eta": "64:27:26", "max_grad_norm": 0.8, "loss": 0.8527485132217407, "grad_norm": 1.9626725912094116, "learning_rate": 2.830188679245283e-05}
+{"ts": "2025-12-23T15:53:38", "event": "train_log", "step": 318, "epoch": 0.1372168284789644, "progress_pct": 2.29, "epoch_pct": 2.29, "eta": "64:17:07", "max_grad_norm": 0.8, "loss": 0.8136687278747559, "grad_norm": 1.9569967985153198, "learning_rate": 2.8481581311769995e-05}
+{"ts": "2025-12-23T15:53:56", "event": "train_log", "step": 320, "epoch": 0.13807982740021574, "progress_pct": 2.3, "epoch_pct": 2.3, "eta": "64:05:39", "max_grad_norm": 0.8, "loss": 0.8999513387680054, "grad_norm": 2.301692247390747, "learning_rate": 2.8661275831087153e-05}
+{"ts": "2025-12-23T15:54:15", "event": "train_log", "step": 322, "epoch": 0.1389428263214671, "progress_pct": 2.32, "epoch_pct": 2.32, "eta": "63:54:16", "max_grad_norm": 0.8, "loss": 0.9751577973365784, "grad_norm": 2.0247251987457275, "learning_rate": 2.884097035040431e-05}
+{"ts": "2025-12-23T15:54:35", "event": "train_log", "step": 324, "epoch": 0.13980582524271845, "progress_pct": 2.33, "epoch_pct": 2.33, "eta": "63:44:00", "max_grad_norm": 0.8, "loss": 0.8741729259490967, "grad_norm": 2.019547462463379, "learning_rate": 2.9020664869721477e-05}
+{"ts": "2025-12-23T15:54:53", "event": "train_log", "step": 326, "epoch": 0.1406688241639698, "progress_pct": 2.34, "epoch_pct": 2.34, "eta": "63:32:24", "max_grad_norm": 0.8, "loss": 0.948066234588623, "grad_norm": 2.3503618240356445, "learning_rate": 2.9200359389038635e-05}
+{"ts": "2025-12-23T15:55:11", "event": "train_log", "step": 328, "epoch": 0.14153182308522114, "progress_pct": 2.36, "epoch_pct": 2.36, "eta": "63:20:55", "max_grad_norm": 0.8, "loss": 1.008673906326294, "grad_norm": 2.544365406036377, "learning_rate": 2.9380053908355796e-05}
+{"ts": "2025-12-23T15:55:30", "event": "train_log", "step": 330, "epoch": 0.1423948220064725, "progress_pct": 2.37, "epoch_pct": 2.37, "eta": "63:10:26", "max_grad_norm": 0.8, "loss": 0.9492161273956299, "grad_norm": 3.1647145748138428, "learning_rate": 2.9559748427672958e-05}
+{"ts": "2025-12-23T15:55:48", "event": "train_log", "step": 332, "epoch": 0.14325782092772385, "progress_pct": 2.39, "epoch_pct": 2.39, "eta": "62:59:47", "max_grad_norm": 0.8, "loss": 0.9077466130256653, "grad_norm": 2.0551259517669678, "learning_rate": 2.973944294699012e-05}
+{"ts": "2025-12-23T15:56:08", "event": "train_log", "step": 334, "epoch": 0.14412081984897518, "progress_pct": 2.4, "epoch_pct": 2.4, "eta": "62:49:55", "max_grad_norm": 0.8, "loss": 0.8839989304542542, "grad_norm": 1.6792620420455933, "learning_rate": 2.9919137466307278e-05}
+{"ts": "2025-12-23T15:56:28", "event": "train_log", "step": 336, "epoch": 0.14498381877022654, "progress_pct": 2.42, "epoch_pct": 2.42, "eta": "62:40:05", "max_grad_norm": 0.8, "loss": 0.855834424495697, "grad_norm": 1.7771215438842773, "learning_rate": 3.0098831985624436e-05}
+{"ts": "2025-12-23T15:56:48", "event": "train_log", "step": 338, "epoch": 0.1458468176914779, "progress_pct": 2.43, "epoch_pct": 2.43, "eta": "62:30:49", "max_grad_norm": 0.8, "loss": 0.8477111458778381, "grad_norm": 2.030884027481079, "learning_rate": 3.02785265049416e-05}
+{"ts": "2025-12-23T15:57:06", "event": "train_log", "step": 340, "epoch": 0.14670981661272922, "progress_pct": 2.44, "epoch_pct": 2.45, "eta": "62:20:33", "max_grad_norm": 0.8, "loss": 0.9139004945755005, "grad_norm": 2.4177005290985107, "learning_rate": 3.045822102425876e-05}
+{"ts": "2025-12-23T15:57:25", "event": "train_log", "step": 342, "epoch": 0.14757281553398058, "progress_pct": 2.46, "epoch_pct": 2.46, "eta": "62:10:28", "max_grad_norm": 0.8, "loss": 0.8836029767990112, "grad_norm": 2.094775438308716, "learning_rate": 3.063791554357592e-05}
+{"ts": "2025-12-23T15:57:44", "event": "train_log", "step": 344, "epoch": 0.14843581445523193, "progress_pct": 2.47, "epoch_pct": 2.47, "eta": "62:00:44", "max_grad_norm": 0.8, "loss": 0.9318807125091553, "grad_norm": 2.0046238899230957, "learning_rate": 3.081761006289308e-05}
+{"ts": "2025-12-23T15:58:03", "event": "train_log", "step": 346, "epoch": 0.1492988133764833, "progress_pct": 2.49, "epoch_pct": 2.49, "eta": "61:51:23", "max_grad_norm": 0.8, "loss": 0.8916489481925964, "grad_norm": 2.40739369392395, "learning_rate": 3.0997304582210244e-05}
+{"ts": "2025-12-23T15:58:23", "event": "train_log", "step": 348, "epoch": 0.15016181229773462, "progress_pct": 2.5, "epoch_pct": 2.5, "eta": "61:42:26", "max_grad_norm": 0.8, "loss": 0.90766441822052, "grad_norm": 2.1190497875213623, "learning_rate": 3.11769991015274e-05}
+{"ts": "2025-12-23T15:58:41", "event": "train_log", "step": 350, "epoch": 0.15102481121898598, "progress_pct": 2.52, "epoch_pct": 2.52, "eta": "61:32:00", "max_grad_norm": 0.8, "loss": 0.9210623502731323, "grad_norm": 2.721255302429199, "learning_rate": 3.135669362084457e-05}
+{"ts": "2025-12-23T15:58:59", "event": "train_log", "step": 352, "epoch": 0.15188781014023733, "progress_pct": 2.53, "epoch_pct": 2.53, "eta": "61:22:05", "max_grad_norm": 0.8, "loss": 0.9134353399276733, "grad_norm": 2.4988787174224854, "learning_rate": 3.1536388140161726e-05}
+{"ts": "2025-12-23T15:59:18", "event": "train_log", "step": 354, "epoch": 0.15275080906148866, "progress_pct": 2.55, "epoch_pct": 2.55, "eta": "61:12:38", "max_grad_norm": 0.8, "loss": 0.9127986431121826, "grad_norm": 2.6075687408447266, "learning_rate": 3.1716082659478884e-05}
+{"ts": "2025-12-23T15:59:38", "event": "train_log", "step": 356, "epoch": 0.15361380798274002, "progress_pct": 2.56, "epoch_pct": 2.56, "eta": "61:04:20", "max_grad_norm": 0.8, "loss": 0.8019667863845825, "grad_norm": 1.9883074760437012, "learning_rate": 3.189577717879605e-05}
+{"ts": "2025-12-23T15:59:57", "event": "train_log", "step": 358, "epoch": 0.15447680690399138, "progress_pct": 2.57, "epoch_pct": 2.57, "eta": "60:55:36", "max_grad_norm": 0.8, "loss": 0.8417298197746277, "grad_norm": 2.058591365814209, "learning_rate": 3.207547169811321e-05}
+{"ts": "2025-12-23T16:00:15", "event": "train_log", "step": 360, "epoch": 0.1553398058252427, "progress_pct": 2.59, "epoch_pct": 2.59, "eta": "60:46:08", "max_grad_norm": 0.8, "loss": 0.909717321395874, "grad_norm": 2.0510308742523193, "learning_rate": 3.225516621743037e-05}
+{"ts": "2025-12-23T16:00:34", "event": "train_log", "step": 362, "epoch": 0.15620280474649406, "progress_pct": 2.6, "epoch_pct": 2.6, "eta": "60:37:03", "max_grad_norm": 0.8, "loss": 0.9414666295051575, "grad_norm": 2.3204104900360107, "learning_rate": 3.243486073674753e-05}
+{"ts": "2025-12-23T16:00:54", "event": "train_log", "step": 364, "epoch": 0.15706580366774542, "progress_pct": 2.62, "epoch_pct": 2.62, "eta": "60:28:52", "max_grad_norm": 0.8, "loss": 0.8631396889686584, "grad_norm": 1.9658725261688232, "learning_rate": 3.2614555256064695e-05}
+{"ts": "2025-12-23T16:01:12", "event": "train_log", "step": 366, "epoch": 0.15792880258899678, "progress_pct": 2.63, "epoch_pct": 2.63, "eta": "60:19:47", "max_grad_norm": 0.8, "loss": 0.8996705412864685, "grad_norm": 2.2223057746887207, "learning_rate": 3.2794249775381853e-05}
+{"ts": "2025-12-23T16:01:31", "event": "train_log", "step": 368, "epoch": 0.1587918015102481, "progress_pct": 2.65, "epoch_pct": 2.65, "eta": "60:11:14", "max_grad_norm": 0.8, "loss": 0.8185732364654541, "grad_norm": 1.9815107583999634, "learning_rate": 3.297394429469901e-05}
+{"ts": "2025-12-23T16:01:51", "event": "train_log", "step": 370, "epoch": 0.15965480043149946, "progress_pct": 2.66, "epoch_pct": 2.66, "eta": "60:03:05", "max_grad_norm": 0.8, "loss": 0.9952898025512695, "grad_norm": 2.319581985473633, "learning_rate": 3.315363881401618e-05}
+{"ts": "2025-12-23T16:02:12", "event": "train_log", "step": 372, "epoch": 0.16051779935275082, "progress_pct": 2.67, "epoch_pct": 2.68, "eta": "59:56:00", "max_grad_norm": 0.8, "loss": 0.7604310512542725, "grad_norm": 1.7916827201843262, "learning_rate": 3.3333333333333335e-05}
+{"ts": "2025-12-23T16:02:31", "event": "train_log", "step": 374, "epoch": 0.16138079827400215, "progress_pct": 2.69, "epoch_pct": 2.69, "eta": "59:47:54", "max_grad_norm": 0.8, "loss": 0.7867834568023682, "grad_norm": 1.955737590789795, "learning_rate": 3.35130278526505e-05}
+{"ts": "2025-12-23T16:02:51", "event": "train_log", "step": 376, "epoch": 0.1622437971952535, "progress_pct": 2.7, "epoch_pct": 2.7, "eta": "59:40:00", "max_grad_norm": 0.8, "loss": 0.9106539487838745, "grad_norm": 2.409783124923706, "learning_rate": 3.369272237196766e-05}
+{"ts": "2025-12-23T16:03:10", "event": "train_log", "step": 378, "epoch": 0.16310679611650486, "progress_pct": 2.72, "epoch_pct": 2.72, "eta": "59:32:06", "max_grad_norm": 0.8, "loss": 0.8945732712745667, "grad_norm": 2.2521097660064697, "learning_rate": 3.3872416891284816e-05}
+{"ts": "2025-12-23T16:03:30", "event": "train_log", "step": 380, "epoch": 0.16396979503775622, "progress_pct": 2.73, "epoch_pct": 2.73, "eta": "59:24:33", "max_grad_norm": 0.8, "loss": 0.8950189352035522, "grad_norm": 2.357815742492676, "learning_rate": 3.4052111410601975e-05}
+{"ts": "2025-12-23T16:03:48", "event": "train_log", "step": 382, "epoch": 0.16483279395900755, "progress_pct": 2.75, "epoch_pct": 2.75, "eta": "59:15:58", "max_grad_norm": 0.8, "loss": 0.9965803623199463, "grad_norm": 2.347301721572876, "learning_rate": 3.423180592991914e-05}
+{"ts": "2025-12-23T16:04:08", "event": "train_log", "step": 384, "epoch": 0.1656957928802589, "progress_pct": 2.76, "epoch_pct": 2.76, "eta": "59:08:40", "max_grad_norm": 0.8, "loss": 0.9081383347511292, "grad_norm": 2.054586887359619, "learning_rate": 3.44115004492363e-05}
+{"ts": "2025-12-23T16:04:28", "event": "train_log", "step": 386, "epoch": 0.16655879180151026, "progress_pct": 2.78, "epoch_pct": 2.78, "eta": "59:01:43", "max_grad_norm": 0.8, "loss": 0.874001681804657, "grad_norm": 2.1550471782684326, "learning_rate": 3.4591194968553456e-05}
+{"ts": "2025-12-23T16:04:49", "event": "train_log", "step": 388, "epoch": 0.1674217907227616, "progress_pct": 2.79, "epoch_pct": 2.79, "eta": "58:54:45", "max_grad_norm": 0.8, "loss": 0.9111959934234619, "grad_norm": 1.6929951906204224, "learning_rate": 3.477088948787062e-05}
+{"ts": "2025-12-23T16:05:09", "event": "train_log", "step": 390, "epoch": 0.16828478964401294, "progress_pct": 2.8, "epoch_pct": 2.8, "eta": "58:48:04", "max_grad_norm": 0.8, "loss": 0.8915985226631165, "grad_norm": 2.0102803707122803, "learning_rate": 3.495058400718778e-05}
+{"ts": "2025-12-23T16:05:29", "event": "train_log", "step": 392, "epoch": 0.1691477885652643, "progress_pct": 2.82, "epoch_pct": 2.82, "eta": "58:40:52", "max_grad_norm": 0.8, "loss": 1.0158213376998901, "grad_norm": 4.215941429138184, "learning_rate": 3.5130278526504944e-05}
+{"ts": "2025-12-23T16:05:49", "event": "train_log", "step": 394, "epoch": 0.17001078748651563, "progress_pct": 2.83, "epoch_pct": 2.83, "eta": "58:34:00", "max_grad_norm": 0.8, "loss": 0.7811702489852905, "grad_norm": 1.864119529724121, "learning_rate": 3.53099730458221e-05}
+{"ts": "2025-12-23T16:06:09", "event": "train_log", "step": 396, "epoch": 0.170873786407767, "progress_pct": 2.85, "epoch_pct": 2.85, "eta": "58:26:49", "max_grad_norm": 0.8, "loss": 0.931502103805542, "grad_norm": 1.9778105020523071, "learning_rate": 3.548966756513927e-05}
+{"ts": "2025-12-23T16:06:28", "event": "train_log", "step": 398, "epoch": 0.17173678532901834, "progress_pct": 2.86, "epoch_pct": 2.86, "eta": "58:19:42", "max_grad_norm": 0.8, "loss": 0.9409074187278748, "grad_norm": 2.0757675170898438, "learning_rate": 3.5669362084456426e-05}
+{"ts": "2025-12-23T16:06:50", "event": "train_log", "step": 400, "epoch": 0.1725997842502697, "progress_pct": 2.88, "epoch_pct": 2.88, "eta": "58:14:00", "max_grad_norm": 0.8, "loss": 0.8159269094467163, "grad_norm": 1.5847891569137573, "learning_rate": 3.5849056603773584e-05}
+{"ts": "2025-12-23T16:20:43", "event": "train_log", "step": 400, "epoch": 0.1725997842502697, "progress_pct": 2.88, "epoch_pct": 2.88, "eta": "66:02:38", "max_grad_norm": 0.8, "eval_loss": 0.9124138355255127, "eval_runtime": 832.6331, "eval_samples_per_second": 2.474, "eval_steps_per_second": 2.474}
+{"ts": "2025-12-23T16:21:02", "event": "train_log", "step": 402, "epoch": 0.17346278317152103, "progress_pct": 2.89, "epoch_pct": 2.89, "eta": "65:53:12", "max_grad_norm": 0.8, "loss": 0.8901475667953491, "grad_norm": 1.8158049583435059, "learning_rate": 3.602875112309075e-05}
+{"ts": "2025-12-23T16:21:21", "event": "train_log", "step": 404, "epoch": 0.17432578209277239, "progress_pct": 2.9, "epoch_pct": 2.91, "eta": "65:43:41", "max_grad_norm": 0.8, "loss": 0.8330907821655273, "grad_norm": 1.9448504447937012, "learning_rate": 3.620844564240791e-05}
+{"ts": "2025-12-23T16:21:40", "event": "train_log", "step": 406, "epoch": 0.17518878101402374, "progress_pct": 2.92, "epoch_pct": 2.92, "eta": "65:34:12", "max_grad_norm": 0.8, "loss": 0.870513916015625, "grad_norm": 2.028280019760132, "learning_rate": 3.638814016172507e-05}
+{"ts": "2025-12-23T16:21:59", "event": "train_log", "step": 408, "epoch": 0.17605177993527507, "progress_pct": 2.93, "epoch_pct": 2.93, "eta": "65:24:54", "max_grad_norm": 0.8, "loss": 0.8064979910850525, "grad_norm": 2.2005157470703125, "learning_rate": 3.656783468104223e-05}
+{"ts": "2025-12-23T16:22:19", "event": "train_log", "step": 410, "epoch": 0.17691477885652643, "progress_pct": 2.95, "epoch_pct": 2.95, "eta": "65:15:40", "max_grad_norm": 0.8, "loss": 0.8457443714141846, "grad_norm": 2.2756290435791016, "learning_rate": 3.6747529200359395e-05}
+{"ts": "2025-12-23T16:22:39", "event": "train_log", "step": 412, "epoch": 0.17777777777777778, "progress_pct": 2.96, "epoch_pct": 2.96, "eta": "65:07:04", "max_grad_norm": 0.8, "loss": 0.7478897571563721, "grad_norm": 2.0970261096954346, "learning_rate": 3.6927223719676554e-05}
+{"ts": "2025-12-23T16:22:58", "event": "train_log", "step": 414, "epoch": 0.1786407766990291, "progress_pct": 2.98, "epoch_pct": 2.98, "eta": "64:58:12", "max_grad_norm": 0.8, "loss": 0.863377034664154, "grad_norm": 1.8917485475540161, "learning_rate": 3.710691823899371e-05}
+{"ts": "2025-12-23T16:23:17", "event": "train_log", "step": 416, "epoch": 0.17950377562028047, "progress_pct": 2.99, "epoch_pct": 2.99, "eta": "64:49:18", "max_grad_norm": 0.8, "loss": 0.8654620051383972, "grad_norm": 2.0429933071136475, "learning_rate": 3.728661275831088e-05}
+{"ts": "2025-12-23T16:23:37", "event": "train_log", "step": 418, "epoch": 0.18036677454153183, "progress_pct": 3.01, "epoch_pct": 3.01, "eta": "64:40:48", "max_grad_norm": 0.8, "loss": 0.8295455574989319, "grad_norm": 2.005125045776367, "learning_rate": 3.7466307277628035e-05}
+{"ts": "2025-12-23T16:23:56", "event": "train_log", "step": 420, "epoch": 0.18122977346278318, "progress_pct": 3.02, "epoch_pct": 3.02, "eta": "64:31:47", "max_grad_norm": 0.8, "loss": 0.8728633522987366, "grad_norm": 2.3852169513702393, "learning_rate": 3.764600179694519e-05}
+{"ts": "2025-12-23T16:24:16", "event": "train_log", "step": 422, "epoch": 0.1820927723840345, "progress_pct": 3.03, "epoch_pct": 3.03, "eta": "64:23:14", "max_grad_norm": 0.8, "loss": 0.8217283487319946, "grad_norm": 1.8731592893600464, "learning_rate": 3.782569631626235e-05}
+{"ts": "2025-12-23T16:24:34", "event": "train_log", "step": 424, "epoch": 0.18295577130528587, "progress_pct": 3.05, "epoch_pct": 3.05, "eta": "64:14:23", "max_grad_norm": 0.8, "loss": 0.8869860172271729, "grad_norm": 2.2818236351013184, "learning_rate": 3.8005390835579516e-05}
+{"ts": "2025-12-23T16:24:53", "event": "train_log", "step": 426, "epoch": 0.18381877022653723, "progress_pct": 3.06, "epoch_pct": 3.06, "eta": "64:05:27", "max_grad_norm": 0.8, "loss": 0.9732091426849365, "grad_norm": 2.0646233558654785, "learning_rate": 3.8185085354896675e-05}
+{"ts": "2025-12-23T16:25:13", "event": "train_log", "step": 428, "epoch": 0.18468176914778855, "progress_pct": 3.08, "epoch_pct": 3.08, "eta": "63:57:31", "max_grad_norm": 0.8, "loss": 0.9012333154678345, "grad_norm": 1.9991949796676636, "learning_rate": 3.836477987421384e-05}
+{"ts": "2025-12-23T16:25:32", "event": "train_log", "step": 430, "epoch": 0.1855447680690399, "progress_pct": 3.09, "epoch_pct": 3.09, "eta": "63:49:03", "max_grad_norm": 0.8, "loss": 0.867533802986145, "grad_norm": 2.209775447845459, "learning_rate": 3.8544474393531e-05}
+{"ts": "2025-12-23T16:25:51", "event": "train_log", "step": 432, "epoch": 0.18640776699029127, "progress_pct": 3.11, "epoch_pct": 3.11, "eta": "63:40:23", "max_grad_norm": 0.8, "loss": 0.8697446584701538, "grad_norm": 2.1242146492004395, "learning_rate": 3.8724168912848156e-05}
+{"ts": "2025-12-23T16:26:11", "event": "train_log", "step": 434, "epoch": 0.1872707659115426, "progress_pct": 3.12, "epoch_pct": 3.12, "eta": "63:32:36", "max_grad_norm": 0.8, "loss": 0.8712933659553528, "grad_norm": 1.771337628364563, "learning_rate": 3.890386343216532e-05}
+{"ts": "2025-12-23T16:26:29", "event": "train_log", "step": 436, "epoch": 0.18813376483279395, "progress_pct": 3.13, "epoch_pct": 3.14, "eta": "63:24:13", "max_grad_norm": 0.8, "loss": 0.8638070821762085, "grad_norm": 2.133652925491333, "learning_rate": 3.908355795148248e-05}
+{"ts": "2025-12-23T16:26:50", "event": "train_log", "step": 438, "epoch": 0.1889967637540453, "progress_pct": 3.15, "epoch_pct": 3.15, "eta": "63:17:01", "max_grad_norm": 0.8, "loss": 0.8255205750465393, "grad_norm": 1.9354501962661743, "learning_rate": 3.9263252470799644e-05}
+{"ts": "2025-12-23T16:27:10", "event": "train_log", "step": 440, "epoch": 0.18985976267529667, "progress_pct": 3.16, "epoch_pct": 3.16, "eta": "63:09:10", "max_grad_norm": 0.8, "loss": 0.9652490019798279, "grad_norm": 1.9564658403396606, "learning_rate": 3.94429469901168e-05}
+{"ts": "2025-12-23T16:27:30", "event": "train_log", "step": 442, "epoch": 0.190722761596548, "progress_pct": 3.18, "epoch_pct": 3.18, "eta": "63:01:52", "max_grad_norm": 0.8, "loss": 0.8960518836975098, "grad_norm": 1.9733822345733643, "learning_rate": 3.962264150943397e-05}
+{"ts": "2025-12-23T16:27:50", "event": "train_log", "step": 444, "epoch": 0.19158576051779935, "progress_pct": 3.19, "epoch_pct": 3.19, "eta": "62:53:59", "max_grad_norm": 0.8, "loss": 0.9555241465568542, "grad_norm": 2.1670451164245605, "learning_rate": 3.9802336028751126e-05}
+{"ts": "2025-12-23T16:28:09", "event": "train_log", "step": 446, "epoch": 0.1924487594390507, "progress_pct": 3.21, "epoch_pct": 3.21, "eta": "62:46:32", "max_grad_norm": 0.8, "loss": 0.8238654732704163, "grad_norm": 2.0059614181518555, "learning_rate": 3.9982030548068284e-05}
+{"ts": "2025-12-23T16:28:30", "event": "train_log", "step": 448, "epoch": 0.19331175836030204, "progress_pct": 3.22, "epoch_pct": 3.22, "eta": "62:39:27", "max_grad_norm": 0.8, "loss": 0.8701026439666748, "grad_norm": 1.9803026914596558, "learning_rate": 4.016172506738545e-05}
+{"ts": "2025-12-23T16:28:50", "event": "train_log", "step": 450, "epoch": 0.1941747572815534, "progress_pct": 3.24, "epoch_pct": 3.24, "eta": "62:31:56", "max_grad_norm": 0.8, "loss": 0.8927714228630066, "grad_norm": 1.8800172805786133, "learning_rate": 4.034141958670261e-05}
+{"ts": "2025-12-23T16:29:11", "event": "train_log", "step": 452, "epoch": 0.19503775620280475, "progress_pct": 3.25, "epoch_pct": 3.25, "eta": "62:25:21", "max_grad_norm": 0.8, "loss": 0.8635751008987427, "grad_norm": 1.8555818796157837, "learning_rate": 4.052111410601977e-05}
+{"ts": "2025-12-23T16:29:31", "event": "train_log", "step": 454, "epoch": 0.1959007551240561, "progress_pct": 3.26, "epoch_pct": 3.27, "eta": "62:18:05", "max_grad_norm": 0.8, "loss": 0.925403892993927, "grad_norm": 2.5119166374206543, "learning_rate": 4.070080862533693e-05}
+{"ts": "2025-12-23T16:29:50", "event": "train_log", "step": 456, "epoch": 0.19676375404530744, "progress_pct": 3.28, "epoch_pct": 3.28, "eta": "62:10:31", "max_grad_norm": 0.8, "loss": 0.8917028307914734, "grad_norm": 2.0825788974761963, "learning_rate": 4.088050314465409e-05}
+{"ts": "2025-12-23T16:30:09", "event": "train_log", "step": 458, "epoch": 0.1976267529665588, "progress_pct": 3.29, "epoch_pct": 3.29, "eta": "62:03:17", "max_grad_norm": 0.8, "loss": 0.9097893834114075, "grad_norm": 2.4413394927978516, "learning_rate": 4.1060197663971254e-05}
+{"ts": "2025-12-23T16:30:29", "event": "train_log", "step": 460, "epoch": 0.19848975188781015, "progress_pct": 3.31, "epoch_pct": 3.31, "eta": "61:56:03", "max_grad_norm": 0.8, "loss": 0.8284633755683899, "grad_norm": 2.503617286682129, "learning_rate": 4.123989218328841e-05}
+{"ts": "2025-12-23T16:30:48", "event": "train_log", "step": 462, "epoch": 0.19935275080906148, "progress_pct": 3.32, "epoch_pct": 3.32, "eta": "61:48:43", "max_grad_norm": 0.8, "loss": 0.8743083477020264, "grad_norm": 1.9360398054122925, "learning_rate": 4.141958670260557e-05}
+{"ts": "2025-12-23T16:31:10", "event": "train_log", "step": 464, "epoch": 0.20021574973031284, "progress_pct": 3.34, "epoch_pct": 3.34, "eta": "61:42:46", "max_grad_norm": 0.8, "loss": 0.837934136390686, "grad_norm": 1.9228297472000122, "learning_rate": 4.159928122192273e-05}
+{"ts": "2025-12-23T16:31:30", "event": "train_log", "step": 466, "epoch": 0.2010787486515642, "progress_pct": 3.35, "epoch_pct": 3.35, "eta": "61:35:42", "max_grad_norm": 0.8, "loss": 0.831945538520813, "grad_norm": 2.3240439891815186, "learning_rate": 4.1778975741239893e-05}
+{"ts": "2025-12-23T16:31:48", "event": "train_log", "step": 468, "epoch": 0.20194174757281552, "progress_pct": 3.36, "epoch_pct": 3.37, "eta": "61:28:26", "max_grad_norm": 0.8, "loss": 0.8012097477912903, "grad_norm": 1.879684329032898, "learning_rate": 4.195867026055705e-05}
+{"ts": "2025-12-23T16:32:07", "event": "train_log", "step": 470, "epoch": 0.20280474649406688, "progress_pct": 3.38, "epoch_pct": 3.38, "eta": "61:20:51", "max_grad_norm": 0.8, "loss": 0.8670937418937683, "grad_norm": 2.2995200157165527, "learning_rate": 4.213836477987422e-05}
+{"ts": "2025-12-23T16:32:27", "event": "train_log", "step": 472, "epoch": 0.20366774541531824, "progress_pct": 3.39, "epoch_pct": 3.39, "eta": "61:14:29", "max_grad_norm": 0.8, "loss": 0.88299161195755, "grad_norm": 2.098020553588867, "learning_rate": 4.2318059299191375e-05}
+{"ts": "2025-12-23T16:32:47", "event": "train_log", "step": 474, "epoch": 0.2045307443365696, "progress_pct": 3.41, "epoch_pct": 3.41, "eta": "61:07:44", "max_grad_norm": 0.8, "loss": 0.9045401811599731, "grad_norm": 2.0588274002075195, "learning_rate": 4.249775381850854e-05}
+{"ts": "2025-12-23T16:33:06", "event": "train_log", "step": 476, "epoch": 0.20539374325782092, "progress_pct": 3.42, "epoch_pct": 3.42, "eta": "61:00:42", "max_grad_norm": 0.8, "loss": 0.9787635803222656, "grad_norm": 1.8832436800003052, "learning_rate": 4.26774483378257e-05}
+{"ts": "2025-12-23T16:33:24", "event": "train_log", "step": 478, "epoch": 0.20625674217907228, "progress_pct": 3.44, "epoch_pct": 3.44, "eta": "60:53:28", "max_grad_norm": 0.8, "loss": 0.8874871730804443, "grad_norm": 2.3451809883117676, "learning_rate": 4.2857142857142856e-05}
+{"ts": "2025-12-23T16:33:46", "event": "train_log", "step": 480, "epoch": 0.20711974110032363, "progress_pct": 3.45, "epoch_pct": 3.45, "eta": "60:47:38", "max_grad_norm": 0.8, "loss": 0.794157087802887, "grad_norm": 1.8093878030776978, "learning_rate": 4.303683737646002e-05}
+{"ts": "2025-12-23T16:34:06", "event": "train_log", "step": 482, "epoch": 0.20798274002157496, "progress_pct": 3.47, "epoch_pct": 3.47, "eta": "60:41:26", "max_grad_norm": 0.8, "loss": 0.792723536491394, "grad_norm": 1.8396174907684326, "learning_rate": 4.321653189577718e-05}
+{"ts": "2025-12-23T16:34:26", "event": "train_log", "step": 484, "epoch": 0.20884573894282632, "progress_pct": 3.48, "epoch_pct": 3.48, "eta": "60:35:07", "max_grad_norm": 0.8, "loss": 0.9221814274787903, "grad_norm": 2.3939783573150635, "learning_rate": 4.3396226415094345e-05}
+{"ts": "2025-12-23T16:34:46", "event": "train_log", "step": 486, "epoch": 0.20970873786407768, "progress_pct": 3.49, "epoch_pct": 3.5, "eta": "60:28:46", "max_grad_norm": 0.8, "loss": 0.8716039061546326, "grad_norm": 2.2652194499969482, "learning_rate": 4.35759209344115e-05}
+{"ts": "2025-12-23T16:35:04", "event": "train_log", "step": 488, "epoch": 0.210571736785329, "progress_pct": 3.51, "epoch_pct": 3.51, "eta": "60:21:45", "max_grad_norm": 0.8, "loss": 0.8517807126045227, "grad_norm": 2.068258762359619, "learning_rate": 4.375561545372867e-05}
+{"ts": "2025-12-23T16:35:24", "event": "train_log", "step": 490, "epoch": 0.21143473570658036, "progress_pct": 3.52, "epoch_pct": 3.52, "eta": "60:15:37", "max_grad_norm": 0.8, "loss": 0.8699264526367188, "grad_norm": 2.264620542526245, "learning_rate": 4.3935309973045826e-05}
+{"ts": "2025-12-23T16:35:44", "event": "train_log", "step": 492, "epoch": 0.21229773462783172, "progress_pct": 3.54, "epoch_pct": 3.54, "eta": "60:09:10", "max_grad_norm": 0.8, "loss": 0.8760254383087158, "grad_norm": 2.0821611881256104, "learning_rate": 4.4115004492362984e-05}
+{"ts": "2025-12-23T16:36:03", "event": "train_log", "step": 494, "epoch": 0.21316073354908308, "progress_pct": 3.55, "epoch_pct": 3.55, "eta": "60:02:34", "max_grad_norm": 0.8, "loss": 0.8957411050796509, "grad_norm": 1.9486546516418457, "learning_rate": 4.429469901168015e-05}
+{"ts": "2025-12-23T16:36:21", "event": "train_log", "step": 496, "epoch": 0.2140237324703344, "progress_pct": 3.57, "epoch_pct": 3.57, "eta": "59:55:38", "max_grad_norm": 0.8, "loss": 0.981200635433197, "grad_norm": 2.092957019805908, "learning_rate": 4.447439353099731e-05}
+{"ts": "2025-12-23T16:36:40", "event": "train_log", "step": 498, "epoch": 0.21488673139158576, "progress_pct": 3.58, "epoch_pct": 3.58, "eta": "59:49:31", "max_grad_norm": 0.8, "loss": 0.8295405507087708, "grad_norm": 2.325758218765259, "learning_rate": 4.4654088050314466e-05}
+{"ts": "2025-12-23T16:36:58", "event": "train_log", "step": 500, "epoch": 0.21574973031283712, "progress_pct": 3.6, "epoch_pct": 3.6, "eta": "59:42:29", "max_grad_norm": 0.8, "loss": 0.8463165163993835, "grad_norm": 2.251295566558838, "learning_rate": 4.4833782569631624e-05}
+{"ts": "2025-12-23T16:50:41", "event": "train_log", "step": 500, "epoch": 0.21574973031283712, "progress_pct": 3.6, "epoch_pct": 3.6, "eta": "65:50:19", "max_grad_norm": 0.8, "eval_loss": 0.8945327997207642, "eval_runtime": 822.9776, "eval_samples_per_second": 2.503, "eval_steps_per_second": 2.503}
+{"ts": "2025-12-23T16:51:00", "event": "train_log", "step": 502, "epoch": 0.21661272923408845, "progress_pct": 3.61, "epoch_pct": 3.61, "eta": "65:42:36", "max_grad_norm": 0.8, "loss": 0.9099015593528748, "grad_norm": 1.8110703229904175, "learning_rate": 4.501347708894879e-05}
+{"ts": "2025-12-23T16:51:19", "event": "train_log", "step": 504, "epoch": 0.2174757281553398, "progress_pct": 3.62, "epoch_pct": 3.62, "eta": "65:34:45", "max_grad_norm": 0.8, "loss": 0.875080943107605, "grad_norm": 2.0977623462677, "learning_rate": 4.519317160826595e-05}
+{"ts": "2025-12-23T16:51:38", "event": "train_log", "step": 506, "epoch": 0.21833872707659116, "progress_pct": 3.64, "epoch_pct": 3.64, "eta": "65:26:49", "max_grad_norm": 0.8, "loss": 0.8511829376220703, "grad_norm": 1.8648077249526978, "learning_rate": 4.537286612758311e-05}
+{"ts": "2025-12-23T16:51:58", "event": "train_log", "step": 508, "epoch": 0.2192017259978425, "progress_pct": 3.65, "epoch_pct": 3.65, "eta": "65:19:32", "max_grad_norm": 0.8, "loss": 0.8408182859420776, "grad_norm": 1.754744291305542, "learning_rate": 4.555256064690027e-05}
+{"ts": "2025-12-23T16:52:17", "event": "train_log", "step": 510, "epoch": 0.22006472491909385, "progress_pct": 3.67, "epoch_pct": 3.67, "eta": "65:12:11", "max_grad_norm": 0.8, "loss": 0.8824991583824158, "grad_norm": 1.9869136810302734, "learning_rate": 4.573225516621743e-05}
+{"ts": "2025-12-23T16:52:35", "event": "train_log", "step": 512, "epoch": 0.2209277238403452, "progress_pct": 3.68, "epoch_pct": 3.68, "eta": "65:04:00", "max_grad_norm": 0.8, "loss": 0.9949447512626648, "grad_norm": 2.4601082801818848, "learning_rate": 4.5911949685534594e-05}
+{"ts": "2025-12-23T16:52:54", "event": "train_log", "step": 514, "epoch": 0.22179072276159656, "progress_pct": 3.7, "epoch_pct": 3.7, "eta": "64:56:37", "max_grad_norm": 0.8, "loss": 0.8979368805885315, "grad_norm": 2.1224238872528076, "learning_rate": 4.609164420485175e-05}
+{"ts": "2025-12-23T16:53:14", "event": "train_log", "step": 516, "epoch": 0.2226537216828479, "progress_pct": 3.71, "epoch_pct": 3.71, "eta": "64:49:39", "max_grad_norm": 0.8, "loss": 0.8728249669075012, "grad_norm": 1.6812392473220825, "learning_rate": 4.627133872416892e-05}
+{"ts": "2025-12-23T16:53:34", "event": "train_log", "step": 518, "epoch": 0.22351672060409924, "progress_pct": 3.72, "epoch_pct": 3.73, "eta": "64:42:33", "max_grad_norm": 0.8, "loss": 0.8083430528640747, "grad_norm": 1.6505497694015503, "learning_rate": 4.6451033243486075e-05}
+{"ts": "2025-12-23T16:53:55", "event": "train_log", "step": 520, "epoch": 0.2243797195253506, "progress_pct": 3.74, "epoch_pct": 3.74, "eta": "64:35:53", "max_grad_norm": 0.8, "loss": 0.8042294383049011, "grad_norm": 1.6454377174377441, "learning_rate": 4.663072776280324e-05}
+{"ts": "2025-12-23T16:54:14", "event": "train_log", "step": 522, "epoch": 0.22524271844660193, "progress_pct": 3.75, "epoch_pct": 3.75, "eta": "64:28:44", "max_grad_norm": 0.8, "loss": 0.9263730049133301, "grad_norm": 2.0974535942077637, "learning_rate": 4.68104222821204e-05}
+{"ts": "2025-12-23T16:54:34", "event": "train_log", "step": 524, "epoch": 0.2261057173678533, "progress_pct": 3.77, "epoch_pct": 3.77, "eta": "64:21:51", "max_grad_norm": 0.8, "loss": 0.9090206623077393, "grad_norm": 1.90152108669281, "learning_rate": 4.6990116801437556e-05}
+{"ts": "2025-12-23T16:54:53", "event": "train_log", "step": 526, "epoch": 0.22696871628910464, "progress_pct": 3.78, "epoch_pct": 3.78, "eta": "64:14:26", "max_grad_norm": 0.8, "loss": 0.8310918807983398, "grad_norm": 1.9843928813934326, "learning_rate": 4.716981132075472e-05}
+{"ts": "2025-12-23T16:55:12", "event": "train_log", "step": 528, "epoch": 0.227831715210356, "progress_pct": 3.8, "epoch_pct": 3.8, "eta": "64:07:37", "max_grad_norm": 0.8, "loss": 0.8838925957679749, "grad_norm": 1.7880226373672485, "learning_rate": 4.734950584007188e-05}
+{"ts": "2025-12-23T16:55:33", "event": "train_log", "step": 530, "epoch": 0.22869471413160733, "progress_pct": 3.81, "epoch_pct": 3.81, "eta": "64:01:04", "max_grad_norm": 0.8, "loss": 0.9431065917015076, "grad_norm": 1.8534979820251465, "learning_rate": 4.7529200359389045e-05}
+{"ts": "2025-12-23T16:55:52", "event": "train_log", "step": 532, "epoch": 0.2295577130528587, "progress_pct": 3.83, "epoch_pct": 3.83, "eta": "63:54:20", "max_grad_norm": 0.8, "loss": 0.7988291382789612, "grad_norm": 1.8936293125152588, "learning_rate": 4.77088948787062e-05}
+{"ts": "2025-12-23T16:56:11", "event": "train_log", "step": 534, "epoch": 0.23042071197411004, "progress_pct": 3.84, "epoch_pct": 3.84, "eta": "63:47:20", "max_grad_norm": 0.8, "loss": 0.8477044105529785, "grad_norm": 1.919683814048767, "learning_rate": 4.788858939802337e-05}
+{"ts": "2025-12-23T16:56:31", "event": "train_log", "step": 536, "epoch": 0.23128371089536137, "progress_pct": 3.85, "epoch_pct": 3.85, "eta": "63:40:29", "max_grad_norm": 0.8, "loss": 0.8482418060302734, "grad_norm": 1.8022494316101074, "learning_rate": 4.8068283917340526e-05}
+{"ts": "2025-12-23T16:56:51", "event": "train_log", "step": 538, "epoch": 0.23214670981661273, "progress_pct": 3.87, "epoch_pct": 3.87, "eta": "63:34:15", "max_grad_norm": 0.8, "loss": 0.8553584814071655, "grad_norm": 1.8628984689712524, "learning_rate": 4.8247978436657684e-05}
+{"ts": "2025-12-23T16:57:13", "event": "train_log", "step": 540, "epoch": 0.23300970873786409, "progress_pct": 3.88, "epoch_pct": 3.88, "eta": "63:28:31", "max_grad_norm": 0.8, "loss": 0.8287713527679443, "grad_norm": 1.8644732236862183, "learning_rate": 4.842767295597484e-05}
+{"ts": "2025-12-23T16:57:31", "event": "train_log", "step": 542, "epoch": 0.23387270765911541, "progress_pct": 3.9, "epoch_pct": 3.9, "eta": "63:21:28", "max_grad_norm": 0.8, "loss": 0.8984123468399048, "grad_norm": 2.162926435470581, "learning_rate": 4.8607367475292e-05}
+{"ts": "2025-12-23T16:57:53", "event": "train_log", "step": 544, "epoch": 0.23473570658036677, "progress_pct": 3.91, "epoch_pct": 3.91, "eta": "63:15:56", "max_grad_norm": 0.8, "loss": 0.725272536277771, "grad_norm": 1.453223466873169, "learning_rate": 4.8787061994609166e-05}
+{"ts": "2025-12-23T16:58:12", "event": "train_log", "step": 546, "epoch": 0.23559870550161813, "progress_pct": 3.93, "epoch_pct": 3.93, "eta": "63:09:03", "max_grad_norm": 0.8, "loss": 0.9128344655036926, "grad_norm": 2.2145543098449707, "learning_rate": 4.8966756513926324e-05}
+{"ts": "2025-12-23T16:58:31", "event": "train_log", "step": 548, "epoch": 0.23646170442286948, "progress_pct": 3.94, "epoch_pct": 3.94, "eta": "63:02:20", "max_grad_norm": 0.8, "loss": 0.782098114490509, "grad_norm": 1.7639187574386597, "learning_rate": 4.914645103324349e-05}
+{"ts": "2025-12-23T16:58:50", "event": "train_log", "step": 550, "epoch": 0.2373247033441208, "progress_pct": 3.95, "epoch_pct": 3.96, "eta": "62:55:36", "max_grad_norm": 0.8, "loss": 0.8212350010871887, "grad_norm": 1.6597305536270142, "learning_rate": 4.932614555256065e-05}
+{"ts": "2025-12-23T16:59:07", "event": "train_log", "step": 552, "epoch": 0.23818770226537217, "progress_pct": 3.97, "epoch_pct": 3.97, "eta": "62:48:31", "max_grad_norm": 0.8, "loss": 0.9033122062683105, "grad_norm": 1.9441677331924438, "learning_rate": 4.950584007187781e-05}
+{"ts": "2025-12-23T16:59:25", "event": "train_log", "step": 554, "epoch": 0.23905070118662353, "progress_pct": 3.98, "epoch_pct": 3.98, "eta": "62:41:30", "max_grad_norm": 0.8, "loss": 0.9195884466171265, "grad_norm": 2.0332281589508057, "learning_rate": 4.968553459119497e-05}
+{"ts": "2025-12-23T16:59:45", "event": "train_log", "step": 556, "epoch": 0.23991370010787486, "progress_pct": 4.0, "epoch_pct": 4.0, "eta": "62:35:20", "max_grad_norm": 0.8, "loss": 0.8828456401824951, "grad_norm": 1.9222822189331055, "learning_rate": 4.986522911051213e-05}
+{"ts": "2025-12-23T17:00:05", "event": "train_log", "step": 558, "epoch": 0.2407766990291262, "progress_pct": 4.01, "epoch_pct": 4.01, "eta": "62:29:12", "max_grad_norm": 0.8, "loss": 0.9089829921722412, "grad_norm": 1.9830708503723145, "learning_rate": 5.004492362982929e-05}
+{"ts": "2025-12-23T17:00:24", "event": "train_log", "step": 560, "epoch": 0.24163969795037757, "progress_pct": 4.03, "epoch_pct": 4.03, "eta": "62:22:45", "max_grad_norm": 0.8, "loss": 0.8747812509536743, "grad_norm": 1.99044930934906, "learning_rate": 5.022461814914645e-05}
+{"ts": "2025-12-23T17:00:43", "event": "train_log", "step": 562, "epoch": 0.2425026968716289, "progress_pct": 4.04, "epoch_pct": 4.04, "eta": "62:16:37", "max_grad_norm": 0.8, "loss": 0.8421230912208557, "grad_norm": 1.7961562871932983, "learning_rate": 5.040431266846362e-05}
+{"ts": "2025-12-23T17:01:03", "event": "train_log", "step": 564, "epoch": 0.24336569579288025, "progress_pct": 4.06, "epoch_pct": 4.06, "eta": "62:10:26", "max_grad_norm": 0.8, "loss": 0.8669922351837158, "grad_norm": 1.9798882007598877, "learning_rate": 5.058400718778078e-05}
+{"ts": "2025-12-23T17:01:23", "event": "train_log", "step": 566, "epoch": 0.2442286947141316, "progress_pct": 4.07, "epoch_pct": 4.07, "eta": "62:04:33", "max_grad_norm": 0.8, "loss": 0.8596820831298828, "grad_norm": 2.0138750076293945, "learning_rate": 5.076370170709793e-05}
+{"ts": "2025-12-23T17:01:43", "event": "train_log", "step": 568, "epoch": 0.24509169363538297, "progress_pct": 4.08, "epoch_pct": 4.08, "eta": "61:58:46", "max_grad_norm": 0.8, "loss": 0.8288601636886597, "grad_norm": 2.069021701812744, "learning_rate": 5.09433962264151e-05}
+{"ts": "2025-12-23T17:02:01", "event": "train_log", "step": 570, "epoch": 0.2459546925566343, "progress_pct": 4.1, "epoch_pct": 4.1, "eta": "61:52:26", "max_grad_norm": 0.8, "loss": 0.8321394324302673, "grad_norm": 2.6913585662841797, "learning_rate": 5.112309074573226e-05}
+{"ts": "2025-12-23T17:02:22", "event": "train_log", "step": 572, "epoch": 0.24681769147788565, "progress_pct": 4.11, "epoch_pct": 4.11, "eta": "61:47:00", "max_grad_norm": 0.8, "loss": 0.842660129070282, "grad_norm": 1.8057955503463745, "learning_rate": 5.1302785265049415e-05}
+{"ts": "2025-12-23T17:02:44", "event": "train_log", "step": 574, "epoch": 0.247680690399137, "progress_pct": 4.13, "epoch_pct": 4.13, "eta": "61:41:49", "max_grad_norm": 0.8, "loss": 0.8431437015533447, "grad_norm": 1.5741548538208008, "learning_rate": 5.148247978436658e-05}
+{"ts": "2025-12-23T17:03:02", "event": "train_log", "step": 576, "epoch": 0.24854368932038834, "progress_pct": 4.14, "epoch_pct": 4.14, "eta": "61:35:38", "max_grad_norm": 0.8, "loss": 0.9046191573143005, "grad_norm": 2.0329413414001465, "learning_rate": 5.166217430368374e-05}
+{"ts": "2025-12-23T17:03:23", "event": "train_log", "step": 578, "epoch": 0.2494066882416397, "progress_pct": 4.16, "epoch_pct": 4.16, "eta": "61:30:03", "max_grad_norm": 0.8, "loss": 0.8108726739883423, "grad_norm": 1.7638250589370728, "learning_rate": 5.18418688230009e-05}
+{"ts": "2025-12-23T17:03:41", "event": "train_log", "step": 580, "epoch": 0.25026968716289105, "progress_pct": 4.17, "epoch_pct": 4.17, "eta": "61:24:00", "max_grad_norm": 0.8, "loss": 0.8621282577514648, "grad_norm": 2.1088576316833496, "learning_rate": 5.202156334231806e-05}
+{"ts": "2025-12-23T17:04:01", "event": "train_log", "step": 582, "epoch": 0.2511326860841424, "progress_pct": 4.18, "epoch_pct": 4.19, "eta": "61:18:09", "max_grad_norm": 0.8, "loss": 0.8774182200431824, "grad_norm": 2.072145462036133, "learning_rate": 5.220125786163522e-05}
+{"ts": "2025-12-23T17:04:20", "event": "train_log", "step": 584, "epoch": 0.25199568500539377, "progress_pct": 4.2, "epoch_pct": 4.2, "eta": "61:12:14", "max_grad_norm": 0.8, "loss": 0.863322377204895, "grad_norm": 1.8093693256378174, "learning_rate": 5.2380952380952384e-05}
+{"ts": "2025-12-23T17:04:39", "event": "train_log", "step": 586, "epoch": 0.25285868392664507, "progress_pct": 4.21, "epoch_pct": 4.21, "eta": "61:06:35", "max_grad_norm": 0.8, "loss": 0.8493514657020569, "grad_norm": 1.7119505405426025, "learning_rate": 5.2560646900269536e-05}
+{"ts": "2025-12-23T17:04:58", "event": "train_log", "step": 588, "epoch": 0.2537216828478964, "progress_pct": 4.23, "epoch_pct": 4.23, "eta": "61:00:43", "max_grad_norm": 0.8, "loss": 0.8278731107711792, "grad_norm": 2.042052745819092, "learning_rate": 5.27403414195867e-05}
+{"ts": "2025-12-23T17:05:17", "event": "train_log", "step": 590, "epoch": 0.2545846817691478, "progress_pct": 4.24, "epoch_pct": 4.24, "eta": "60:54:46", "max_grad_norm": 0.8, "loss": 0.8388869166374207, "grad_norm": 1.9019906520843506, "learning_rate": 5.2920035938903866e-05}
+{"ts": "2025-12-23T17:05:36", "event": "train_log", "step": 592, "epoch": 0.25544768069039914, "progress_pct": 4.26, "epoch_pct": 4.26, "eta": "60:49:03", "max_grad_norm": 0.8, "loss": 0.8666917085647583, "grad_norm": 1.8072630167007446, "learning_rate": 5.309973045822103e-05}
+{"ts": "2025-12-23T17:05:56", "event": "train_log", "step": 594, "epoch": 0.2563106796116505, "progress_pct": 4.27, "epoch_pct": 4.27, "eta": "60:43:48", "max_grad_norm": 0.8, "loss": 0.8114299774169922, "grad_norm": 1.7099417448043823, "learning_rate": 5.327942497753818e-05}
+{"ts": "2025-12-23T17:06:16", "event": "train_log", "step": 596, "epoch": 0.25717367853290185, "progress_pct": 4.29, "epoch_pct": 4.29, "eta": "60:38:24", "max_grad_norm": 0.8, "loss": 0.8842664361000061, "grad_norm": 1.8132266998291016, "learning_rate": 5.345911949685535e-05}
+{"ts": "2025-12-23T17:06:35", "event": "train_log", "step": 598, "epoch": 0.2580366774541532, "progress_pct": 4.3, "epoch_pct": 4.3, "eta": "60:32:52", "max_grad_norm": 0.8, "loss": 0.7954385876655579, "grad_norm": 1.9578651189804077, "learning_rate": 5.363881401617251e-05}
+{"ts": "2025-12-23T17:06:54", "event": "train_log", "step": 600, "epoch": 0.2588996763754045, "progress_pct": 4.31, "epoch_pct": 4.31, "eta": "60:27:07", "max_grad_norm": 0.8, "loss": 0.912541925907135, "grad_norm": 1.832077145576477, "learning_rate": 5.381850853548968e-05}
+{"ts": "2025-12-23T17:20:42", "event": "train_log", "step": 600, "epoch": 0.2588996763754045, "progress_pct": 4.31, "epoch_pct": 4.31, "eta": "65:33:09", "max_grad_norm": 0.8, "eval_loss": 0.8770867586135864, "eval_runtime": 827.8893, "eval_samples_per_second": 2.488, "eval_steps_per_second": 2.488}
+{"ts": "2025-12-23T17:21:02", "event": "train_log", "step": 602, "epoch": 0.25976267529665586, "progress_pct": 4.33, "epoch_pct": 4.33, "eta": "65:26:52", "max_grad_norm": 0.8, "loss": 0.8436558246612549, "grad_norm": 1.7565301656723022, "learning_rate": 5.399820305480683e-05}
+{"ts": "2025-12-23T17:21:20", "event": "train_log", "step": 604, "epoch": 0.2606256742179072, "progress_pct": 4.34, "epoch_pct": 4.34, "eta": "65:20:02", "max_grad_norm": 0.8, "loss": 0.9384433031082153, "grad_norm": 1.9199552536010742, "learning_rate": 5.4177897574123994e-05}
+{"ts": "2025-12-23T17:21:40", "event": "train_log", "step": 606, "epoch": 0.2614886731391586, "progress_pct": 4.36, "epoch_pct": 4.36, "eta": "65:13:43", "max_grad_norm": 0.8, "loss": 0.8695753812789917, "grad_norm": 1.7609341144561768, "learning_rate": 5.435759209344116e-05}
+{"ts": "2025-12-23T17:22:00", "event": "train_log", "step": 608, "epoch": 0.26235167206040994, "progress_pct": 4.37, "epoch_pct": 4.37, "eta": "65:07:24", "max_grad_norm": 0.8, "loss": 0.8462895154953003, "grad_norm": 1.771958589553833, "learning_rate": 5.453728661275831e-05}
+{"ts": "2025-12-23T17:22:19", "event": "train_log", "step": 610, "epoch": 0.2632146709816613, "progress_pct": 4.39, "epoch_pct": 4.39, "eta": "65:01:05", "max_grad_norm": 0.8, "loss": 0.9074611663818359, "grad_norm": 1.8958079814910889, "learning_rate": 5.4716981132075475e-05}
+{"ts": "2025-12-23T17:22:39", "event": "train_log", "step": 612, "epoch": 0.26407766990291265, "progress_pct": 4.4, "epoch_pct": 4.4, "eta": "64:54:49", "max_grad_norm": 0.8, "loss": 0.8603307008743286, "grad_norm": 1.8668403625488281, "learning_rate": 5.489667565139264e-05}
+{"ts": "2025-12-23T17:22:58", "event": "train_log", "step": 614, "epoch": 0.26494066882416395, "progress_pct": 4.41, "epoch_pct": 4.42, "eta": "64:48:28", "max_grad_norm": 0.8, "loss": 0.8925216794013977, "grad_norm": 1.8902733325958252, "learning_rate": 5.50763701707098e-05}
+{"ts": "2025-12-23T17:23:18", "event": "train_log", "step": 616, "epoch": 0.2658036677454153, "progress_pct": 4.43, "epoch_pct": 4.43, "eta": "64:42:37", "max_grad_norm": 0.8, "loss": 0.7872517704963684, "grad_norm": 1.7172386646270752, "learning_rate": 5.525606469002696e-05}
+{"ts": "2025-12-23T17:23:38", "event": "train_log", "step": 618, "epoch": 0.26666666666666666, "progress_pct": 4.44, "epoch_pct": 4.44, "eta": "64:36:29", "max_grad_norm": 0.8, "loss": 0.7951212525367737, "grad_norm": 1.7814518213272095, "learning_rate": 5.5435759209344115e-05}
+{"ts": "2025-12-23T17:23:58", "event": "train_log", "step": 620, "epoch": 0.267529665587918, "progress_pct": 4.46, "epoch_pct": 4.46, "eta": "64:30:37", "max_grad_norm": 0.8, "loss": 0.859940230846405, "grad_norm": 1.8714463710784912, "learning_rate": 5.561545372866128e-05}
+{"ts": "2025-12-23T17:24:19", "event": "train_log", "step": 622, "epoch": 0.2683926645091694, "progress_pct": 4.47, "epoch_pct": 4.47, "eta": "64:24:52", "max_grad_norm": 0.8, "loss": 0.8953307867050171, "grad_norm": 1.9614174365997314, "learning_rate": 5.579514824797844e-05}
+{"ts": "2025-12-23T17:24:39", "event": "train_log", "step": 624, "epoch": 0.26925566343042073, "progress_pct": 4.49, "epoch_pct": 4.49, "eta": "64:19:01", "max_grad_norm": 0.8, "loss": 0.8346402645111084, "grad_norm": 1.7190991640090942, "learning_rate": 5.5974842767295596e-05}
+{"ts": "2025-12-23T17:24:57", "event": "train_log", "step": 626, "epoch": 0.27011866235167203, "progress_pct": 4.5, "epoch_pct": 4.5, "eta": "64:12:43", "max_grad_norm": 0.8, "loss": 0.8545536398887634, "grad_norm": 1.981070876121521, "learning_rate": 5.615453728661276e-05}
+{"ts": "2025-12-23T17:25:17", "event": "train_log", "step": 628, "epoch": 0.2709816612729234, "progress_pct": 4.52, "epoch_pct": 4.52, "eta": "64:06:42", "max_grad_norm": 0.8, "loss": 0.9064128398895264, "grad_norm": 1.9186289310455322, "learning_rate": 5.6334231805929926e-05}
+{"ts": "2025-12-23T17:25:37", "event": "train_log", "step": 630, "epoch": 0.27184466019417475, "progress_pct": 4.53, "epoch_pct": 4.53, "eta": "64:01:04", "max_grad_norm": 0.8, "loss": 0.883073627948761, "grad_norm": 1.9937374591827393, "learning_rate": 5.651392632524708e-05}
+{"ts": "2025-12-23T17:25:57", "event": "train_log", "step": 632, "epoch": 0.2727076591154261, "progress_pct": 4.54, "epoch_pct": 4.55, "eta": "63:55:15", "max_grad_norm": 0.8, "loss": 0.8777942657470703, "grad_norm": 1.800236701965332, "learning_rate": 5.669362084456424e-05}
+{"ts": "2025-12-23T17:26:15", "event": "train_log", "step": 634, "epoch": 0.27357065803667746, "progress_pct": 4.56, "epoch_pct": 4.56, "eta": "63:48:55", "max_grad_norm": 0.8, "loss": 0.8527862429618835, "grad_norm": 1.9164050817489624, "learning_rate": 5.687331536388141e-05}
+{"ts": "2025-12-23T17:26:34", "event": "train_log", "step": 636, "epoch": 0.2744336569579288, "progress_pct": 4.57, "epoch_pct": 4.57, "eta": "63:43:01", "max_grad_norm": 0.8, "loss": 0.881090521812439, "grad_norm": 1.7494043111801147, "learning_rate": 5.705300988319856e-05}
+{"ts": "2025-12-23T17:26:54", "event": "train_log", "step": 638, "epoch": 0.2752966558791802, "progress_pct": 4.59, "epoch_pct": 4.59, "eta": "63:37:07", "max_grad_norm": 0.8, "loss": 0.8547250628471375, "grad_norm": 1.9433623552322388, "learning_rate": 5.7232704402515724e-05}
+{"ts": "2025-12-23T17:27:12", "event": "train_log", "step": 640, "epoch": 0.2761596548004315, "progress_pct": 4.6, "epoch_pct": 4.6, "eta": "63:30:50", "max_grad_norm": 0.8, "loss": 0.8429408073425293, "grad_norm": 1.7539145946502686, "learning_rate": 5.741239892183289e-05}
+{"ts": "2025-12-23T17:27:30", "event": "train_log", "step": 642, "epoch": 0.27702265372168283, "progress_pct": 4.62, "epoch_pct": 4.62, "eta": "63:24:53", "max_grad_norm": 0.8, "loss": 0.8587079644203186, "grad_norm": 1.904911756515503, "learning_rate": 5.7592093441150054e-05}
+{"ts": "2025-12-23T17:27:49", "event": "train_log", "step": 644, "epoch": 0.2778856526429342, "progress_pct": 4.63, "epoch_pct": 4.63, "eta": "63:18:56", "max_grad_norm": 0.8, "loss": 0.7795872688293457, "grad_norm": 1.5534838438034058, "learning_rate": 5.7771787960467206e-05}
+{"ts": "2025-12-23T17:28:08", "event": "train_log", "step": 646, "epoch": 0.27874865156418555, "progress_pct": 4.64, "epoch_pct": 4.65, "eta": "63:13:09", "max_grad_norm": 0.8, "loss": 0.7657744884490967, "grad_norm": 1.704798698425293, "learning_rate": 5.795148247978437e-05}
+{"ts": "2025-12-23T17:28:27", "event": "train_log", "step": 648, "epoch": 0.2796116504854369, "progress_pct": 4.66, "epoch_pct": 4.66, "eta": "63:07:13", "max_grad_norm": 0.8, "loss": 0.8119702935218811, "grad_norm": 1.8432048559188843, "learning_rate": 5.8131176999101536e-05}
+{"ts": "2025-12-23T17:28:46", "event": "train_log", "step": 650, "epoch": 0.28047464940668826, "progress_pct": 4.67, "epoch_pct": 4.67, "eta": "63:01:23", "max_grad_norm": 0.8, "loss": 0.8846080899238586, "grad_norm": 1.9495919942855835, "learning_rate": 5.831087151841869e-05}
+{"ts": "2025-12-23T17:29:05", "event": "train_log", "step": 652, "epoch": 0.2813376483279396, "progress_pct": 4.69, "epoch_pct": 4.69, "eta": "62:55:36", "max_grad_norm": 0.8, "loss": 0.854324460029602, "grad_norm": 1.9626739025115967, "learning_rate": 5.849056603773585e-05}
+{"ts": "2025-12-23T17:29:23", "event": "train_log", "step": 654, "epoch": 0.2822006472491909, "progress_pct": 4.7, "epoch_pct": 4.7, "eta": "62:49:51", "max_grad_norm": 0.8, "loss": 0.7908785939216614, "grad_norm": 1.767221212387085, "learning_rate": 5.867026055705302e-05}
+{"ts": "2025-12-23T17:29:43", "event": "train_log", "step": 656, "epoch": 0.2830636461704423, "progress_pct": 4.72, "epoch_pct": 4.72, "eta": "62:44:17", "max_grad_norm": 0.8, "loss": 0.8155530691146851, "grad_norm": 1.8866491317749023, "learning_rate": 5.8849955076370175e-05}
+{"ts": "2025-12-23T17:30:01", "event": "train_log", "step": 658, "epoch": 0.28392664509169363, "progress_pct": 4.73, "epoch_pct": 4.73, "eta": "62:38:20", "max_grad_norm": 0.8, "loss": 0.9833307266235352, "grad_norm": 1.9543417692184448, "learning_rate": 5.9029649595687334e-05}
+{"ts": "2025-12-23T17:30:19", "event": "train_log", "step": 660, "epoch": 0.284789644012945, "progress_pct": 4.75, "epoch_pct": 4.75, "eta": "62:32:35", "max_grad_norm": 0.8, "loss": 0.7991418242454529, "grad_norm": 1.867039680480957, "learning_rate": 5.920934411500449e-05}
+{"ts": "2025-12-23T17:30:38", "event": "train_log", "step": 662, "epoch": 0.28565264293419634, "progress_pct": 4.76, "epoch_pct": 4.76, "eta": "62:26:47", "max_grad_norm": 0.8, "loss": 0.8232532143592834, "grad_norm": 1.8487379550933838, "learning_rate": 5.938903863432166e-05}
+{"ts": "2025-12-23T17:30:58", "event": "train_log", "step": 664, "epoch": 0.2865156418554477, "progress_pct": 4.77, "epoch_pct": 4.78, "eta": "62:21:45", "max_grad_norm": 0.8, "loss": 0.76528000831604, "grad_norm": 1.5015742778778076, "learning_rate": 5.956873315363881e-05}
+{"ts": "2025-12-23T17:31:17", "event": "train_log", "step": 666, "epoch": 0.287378640776699, "progress_pct": 4.79, "epoch_pct": 4.79, "eta": "62:16:14", "max_grad_norm": 0.8, "loss": 0.8220192790031433, "grad_norm": 2.726778268814087, "learning_rate": 5.974842767295597e-05}
+{"ts": "2025-12-23T17:31:36", "event": "train_log", "step": 668, "epoch": 0.28824163969795036, "progress_pct": 4.8, "epoch_pct": 4.8, "eta": "62:10:54", "max_grad_norm": 0.8, "loss": 0.7534968256950378, "grad_norm": 1.6754003763198853, "learning_rate": 5.992812219227314e-05}
+{"ts": "2025-12-23T17:31:56", "event": "train_log", "step": 670, "epoch": 0.2891046386192017, "progress_pct": 4.82, "epoch_pct": 4.82, "eta": "62:05:44", "max_grad_norm": 0.8, "loss": 0.8405768871307373, "grad_norm": 1.9948434829711914, "learning_rate": 6.01078167115903e-05}
+{"ts": "2025-12-23T17:32:16", "event": "train_log", "step": 672, "epoch": 0.28996763754045307, "progress_pct": 4.83, "epoch_pct": 4.83, "eta": "62:00:25", "max_grad_norm": 0.8, "loss": 0.8143276572227478, "grad_norm": 1.7982401847839355, "learning_rate": 6.0287511230907455e-05}
+{"ts": "2025-12-23T17:32:34", "event": "train_log", "step": 674, "epoch": 0.29083063646170443, "progress_pct": 4.85, "epoch_pct": 4.85, "eta": "61:54:51", "max_grad_norm": 0.8, "loss": 0.9602267742156982, "grad_norm": 1.7311779260635376, "learning_rate": 6.046720575022462e-05}
+{"ts": "2025-12-23T17:32:53", "event": "train_log", "step": 676, "epoch": 0.2916936353829558, "progress_pct": 4.86, "epoch_pct": 4.86, "eta": "61:49:22", "max_grad_norm": 0.8, "loss": 0.8628544807434082, "grad_norm": 1.9906021356582642, "learning_rate": 6.0646900269541785e-05}
+{"ts": "2025-12-23T17:33:12", "event": "train_log", "step": 678, "epoch": 0.29255663430420714, "progress_pct": 4.87, "epoch_pct": 4.88, "eta": "61:44:08", "max_grad_norm": 0.8, "loss": 0.7800343632698059, "grad_norm": 1.7477494478225708, "learning_rate": 6.082659478885895e-05}
+{"ts": "2025-12-23T17:33:31", "event": "train_log", "step": 680, "epoch": 0.29341963322545844, "progress_pct": 4.89, "epoch_pct": 4.89, "eta": "61:38:47", "max_grad_norm": 0.8, "loss": 0.8722298741340637, "grad_norm": 1.7978737354278564, "learning_rate": 6.10062893081761e-05}
+{"ts": "2025-12-23T17:33:50", "event": "train_log", "step": 682, "epoch": 0.2942826321467098, "progress_pct": 4.9, "epoch_pct": 4.9, "eta": "61:33:27", "max_grad_norm": 0.8, "loss": 0.8895675539970398, "grad_norm": 2.1463348865509033, "learning_rate": 6.118598382749326e-05}
+{"ts": "2025-12-23T17:34:08", "event": "train_log", "step": 684, "epoch": 0.29514563106796116, "progress_pct": 4.92, "epoch_pct": 4.92, "eta": "61:27:57", "max_grad_norm": 0.8, "loss": 0.8579491972923279, "grad_norm": 1.9256783723831177, "learning_rate": 6.136567834681042e-05}
+{"ts": "2025-12-23T17:34:27", "event": "train_log", "step": 686, "epoch": 0.2960086299892125, "progress_pct": 4.93, "epoch_pct": 4.93, "eta": "61:22:58", "max_grad_norm": 0.8, "loss": 0.7811232805252075, "grad_norm": 1.4459543228149414, "learning_rate": 6.154537286612758e-05}
+{"ts": "2025-12-23T17:34:46", "event": "train_log", "step": 688, "epoch": 0.29687162891046387, "progress_pct": 4.95, "epoch_pct": 4.95, "eta": "61:17:49", "max_grad_norm": 0.8, "loss": 0.7827678322792053, "grad_norm": 1.6202397346496582, "learning_rate": 6.172506738544474e-05}
+{"ts": "2025-12-23T17:35:05", "event": "train_log", "step": 690, "epoch": 0.2977346278317152, "progress_pct": 4.96, "epoch_pct": 4.96, "eta": "61:12:32", "max_grad_norm": 0.8, "loss": 0.8186459541320801, "grad_norm": 2.0524775981903076, "learning_rate": 6.19047619047619e-05}
+{"ts": "2025-12-23T17:35:23", "event": "train_log", "step": 692, "epoch": 0.2985976267529666, "progress_pct": 4.98, "epoch_pct": 4.98, "eta": "61:07:12", "max_grad_norm": 0.8, "loss": 0.8691027760505676, "grad_norm": 1.7867428064346313, "learning_rate": 6.208445642407907e-05}
+{"ts": "2025-12-23T17:35:43", "event": "train_log", "step": 694, "epoch": 0.2994606256742179, "progress_pct": 4.99, "epoch_pct": 4.99, "eta": "61:02:11", "max_grad_norm": 0.8, "loss": 0.8233416080474854, "grad_norm": 1.9605737924575806, "learning_rate": 6.226415094339622e-05}
+{"ts": "2025-12-23T17:36:01", "event": "train_log", "step": 696, "epoch": 0.30032362459546924, "progress_pct": 5.0, "epoch_pct": 5.01, "eta": "60:57:03", "max_grad_norm": 0.8, "loss": 0.8493421077728271, "grad_norm": 1.7035984992980957, "learning_rate": 6.244384546271339e-05}
+{"ts": "2025-12-23T17:36:19", "event": "train_log", "step": 698, "epoch": 0.3011866235167206, "progress_pct": 5.02, "epoch_pct": 5.02, "eta": "60:51:37", "max_grad_norm": 0.8, "loss": 0.8114250302314758, "grad_norm": 1.9402128458023071, "learning_rate": 6.262353998203055e-05}
+{"ts": "2025-12-23T17:36:37", "event": "train_log", "step": 700, "epoch": 0.30204962243797195, "progress_pct": 5.03, "epoch_pct": 5.03, "eta": "60:46:21", "max_grad_norm": 0.8, "loss": 0.8996106386184692, "grad_norm": 1.8533520698547363, "learning_rate": 6.28032345013477e-05}
+{"ts": "2025-12-23T17:50:37", "event": "train_log", "step": 700, "epoch": 0.30204962243797195, "progress_pct": 5.03, "epoch_pct": 5.03, "eta": "65:10:24", "max_grad_norm": 0.8, "eval_loss": 0.8631294369697571, "eval_runtime": 839.6269, "eval_samples_per_second": 2.453, "eval_steps_per_second": 2.453}
+{"ts": "2025-12-23T17:50:55", "event": "train_log", "step": 702, "epoch": 0.3029126213592233, "progress_pct": 5.05, "epoch_pct": 5.05, "eta": "65:04:24", "max_grad_norm": 0.8, "loss": 0.7978370189666748, "grad_norm": 1.727793574333191, "learning_rate": 6.298292902066487e-05}
+{"ts": "2025-12-23T17:51:14", "event": "train_log", "step": 704, "epoch": 0.30377562028047467, "progress_pct": 5.06, "epoch_pct": 5.06, "eta": "64:58:27", "max_grad_norm": 0.8, "loss": 0.8120381236076355, "grad_norm": 1.9601492881774902, "learning_rate": 6.316262353998203e-05}
+{"ts": "2025-12-23T17:51:32", "event": "train_log", "step": 706, "epoch": 0.304638619201726, "progress_pct": 5.08, "epoch_pct": 5.08, "eta": "64:52:41", "max_grad_norm": 0.8, "loss": 0.8005605340003967, "grad_norm": 1.8380948305130005, "learning_rate": 6.33423180592992e-05}
+{"ts": "2025-12-23T17:51:51", "event": "train_log", "step": 708, "epoch": 0.3055016181229773, "progress_pct": 5.09, "epoch_pct": 5.09, "eta": "64:46:57", "max_grad_norm": 0.8, "loss": 0.8967245817184448, "grad_norm": 1.8210986852645874, "learning_rate": 6.352201257861635e-05}
+{"ts": "2025-12-23T17:52:10", "event": "train_log", "step": 710, "epoch": 0.3063646170442287, "progress_pct": 5.1, "epoch_pct": 5.11, "eta": "64:41:09", "max_grad_norm": 0.8, "loss": 0.8271524906158447, "grad_norm": 3.5359878540039062, "learning_rate": 6.370170709793352e-05}
+{"ts": "2025-12-23T17:52:29", "event": "train_log", "step": 712, "epoch": 0.30722761596548004, "progress_pct": 5.12, "epoch_pct": 5.12, "eta": "64:35:28", "max_grad_norm": 0.8, "loss": 0.8759040236473083, "grad_norm": 1.7943733930587769, "learning_rate": 6.388140161725068e-05}
+{"ts": "2025-12-23T17:52:47", "event": "train_log", "step": 714, "epoch": 0.3080906148867314, "progress_pct": 5.13, "epoch_pct": 5.13, "eta": "64:29:52", "max_grad_norm": 0.8, "loss": 0.7872476577758789, "grad_norm": 1.5483570098876953, "learning_rate": 6.406109613656783e-05}
+{"ts": "2025-12-23T17:53:07", "event": "train_log", "step": 716, "epoch": 0.30895361380798275, "progress_pct": 5.15, "epoch_pct": 5.15, "eta": "64:24:25", "max_grad_norm": 0.8, "loss": 0.849986732006073, "grad_norm": 1.8044816255569458, "learning_rate": 6.4240790655885e-05}
+{"ts": "2025-12-23T17:53:26", "event": "train_log", "step": 718, "epoch": 0.3098166127292341, "progress_pct": 5.16, "epoch_pct": 5.16, "eta": "64:18:56", "max_grad_norm": 0.8, "loss": 0.8292878866195679, "grad_norm": 1.548281192779541, "learning_rate": 6.442048517520216e-05}
+{"ts": "2025-12-23T17:53:46", "event": "train_log", "step": 720, "epoch": 0.3106796116504854, "progress_pct": 5.18, "epoch_pct": 5.18, "eta": "64:13:40", "max_grad_norm": 0.8, "loss": 0.8796502351760864, "grad_norm": 1.643020749092102, "learning_rate": 6.460017969451933e-05}
+{"ts": "2025-12-23T17:54:06", "event": "train_log", "step": 722, "epoch": 0.31154261057173677, "progress_pct": 5.19, "epoch_pct": 5.19, "eta": "64:08:32", "max_grad_norm": 0.8, "loss": 0.806233286857605, "grad_norm": 1.633881688117981, "learning_rate": 6.477987421383648e-05}
+{"ts": "2025-12-23T17:54:25", "event": "train_log", "step": 724, "epoch": 0.3124056094929881, "progress_pct": 5.21, "epoch_pct": 5.21, "eta": "64:03:12", "max_grad_norm": 0.8, "loss": 0.802722156047821, "grad_norm": 1.7712864875793457, "learning_rate": 6.495956873315364e-05}
+{"ts": "2025-12-23T17:54:44", "event": "train_log", "step": 726, "epoch": 0.3132686084142395, "progress_pct": 5.22, "epoch_pct": 5.22, "eta": "63:57:36", "max_grad_norm": 0.8, "loss": 0.8735752105712891, "grad_norm": 1.9158337116241455, "learning_rate": 6.513926325247081e-05}
+{"ts": "2025-12-23T17:55:03", "event": "train_log", "step": 728, "epoch": 0.31413160733549084, "progress_pct": 5.23, "epoch_pct": 5.24, "eta": "63:52:20", "max_grad_norm": 0.8, "loss": 0.793403148651123, "grad_norm": 1.8648416996002197, "learning_rate": 6.531895777178796e-05}
+{"ts": "2025-12-23T17:55:22", "event": "train_log", "step": 730, "epoch": 0.3149946062567422, "progress_pct": 5.25, "epoch_pct": 5.25, "eta": "63:47:05", "max_grad_norm": 0.8, "loss": 0.7772542834281921, "grad_norm": 1.5917021036148071, "learning_rate": 6.549865229110512e-05}
+{"ts": "2025-12-23T17:55:41", "event": "train_log", "step": 732, "epoch": 0.31585760517799355, "progress_pct": 5.26, "epoch_pct": 5.26, "eta": "63:41:33", "max_grad_norm": 0.8, "loss": 0.8731698989868164, "grad_norm": 1.8560330867767334, "learning_rate": 6.567834681042229e-05}
+{"ts": "2025-12-23T17:56:00", "event": "train_log", "step": 734, "epoch": 0.31672060409924485, "progress_pct": 5.28, "epoch_pct": 5.28, "eta": "63:36:26", "max_grad_norm": 0.8, "loss": 0.8167505264282227, "grad_norm": 1.6038225889205933, "learning_rate": 6.585804132973945e-05}
+{"ts": "2025-12-23T17:56:20", "event": "train_log", "step": 736, "epoch": 0.3175836030204962, "progress_pct": 5.29, "epoch_pct": 5.29, "eta": "63:31:25", "max_grad_norm": 0.8, "loss": 0.7810112237930298, "grad_norm": 1.6221873760223389, "learning_rate": 6.60377358490566e-05}
+{"ts": "2025-12-23T17:56:38", "event": "train_log", "step": 738, "epoch": 0.31844660194174756, "progress_pct": 5.31, "epoch_pct": 5.31, "eta": "63:25:53", "max_grad_norm": 0.8, "loss": 0.8735228776931763, "grad_norm": 1.8528746366500854, "learning_rate": 6.621743036837377e-05}
+{"ts": "2025-12-23T17:56:58", "event": "train_log", "step": 740, "epoch": 0.3193096008629989, "progress_pct": 5.32, "epoch_pct": 5.32, "eta": "63:20:46", "max_grad_norm": 0.8, "loss": 0.8231552243232727, "grad_norm": 1.473233938217163, "learning_rate": 6.639712488769094e-05}
+{"ts": "2025-12-23T17:57:17", "event": "train_log", "step": 742, "epoch": 0.3201725997842503, "progress_pct": 5.34, "epoch_pct": 5.34, "eta": "63:15:34", "max_grad_norm": 0.8, "loss": 0.8400886058807373, "grad_norm": 1.7726222276687622, "learning_rate": 6.657681940700809e-05}
+{"ts": "2025-12-23T17:57:36", "event": "train_log", "step": 744, "epoch": 0.32103559870550163, "progress_pct": 5.35, "epoch_pct": 5.35, "eta": "63:10:32", "max_grad_norm": 0.8, "loss": 0.7371112108230591, "grad_norm": 1.7245025634765625, "learning_rate": 6.675651392632525e-05}
+{"ts": "2025-12-23T17:57:56", "event": "train_log", "step": 746, "epoch": 0.321898597626753, "progress_pct": 5.36, "epoch_pct": 5.36, "eta": "63:05:39", "max_grad_norm": 0.8, "loss": 0.8057155013084412, "grad_norm": 1.6225395202636719, "learning_rate": 6.693620844564242e-05}
+{"ts": "2025-12-23T17:58:15", "event": "train_log", "step": 748, "epoch": 0.3227615965480043, "progress_pct": 5.38, "epoch_pct": 5.38, "eta": "63:00:34", "max_grad_norm": 0.8, "loss": 0.8471131324768066, "grad_norm": 1.607007384300232, "learning_rate": 6.711590296495958e-05}
+{"ts": "2025-12-23T17:58:35", "event": "train_log", "step": 750, "epoch": 0.32362459546925565, "progress_pct": 5.39, "epoch_pct": 5.39, "eta": "62:55:42", "max_grad_norm": 0.8, "loss": 0.8251315951347351, "grad_norm": 1.760038137435913, "learning_rate": 6.729559748427673e-05}
+{"ts": "2025-12-23T17:58:55", "event": "train_log", "step": 752, "epoch": 0.324487594390507, "progress_pct": 5.41, "epoch_pct": 5.41, "eta": "62:51:00", "max_grad_norm": 0.8, "loss": 0.8215667605400085, "grad_norm": 1.8012818098068237, "learning_rate": 6.74752920035939e-05}
+{"ts": "2025-12-23T17:59:15", "event": "train_log", "step": 754, "epoch": 0.32535059331175836, "progress_pct": 5.42, "epoch_pct": 5.42, "eta": "62:46:15", "max_grad_norm": 0.8, "loss": 0.8113123178482056, "grad_norm": 1.769614815711975, "learning_rate": 6.765498652291105e-05}
+{"ts": "2025-12-23T17:59:34", "event": "train_log", "step": 756, "epoch": 0.3262135922330097, "progress_pct": 5.44, "epoch_pct": 5.44, "eta": "62:41:08", "max_grad_norm": 0.8, "loss": 0.8893613219261169, "grad_norm": 1.7554916143417358, "learning_rate": 6.783468104222822e-05}
+{"ts": "2025-12-23T17:59:54", "event": "train_log", "step": 758, "epoch": 0.3270765911542611, "progress_pct": 5.45, "epoch_pct": 5.45, "eta": "62:36:23", "max_grad_norm": 0.8, "loss": 0.8165010809898376, "grad_norm": 1.5720254182815552, "learning_rate": 6.801437556154538e-05}
+{"ts": "2025-12-23T18:00:12", "event": "train_log", "step": 760, "epoch": 0.32793959007551243, "progress_pct": 5.46, "epoch_pct": 5.47, "eta": "62:31:16", "max_grad_norm": 0.8, "loss": 0.8800036311149597, "grad_norm": 2.1887543201446533, "learning_rate": 6.819407008086253e-05}
+{"ts": "2025-12-23T18:00:32", "event": "train_log", "step": 762, "epoch": 0.32880258899676373, "progress_pct": 5.48, "epoch_pct": 5.48, "eta": "62:26:26", "max_grad_norm": 0.8, "loss": 0.7268908619880676, "grad_norm": 1.63350510597229, "learning_rate": 6.83737646001797e-05}
+{"ts": "2025-12-23T18:00:53", "event": "train_log", "step": 764, "epoch": 0.3296655879180151, "progress_pct": 5.49, "epoch_pct": 5.49, "eta": "62:22:00", "max_grad_norm": 0.8, "loss": 0.7695913910865784, "grad_norm": 1.5838189125061035, "learning_rate": 6.855345911949685e-05}
+{"ts": "2025-12-23T18:01:12", "event": "train_log", "step": 766, "epoch": 0.33052858683926645, "progress_pct": 5.51, "epoch_pct": 5.51, "eta": "62:17:06", "max_grad_norm": 0.8, "loss": 0.9005667567253113, "grad_norm": 1.8752859830856323, "learning_rate": 6.873315363881401e-05}
+{"ts": "2025-12-23T18:01:32", "event": "train_log", "step": 768, "epoch": 0.3313915857605178, "progress_pct": 5.52, "epoch_pct": 5.52, "eta": "62:12:30", "max_grad_norm": 0.8, "loss": 0.7960113883018494, "grad_norm": 1.4201446771621704, "learning_rate": 6.891284815813118e-05}
+{"ts": "2025-12-23T18:01:52", "event": "train_log", "step": 770, "epoch": 0.33225458468176916, "progress_pct": 5.54, "epoch_pct": 5.54, "eta": "62:07:56", "max_grad_norm": 0.8, "loss": 0.851472020149231, "grad_norm": 1.6665233373641968, "learning_rate": 6.909254267744834e-05}
+{"ts": "2025-12-23T18:02:12", "event": "train_log", "step": 772, "epoch": 0.3331175836030205, "progress_pct": 5.55, "epoch_pct": 5.55, "eta": "62:03:31", "max_grad_norm": 0.8, "loss": 0.8403398990631104, "grad_norm": 1.72721266746521, "learning_rate": 6.92722371967655e-05}
+{"ts": "2025-12-23T18:02:31", "event": "train_log", "step": 774, "epoch": 0.3339805825242718, "progress_pct": 5.57, "epoch_pct": 5.57, "eta": "61:58:45", "max_grad_norm": 0.8, "loss": 0.7582393884658813, "grad_norm": 1.648680567741394, "learning_rate": 6.945193171608266e-05}
+{"ts": "2025-12-23T18:02:50", "event": "train_log", "step": 776, "epoch": 0.3348435814455232, "progress_pct": 5.58, "epoch_pct": 5.58, "eta": "61:53:59", "max_grad_norm": 0.8, "loss": 0.8250385522842407, "grad_norm": 1.89604651927948, "learning_rate": 6.963162623539982e-05}
+{"ts": "2025-12-23T18:03:09", "event": "train_log", "step": 778, "epoch": 0.33570658036677453, "progress_pct": 5.59, "epoch_pct": 5.6, "eta": "61:49:03", "max_grad_norm": 0.8, "loss": 0.7916358709335327, "grad_norm": 1.4748259782791138, "learning_rate": 6.981132075471698e-05}
+{"ts": "2025-12-23T18:03:28", "event": "train_log", "step": 780, "epoch": 0.3365695792880259, "progress_pct": 5.61, "epoch_pct": 5.61, "eta": "61:44:27", "max_grad_norm": 0.8, "loss": 0.7432135939598083, "grad_norm": 1.4885735511779785, "learning_rate": 6.999101527403414e-05}
+{"ts": "2025-12-23T18:03:49", "event": "train_log", "step": 782, "epoch": 0.33743257820927725, "progress_pct": 5.62, "epoch_pct": 5.62, "eta": "61:40:20", "max_grad_norm": 0.8, "loss": 0.7369741201400757, "grad_norm": 1.5150227546691895, "learning_rate": 7.01707097933513e-05}
+{"ts": "2025-12-23T18:04:10", "event": "train_log", "step": 784, "epoch": 0.3382955771305286, "progress_pct": 5.64, "epoch_pct": 5.64, "eta": "61:35:58", "max_grad_norm": 0.8, "loss": 0.8347548246383667, "grad_norm": 2.0702431201934814, "learning_rate": 7.035040431266847e-05}
+{"ts": "2025-12-23T18:04:31", "event": "train_log", "step": 786, "epoch": 0.33915857605177996, "progress_pct": 5.65, "epoch_pct": 5.65, "eta": "61:31:48", "max_grad_norm": 0.8, "loss": 0.7757654190063477, "grad_norm": 1.8330671787261963, "learning_rate": 7.053009883198562e-05}
+{"ts": "2025-12-23T18:04:50", "event": "train_log", "step": 788, "epoch": 0.34002157497303126, "progress_pct": 5.67, "epoch_pct": 5.67, "eta": "61:27:16", "max_grad_norm": 0.8, "loss": 0.7966224551200867, "grad_norm": 1.6809033155441284, "learning_rate": 7.070979335130279e-05}
+{"ts": "2025-12-23T18:05:10", "event": "train_log", "step": 790, "epoch": 0.3408845738942826, "progress_pct": 5.68, "epoch_pct": 5.68, "eta": "61:22:50", "max_grad_norm": 0.8, "loss": 0.7977543473243713, "grad_norm": 1.5811399221420288, "learning_rate": 7.088948787061995e-05}
+{"ts": "2025-12-23T18:05:30", "event": "train_log", "step": 792, "epoch": 0.341747572815534, "progress_pct": 5.69, "epoch_pct": 5.7, "eta": "61:18:29", "max_grad_norm": 0.8, "loss": 0.7842527627944946, "grad_norm": 1.6572248935699463, "learning_rate": 7.10691823899371e-05}
+{"ts": "2025-12-23T18:05:50", "event": "train_log", "step": 794, "epoch": 0.34261057173678533, "progress_pct": 5.71, "epoch_pct": 5.71, "eta": "61:14:17", "max_grad_norm": 0.8, "loss": 0.7586495876312256, "grad_norm": 1.6138155460357666, "learning_rate": 7.124887690925427e-05}
+{"ts": "2025-12-23T18:06:10", "event": "train_log", "step": 796, "epoch": 0.3434735706580367, "progress_pct": 5.72, "epoch_pct": 5.72, "eta": "61:09:50", "max_grad_norm": 0.8, "loss": 0.9010837078094482, "grad_norm": 1.9198144674301147, "learning_rate": 7.142857142857143e-05}
+{"ts": "2025-12-23T18:06:29", "event": "train_log", "step": 798, "epoch": 0.34433656957928804, "progress_pct": 5.74, "epoch_pct": 5.74, "eta": "61:05:31", "max_grad_norm": 0.8, "loss": 0.7857459783554077, "grad_norm": 1.6212328672409058, "learning_rate": 7.16082659478886e-05}
+{"ts": "2025-12-23T18:06:48", "event": "train_log", "step": 800, "epoch": 0.3451995685005394, "progress_pct": 5.75, "epoch_pct": 5.75, "eta": "61:00:45", "max_grad_norm": 0.8, "loss": 0.9087778925895691, "grad_norm": 2.1088991165161133, "learning_rate": 7.178796046720575e-05}
+{"ts": "2025-12-23T18:20:39", "event": "train_log", "step": 800, "epoch": 0.3451995685005394, "progress_pct": 5.75, "epoch_pct": 5.75, "eta": "64:47:52", "max_grad_norm": 0.8, "eval_loss": 0.8476243615150452, "eval_runtime": 831.6735, "eval_samples_per_second": 2.477, "eval_steps_per_second": 2.477}
+{"ts": "2025-12-23T18:20:58", "event": "train_log", "step": 802, "epoch": 0.3460625674217907, "progress_pct": 5.77, "epoch_pct": 5.77, "eta": "64:42:43", "max_grad_norm": 0.8, "loss": 0.7561838626861572, "grad_norm": 1.6325052976608276, "learning_rate": 7.196765498652292e-05}
+{"ts": "2025-12-23T18:21:16", "event": "train_log", "step": 804, "epoch": 0.34692556634304206, "progress_pct": 5.78, "epoch_pct": 5.78, "eta": "64:37:26", "max_grad_norm": 0.8, "loss": 0.7940152883529663, "grad_norm": 1.8044698238372803, "learning_rate": 7.214734950584008e-05}
+{"ts": "2025-12-23T18:21:36", "event": "train_log", "step": 806, "epoch": 0.3477885652642934, "progress_pct": 5.8, "epoch_pct": 5.8, "eta": "64:32:30", "max_grad_norm": 0.8, "loss": 0.8002730011940002, "grad_norm": 1.4942690134048462, "learning_rate": 7.232704402515723e-05}
+{"ts": "2025-12-23T18:21:55", "event": "train_log", "step": 808, "epoch": 0.34865156418554477, "progress_pct": 5.81, "epoch_pct": 5.81, "eta": "64:27:34", "max_grad_norm": 0.8, "loss": 0.8211291432380676, "grad_norm": 1.5209156274795532, "learning_rate": 7.25067385444744e-05}
+{"ts": "2025-12-23T18:22:14", "event": "train_log", "step": 810, "epoch": 0.34951456310679613, "progress_pct": 5.82, "epoch_pct": 5.83, "eta": "64:22:37", "max_grad_norm": 0.8, "loss": 0.8690080642700195, "grad_norm": 1.662622094154358, "learning_rate": 7.268643306379156e-05}
+{"ts": "2025-12-23T18:22:33", "event": "train_log", "step": 812, "epoch": 0.3503775620280475, "progress_pct": 5.84, "epoch_pct": 5.84, "eta": "64:17:37", "max_grad_norm": 0.8, "loss": 0.8259872794151306, "grad_norm": 1.4452823400497437, "learning_rate": 7.286612758310873e-05}
+{"ts": "2025-12-23T18:22:53", "event": "train_log", "step": 814, "epoch": 0.3512405609492988, "progress_pct": 5.85, "epoch_pct": 5.85, "eta": "64:12:44", "max_grad_norm": 0.8, "loss": 0.7973082661628723, "grad_norm": 1.6030809879302979, "learning_rate": 7.304582210242588e-05}
+{"ts": "2025-12-23T18:23:12", "event": "train_log", "step": 816, "epoch": 0.35210355987055014, "progress_pct": 5.87, "epoch_pct": 5.87, "eta": "64:07:43", "max_grad_norm": 0.8, "loss": 0.8258423805236816, "grad_norm": 1.5944641828536987, "learning_rate": 7.322551662174304e-05}
+{"ts": "2025-12-23T18:23:30", "event": "train_log", "step": 818, "epoch": 0.3529665587918015, "progress_pct": 5.88, "epoch_pct": 5.88, "eta": "64:02:43", "max_grad_norm": 0.8, "loss": 0.8570499420166016, "grad_norm": 1.6284228563308716, "learning_rate": 7.340521114106021e-05}
+{"ts": "2025-12-23T18:23:50", "event": "train_log", "step": 820, "epoch": 0.35382955771305286, "progress_pct": 5.9, "epoch_pct": 5.9, "eta": "63:58:06", "max_grad_norm": 0.8, "loss": 0.7808573842048645, "grad_norm": 1.399397373199463, "learning_rate": 7.358490566037736e-05}
+{"ts": "2025-12-23T18:24:10", "event": "train_log", "step": 822, "epoch": 0.3546925566343042, "progress_pct": 5.91, "epoch_pct": 5.91, "eta": "63:53:24", "max_grad_norm": 0.8, "loss": 0.7514002323150635, "grad_norm": 1.6233313083648682, "learning_rate": 7.376460017969452e-05}
+{"ts": "2025-12-23T18:24:29", "event": "train_log", "step": 824, "epoch": 0.35555555555555557, "progress_pct": 5.92, "epoch_pct": 5.93, "eta": "63:48:29", "max_grad_norm": 0.8, "loss": 0.8392049670219421, "grad_norm": 3.1610822677612305, "learning_rate": 7.394429469901169e-05}
+{"ts": "2025-12-23T18:24:49", "event": "train_log", "step": 826, "epoch": 0.3564185544768069, "progress_pct": 5.94, "epoch_pct": 5.94, "eta": "63:43:58", "max_grad_norm": 0.8, "loss": 0.7574125528335571, "grad_norm": 1.5170103311538696, "learning_rate": 7.412398921832885e-05}
+{"ts": "2025-12-23T18:25:08", "event": "train_log", "step": 828, "epoch": 0.3572815533980582, "progress_pct": 5.95, "epoch_pct": 5.95, "eta": "63:39:07", "max_grad_norm": 0.8, "loss": 0.8394677639007568, "grad_norm": 1.687654733657837, "learning_rate": 7.4303683737646e-05}
+{"ts": "2025-12-23T18:25:29", "event": "train_log", "step": 830, "epoch": 0.3581445523193096, "progress_pct": 5.97, "epoch_pct": 5.97, "eta": "63:34:54", "max_grad_norm": 0.8, "loss": 0.7760812044143677, "grad_norm": 1.6306912899017334, "learning_rate": 7.448337825696317e-05}
+{"ts": "2025-12-23T18:25:50", "event": "train_log", "step": 832, "epoch": 0.35900755124056094, "progress_pct": 5.98, "epoch_pct": 5.98, "eta": "63:30:38", "max_grad_norm": 0.8, "loss": 0.8146010041236877, "grad_norm": 1.4824501276016235, "learning_rate": 7.466307277628034e-05}
+{"ts": "2025-12-23T18:26:10", "event": "train_log", "step": 834, "epoch": 0.3598705501618123, "progress_pct": 6.0, "epoch_pct": 6.0, "eta": "63:26:11", "max_grad_norm": 0.8, "loss": 0.7824002504348755, "grad_norm": 1.962728500366211, "learning_rate": 7.484276729559749e-05}
+{"ts": "2025-12-23T18:26:29", "event": "train_log", "step": 836, "epoch": 0.36073354908306365, "progress_pct": 6.01, "epoch_pct": 6.01, "eta": "63:21:25", "max_grad_norm": 0.8, "loss": 0.8955144882202148, "grad_norm": 1.805561900138855, "learning_rate": 7.502246181491465e-05}
+{"ts": "2025-12-23T18:26:49", "event": "train_log", "step": 838, "epoch": 0.361596548004315, "progress_pct": 6.03, "epoch_pct": 6.03, "eta": "63:16:51", "max_grad_norm": 0.8, "loss": 0.7638000249862671, "grad_norm": 1.6058090925216675, "learning_rate": 7.52021563342318e-05}
+{"ts": "2025-12-23T18:27:08", "event": "train_log", "step": 840, "epoch": 0.36245954692556637, "progress_pct": 6.04, "epoch_pct": 6.04, "eta": "63:12:11", "max_grad_norm": 0.8, "loss": 0.8465489745140076, "grad_norm": 1.5326330661773682, "learning_rate": 7.538185085354897e-05}
+{"ts": "2025-12-23T18:27:27", "event": "train_log", "step": 842, "epoch": 0.36332254584681767, "progress_pct": 6.05, "epoch_pct": 6.06, "eta": "63:07:35", "max_grad_norm": 0.8, "loss": 0.7809846997261047, "grad_norm": 1.563953161239624, "learning_rate": 7.556154537286612e-05}
+{"ts": "2025-12-23T18:27:48", "event": "train_log", "step": 844, "epoch": 0.364185544768069, "progress_pct": 6.07, "epoch_pct": 6.07, "eta": "63:03:18", "max_grad_norm": 0.8, "loss": 0.7431546449661255, "grad_norm": 1.3842798471450806, "learning_rate": 7.574123989218329e-05}
+{"ts": "2025-12-23T18:28:08", "event": "train_log", "step": 846, "epoch": 0.3650485436893204, "progress_pct": 6.08, "epoch_pct": 6.08, "eta": "62:58:56", "max_grad_norm": 0.8, "loss": 0.8984674215316772, "grad_norm": 1.843582272529602, "learning_rate": 7.592093441150045e-05}
+{"ts": "2025-12-23T18:28:26", "event": "train_log", "step": 848, "epoch": 0.36591154261057174, "progress_pct": 6.1, "epoch_pct": 6.1, "eta": "62:54:17", "max_grad_norm": 0.8, "loss": 0.7809584736824036, "grad_norm": 1.7077877521514893, "learning_rate": 7.610062893081762e-05}
+{"ts": "2025-12-23T18:28:48", "event": "train_log", "step": 850, "epoch": 0.3667745415318231, "progress_pct": 6.11, "epoch_pct": 6.11, "eta": "62:50:21", "max_grad_norm": 0.8, "loss": 0.7576664090156555, "grad_norm": 1.3528809547424316, "learning_rate": 7.628032345013477e-05}
+{"ts": "2025-12-23T18:29:08", "event": "train_log", "step": 852, "epoch": 0.36763754045307445, "progress_pct": 6.13, "epoch_pct": 6.13, "eta": "62:46:01", "max_grad_norm": 0.8, "loss": 0.7666721940040588, "grad_norm": 1.4391332864761353, "learning_rate": 7.646001796945193e-05}
+{"ts": "2025-12-23T18:29:28", "event": "train_log", "step": 854, "epoch": 0.3685005393743258, "progress_pct": 6.14, "epoch_pct": 6.14, "eta": "62:41:38", "max_grad_norm": 0.8, "loss": 0.8356342911720276, "grad_norm": 1.626839518547058, "learning_rate": 7.66397124887691e-05}
+{"ts": "2025-12-23T18:29:47", "event": "train_log", "step": 856, "epoch": 0.3693635382955771, "progress_pct": 6.15, "epoch_pct": 6.16, "eta": "62:37:05", "max_grad_norm": 0.8, "loss": 0.7766274809837341, "grad_norm": 1.5550813674926758, "learning_rate": 7.681940700808625e-05}
+{"ts": "2025-12-23T18:30:05", "event": "train_log", "step": 858, "epoch": 0.37022653721682847, "progress_pct": 6.17, "epoch_pct": 6.17, "eta": "62:32:33", "max_grad_norm": 0.8, "loss": 0.7646660804748535, "grad_norm": 1.4684302806854248, "learning_rate": 7.699910152740341e-05}
+{"ts": "2025-12-23T18:30:23", "event": "train_log", "step": 860, "epoch": 0.3710895361380798, "progress_pct": 6.18, "epoch_pct": 6.18, "eta": "62:27:34", "max_grad_norm": 0.8, "loss": 0.8716844916343689, "grad_norm": 1.6446272134780884, "learning_rate": 7.717879604672058e-05}
+{"ts": "2025-12-23T18:30:43", "event": "train_log", "step": 862, "epoch": 0.3719525350593312, "progress_pct": 6.2, "epoch_pct": 6.2, "eta": "62:23:22", "max_grad_norm": 0.8, "loss": 0.8561879992485046, "grad_norm": 1.5995690822601318, "learning_rate": 7.735849056603774e-05}
+{"ts": "2025-12-23T18:31:02", "event": "train_log", "step": 864, "epoch": 0.37281553398058254, "progress_pct": 6.21, "epoch_pct": 6.21, "eta": "62:19:01", "max_grad_norm": 0.8, "loss": 0.8795325756072998, "grad_norm": 1.632079839706421, "learning_rate": 7.75381850853549e-05}
+{"ts": "2025-12-23T18:31:21", "event": "train_log", "step": 866, "epoch": 0.3736785329018339, "progress_pct": 6.23, "epoch_pct": 6.23, "eta": "62:14:27", "max_grad_norm": 0.8, "loss": 0.8779821991920471, "grad_norm": 1.7169907093048096, "learning_rate": 7.771787960467206e-05}
+{"ts": "2025-12-23T18:31:38", "event": "train_log", "step": 868, "epoch": 0.3745415318230852, "progress_pct": 6.24, "epoch_pct": 6.24, "eta": "62:09:42", "max_grad_norm": 0.8, "loss": 0.8811644911766052, "grad_norm": 1.8590933084487915, "learning_rate": 7.789757412398922e-05}
+{"ts": "2025-12-23T18:31:56", "event": "train_log", "step": 870, "epoch": 0.37540453074433655, "progress_pct": 6.26, "epoch_pct": 6.26, "eta": "62:04:58", "max_grad_norm": 0.8, "loss": 0.8391726016998291, "grad_norm": 1.804978370666504, "learning_rate": 7.807726864330638e-05}
+{"ts": "2025-12-23T18:32:15", "event": "train_log", "step": 872, "epoch": 0.3762675296655879, "progress_pct": 6.27, "epoch_pct": 6.27, "eta": "62:00:42", "max_grad_norm": 0.8, "loss": 0.8425854444503784, "grad_norm": 2.2929575443267822, "learning_rate": 7.825696316262354e-05}
+{"ts": "2025-12-23T18:32:36", "event": "train_log", "step": 874, "epoch": 0.37713052858683926, "progress_pct": 6.28, "epoch_pct": 6.29, "eta": "61:56:48", "max_grad_norm": 0.8, "loss": 0.8409838080406189, "grad_norm": 1.551804780960083, "learning_rate": 7.84366576819407e-05}
+{"ts": "2025-12-23T18:32:57", "event": "train_log", "step": 876, "epoch": 0.3779935275080906, "progress_pct": 6.3, "epoch_pct": 6.3, "eta": "61:52:57", "max_grad_norm": 0.8, "loss": 0.7441553473472595, "grad_norm": 1.4655314683914185, "learning_rate": 7.861635220125787e-05}
+{"ts": "2025-12-23T18:33:17", "event": "train_log", "step": 878, "epoch": 0.378856526429342, "progress_pct": 6.31, "epoch_pct": 6.31, "eta": "61:48:44", "max_grad_norm": 0.8, "loss": 0.8850297331809998, "grad_norm": 1.7886772155761719, "learning_rate": 7.879604672057502e-05}
+{"ts": "2025-12-23T18:33:35", "event": "train_log", "step": 880, "epoch": 0.37971952535059333, "progress_pct": 6.33, "epoch_pct": 6.33, "eta": "61:44:18", "max_grad_norm": 0.8, "loss": 0.7991185188293457, "grad_norm": 1.4744676351547241, "learning_rate": 7.897574123989219e-05}
+{"ts": "2025-12-23T18:33:52", "event": "train_log", "step": 882, "epoch": 0.38058252427184464, "progress_pct": 6.34, "epoch_pct": 6.34, "eta": "61:39:31", "max_grad_norm": 0.8, "loss": 0.7903867363929749, "grad_norm": 1.6844931840896606, "learning_rate": 7.915543575920935e-05}
+{"ts": "2025-12-23T18:34:11", "event": "train_log", "step": 884, "epoch": 0.381445523193096, "progress_pct": 6.36, "epoch_pct": 6.36, "eta": "61:35:15", "max_grad_norm": 0.8, "loss": 0.812190592288971, "grad_norm": 1.674704909324646, "learning_rate": 7.93351302785265e-05}
+{"ts": "2025-12-23T18:34:31", "event": "train_log", "step": 886, "epoch": 0.38230852211434735, "progress_pct": 6.37, "epoch_pct": 6.37, "eta": "61:31:10", "max_grad_norm": 0.8, "loss": 0.781878650188446, "grad_norm": 1.6548097133636475, "learning_rate": 7.951482479784367e-05}
+{"ts": "2025-12-23T18:34:50", "event": "train_log", "step": 888, "epoch": 0.3831715210355987, "progress_pct": 6.38, "epoch_pct": 6.39, "eta": "61:26:57", "max_grad_norm": 0.8, "loss": 0.8223106861114502, "grad_norm": 1.3750375509262085, "learning_rate": 7.969451931716083e-05}
+{"ts": "2025-12-23T18:35:10", "event": "train_log", "step": 890, "epoch": 0.38403451995685006, "progress_pct": 6.4, "epoch_pct": 6.4, "eta": "61:22:54", "max_grad_norm": 0.8, "loss": 0.7170665860176086, "grad_norm": 1.42178475856781, "learning_rate": 7.9874213836478e-05}
+{"ts": "2025-12-23T18:35:29", "event": "train_log", "step": 892, "epoch": 0.3848975188781014, "progress_pct": 6.41, "epoch_pct": 6.41, "eta": "61:18:54", "max_grad_norm": 0.8, "loss": 0.8533666133880615, "grad_norm": 1.7126129865646362, "learning_rate": 8.005390835579515e-05}
+{"ts": "2025-12-23T18:35:49", "event": "train_log", "step": 894, "epoch": 0.3857605177993528, "progress_pct": 6.43, "epoch_pct": 6.43, "eta": "61:14:47", "max_grad_norm": 0.8, "loss": 0.8799889087677002, "grad_norm": 1.552474856376648, "learning_rate": 8.023360287511232e-05}
+{"ts": "2025-12-23T18:36:10", "event": "train_log", "step": 896, "epoch": 0.3866235167206041, "progress_pct": 6.44, "epoch_pct": 6.44, "eta": "61:11:09", "max_grad_norm": 0.8, "loss": 0.7968940734863281, "grad_norm": 1.2612507343292236, "learning_rate": 8.041329739442948e-05}
+{"ts": "2025-12-23T18:36:30", "event": "train_log", "step": 898, "epoch": 0.38748651564185543, "progress_pct": 6.46, "epoch_pct": 6.46, "eta": "61:07:11", "max_grad_norm": 0.8, "loss": 0.7910935282707214, "grad_norm": 1.4770987033843994, "learning_rate": 8.059299191374663e-05}
+{"ts": "2025-12-23T18:36:50", "event": "train_log", "step": 900, "epoch": 0.3883495145631068, "progress_pct": 6.47, "epoch_pct": 6.47, "eta": "61:03:16", "max_grad_norm": 0.8, "loss": 0.7873474955558777, "grad_norm": 1.4640419483184814, "learning_rate": 8.07726864330638e-05}
+{"ts": "2025-12-23T18:50:33", "event": "train_log", "step": 900, "epoch": 0.3883495145631068, "progress_pct": 6.47, "epoch_pct": 6.47, "eta": "64:21:39", "max_grad_norm": 0.8, "eval_loss": 0.8351938128471375, "eval_runtime": 823.5649, "eval_samples_per_second": 2.501, "eval_steps_per_second": 2.501}
+{"ts": "2025-12-23T18:50:53", "event": "train_log", "step": 902, "epoch": 0.38921251348435815, "progress_pct": 6.49, "epoch_pct": 6.49, "eta": "64:17:14", "max_grad_norm": 0.8, "loss": 0.8183950185775757, "grad_norm": 1.3780814409255981, "learning_rate": 8.095238095238096e-05}
+{"ts": "2025-12-23T18:51:13", "event": "train_log", "step": 904, "epoch": 0.3900755124056095, "progress_pct": 6.5, "epoch_pct": 6.5, "eta": "64:13:00", "max_grad_norm": 0.8, "loss": 0.786055326461792, "grad_norm": 1.5895066261291504, "learning_rate": 8.113207547169813e-05}
+{"ts": "2025-12-23T18:51:33", "event": "train_log", "step": 906, "epoch": 0.39093851132686086, "progress_pct": 6.51, "epoch_pct": 6.52, "eta": "64:08:40", "max_grad_norm": 0.8, "loss": 0.8638336658477783, "grad_norm": 1.6340118646621704, "learning_rate": 8.131176999101528e-05}
+{"ts": "2025-12-23T18:51:52", "event": "train_log", "step": 908, "epoch": 0.3918015102481122, "progress_pct": 6.53, "epoch_pct": 6.53, "eta": "64:04:12", "max_grad_norm": 0.8, "loss": 0.702790379524231, "grad_norm": 1.3990486860275269, "learning_rate": 8.149146451033244e-05}
+{"ts": "2025-12-23T18:52:13", "event": "train_log", "step": 910, "epoch": 0.3926645091693635, "progress_pct": 6.54, "epoch_pct": 6.54, "eta": "63:59:56", "max_grad_norm": 0.8, "loss": 0.7845349907875061, "grad_norm": 1.3703010082244873, "learning_rate": 8.167115902964961e-05}
+{"ts": "2025-12-23T18:52:35", "event": "train_log", "step": 912, "epoch": 0.3935275080906149, "progress_pct": 6.56, "epoch_pct": 6.56, "eta": "63:56:13", "max_grad_norm": 0.8, "loss": 0.8117155432701111, "grad_norm": 1.4194090366363525, "learning_rate": 8.185085354896676e-05}
+{"ts": "2025-12-23T18:52:55", "event": "train_log", "step": 914, "epoch": 0.39439050701186623, "progress_pct": 6.57, "epoch_pct": 6.57, "eta": "63:52:00", "max_grad_norm": 0.8, "loss": 0.8229454755783081, "grad_norm": 1.4716848134994507, "learning_rate": 8.203054806828392e-05}
+{"ts": "2025-12-23T18:53:16", "event": "train_log", "step": 916, "epoch": 0.3952535059331176, "progress_pct": 6.59, "epoch_pct": 6.59, "eta": "63:48:00", "max_grad_norm": 0.8, "loss": 0.698765218257904, "grad_norm": 1.51328706741333, "learning_rate": 8.221024258760108e-05}
+{"ts": "2025-12-23T18:53:36", "event": "train_log", "step": 918, "epoch": 0.39611650485436894, "progress_pct": 6.6, "epoch_pct": 6.6, "eta": "63:43:43", "max_grad_norm": 0.8, "loss": 0.7694190144538879, "grad_norm": 1.4755173921585083, "learning_rate": 8.238993710691824e-05}
+{"ts": "2025-12-23T18:53:55", "event": "train_log", "step": 920, "epoch": 0.3969795037756203, "progress_pct": 6.61, "epoch_pct": 6.62, "eta": "63:39:22", "max_grad_norm": 0.8, "loss": 0.779274046421051, "grad_norm": 1.5302451848983765, "learning_rate": 8.25696316262354e-05}
+{"ts": "2025-12-23T18:54:15", "event": "train_log", "step": 922, "epoch": 0.3978425026968716, "progress_pct": 6.63, "epoch_pct": 6.63, "eta": "63:35:15", "max_grad_norm": 0.8, "loss": 0.713699221611023, "grad_norm": 1.474968433380127, "learning_rate": 8.274932614555256e-05}
+{"ts": "2025-12-23T18:54:35", "event": "train_log", "step": 924, "epoch": 0.39870550161812296, "progress_pct": 6.64, "epoch_pct": 6.65, "eta": "63:31:07", "max_grad_norm": 0.8, "loss": 0.7869517803192139, "grad_norm": 1.4024375677108765, "learning_rate": 8.292902066486972e-05}
+{"ts": "2025-12-23T18:54:54", "event": "train_log", "step": 926, "epoch": 0.3995685005393743, "progress_pct": 6.66, "epoch_pct": 6.66, "eta": "63:26:43", "max_grad_norm": 0.8, "loss": 0.9040259718894958, "grad_norm": 1.5729585886001587, "learning_rate": 8.310871518418689e-05}
+{"ts": "2025-12-23T18:55:14", "event": "train_log", "step": 928, "epoch": 0.4004314994606257, "progress_pct": 6.67, "epoch_pct": 6.67, "eta": "63:22:28", "max_grad_norm": 0.8, "loss": 0.7589567303657532, "grad_norm": 1.609850525856018, "learning_rate": 8.328840970350404e-05}
+{"ts": "2025-12-23T18:55:33", "event": "train_log", "step": 930, "epoch": 0.40129449838187703, "progress_pct": 6.69, "epoch_pct": 6.69, "eta": "63:18:10", "max_grad_norm": 0.8, "loss": 0.8725809454917908, "grad_norm": 1.778781533241272, "learning_rate": 8.34681042228212e-05}
+{"ts": "2025-12-23T18:55:52", "event": "train_log", "step": 932, "epoch": 0.4021574973031284, "progress_pct": 6.7, "epoch_pct": 6.7, "eta": "63:13:47", "max_grad_norm": 0.8, "loss": 0.8280068039894104, "grad_norm": 1.6333881616592407, "learning_rate": 8.364779874213837e-05}
+{"ts": "2025-12-23T18:56:09", "event": "train_log", "step": 934, "epoch": 0.40302049622437974, "progress_pct": 6.72, "epoch_pct": 6.72, "eta": "63:09:04", "max_grad_norm": 0.8, "loss": 0.8293300867080688, "grad_norm": 1.6444122791290283, "learning_rate": 8.382749326145552e-05}
+{"ts": "2025-12-23T18:56:30", "event": "train_log", "step": 936, "epoch": 0.40388349514563104, "progress_pct": 6.73, "epoch_pct": 6.73, "eta": "63:05:20", "max_grad_norm": 0.8, "loss": 0.7628244161605835, "grad_norm": 1.2815860509872437, "learning_rate": 8.400718778077269e-05}
+{"ts": "2025-12-23T18:56:49", "event": "train_log", "step": 938, "epoch": 0.4047464940668824, "progress_pct": 6.74, "epoch_pct": 6.75, "eta": "63:00:55", "max_grad_norm": 0.8, "loss": 0.8951035141944885, "grad_norm": 1.7849020957946777, "learning_rate": 8.418688230008985e-05}
+{"ts": "2025-12-23T18:57:08", "event": "train_log", "step": 940, "epoch": 0.40560949298813376, "progress_pct": 6.76, "epoch_pct": 6.76, "eta": "62:56:45", "max_grad_norm": 0.8, "loss": 0.7677619457244873, "grad_norm": 1.800937294960022, "learning_rate": 8.436657681940702e-05}
+{"ts": "2025-12-23T18:57:26", "event": "train_log", "step": 942, "epoch": 0.4064724919093851, "progress_pct": 6.77, "epoch_pct": 6.77, "eta": "62:52:20", "max_grad_norm": 0.8, "loss": 0.8553397059440613, "grad_norm": 1.5004104375839233, "learning_rate": 8.454627133872417e-05}
+{"ts": "2025-12-23T18:57:47", "event": "train_log", "step": 944, "epoch": 0.40733549083063647, "progress_pct": 6.79, "epoch_pct": 6.79, "eta": "62:48:34", "max_grad_norm": 0.8, "loss": 0.7110850811004639, "grad_norm": 1.3399486541748047, "learning_rate": 8.472596585804133e-05}
+{"ts": "2025-12-23T18:58:06", "event": "train_log", "step": 946, "epoch": 0.4081984897518878, "progress_pct": 6.8, "epoch_pct": 6.8, "eta": "62:44:15", "max_grad_norm": 0.8, "loss": 0.9379971623420715, "grad_norm": 1.8936012983322144, "learning_rate": 8.49056603773585e-05}
+{"ts": "2025-12-23T18:58:27", "event": "train_log", "step": 948, "epoch": 0.4090614886731392, "progress_pct": 6.82, "epoch_pct": 6.82, "eta": "62:40:38", "max_grad_norm": 0.8, "loss": 0.7492738366127014, "grad_norm": 1.2648893594741821, "learning_rate": 8.508535489667565e-05}
+{"ts": "2025-12-23T18:58:48", "event": "train_log", "step": 950, "epoch": 0.4099244875943905, "progress_pct": 6.83, "epoch_pct": 6.83, "eta": "62:36:46", "max_grad_norm": 0.8, "loss": 0.7922301292419434, "grad_norm": 1.5587494373321533, "learning_rate": 8.526504941599281e-05}
+{"ts": "2025-12-23T18:59:08", "event": "train_log", "step": 952, "epoch": 0.41078748651564184, "progress_pct": 6.84, "epoch_pct": 6.85, "eta": "62:32:55", "max_grad_norm": 0.8, "loss": 0.7494275569915771, "grad_norm": 1.5042990446090698, "learning_rate": 8.544474393530998e-05}
+{"ts": "2025-12-23T18:59:28", "event": "train_log", "step": 954, "epoch": 0.4116504854368932, "progress_pct": 6.86, "epoch_pct": 6.86, "eta": "62:29:04", "max_grad_norm": 0.8, "loss": 0.7983633279800415, "grad_norm": 1.4338537454605103, "learning_rate": 8.562443845462714e-05}
+{"ts": "2025-12-23T18:59:48", "event": "train_log", "step": 956, "epoch": 0.41251348435814456, "progress_pct": 6.87, "epoch_pct": 6.88, "eta": "62:25:06", "max_grad_norm": 0.8, "loss": 0.813288152217865, "grad_norm": 1.6135673522949219, "learning_rate": 8.58041329739443e-05}
+{"ts": "2025-12-23T19:00:08", "event": "train_log", "step": 958, "epoch": 0.4133764832793959, "progress_pct": 6.89, "epoch_pct": 6.89, "eta": "62:21:13", "max_grad_norm": 0.8, "loss": 0.7794904708862305, "grad_norm": 1.3761460781097412, "learning_rate": 8.598382749326146e-05}
+{"ts": "2025-12-23T19:00:28", "event": "train_log", "step": 960, "epoch": 0.41423948220064727, "progress_pct": 6.9, "epoch_pct": 6.9, "eta": "62:17:16", "max_grad_norm": 0.8, "loss": 0.7544562816619873, "grad_norm": 1.51504385471344, "learning_rate": 8.616352201257863e-05}
+{"ts": "2025-12-23T19:00:49", "event": "train_log", "step": 962, "epoch": 0.41510248112189857, "progress_pct": 6.92, "epoch_pct": 6.92, "eta": "62:13:43", "max_grad_norm": 0.8, "loss": 0.8151395916938782, "grad_norm": 1.3698909282684326, "learning_rate": 8.634321653189578e-05}
+{"ts": "2025-12-23T19:01:07", "event": "train_log", "step": 964, "epoch": 0.4159654800431499, "progress_pct": 6.93, "epoch_pct": 6.93, "eta": "62:09:26", "max_grad_norm": 0.8, "loss": 0.7997464537620544, "grad_norm": 1.5509650707244873, "learning_rate": 8.652291105121294e-05}
+{"ts": "2025-12-23T19:01:27", "event": "train_log", "step": 966, "epoch": 0.4168284789644013, "progress_pct": 6.95, "epoch_pct": 6.95, "eta": "62:05:35", "max_grad_norm": 0.8, "loss": 0.767808735370636, "grad_norm": 1.2967383861541748, "learning_rate": 8.67026055705301e-05}
+{"ts": "2025-12-23T19:01:47", "event": "train_log", "step": 968, "epoch": 0.41769147788565264, "progress_pct": 6.96, "epoch_pct": 6.96, "eta": "62:01:40", "max_grad_norm": 0.8, "loss": 0.7537208795547485, "grad_norm": 1.2444552183151245, "learning_rate": 8.688230008984727e-05}
+{"ts": "2025-12-23T19:02:06", "event": "train_log", "step": 970, "epoch": 0.418554476806904, "progress_pct": 6.97, "epoch_pct": 6.98, "eta": "61:57:38", "max_grad_norm": 0.8, "loss": 0.7796021103858948, "grad_norm": 1.4390850067138672, "learning_rate": 8.706199460916442e-05}
+{"ts": "2025-12-23T19:02:24", "event": "train_log", "step": 972, "epoch": 0.41941747572815535, "progress_pct": 6.99, "epoch_pct": 6.99, "eta": "61:53:29", "max_grad_norm": 0.8, "loss": 0.7812987565994263, "grad_norm": 1.4822322130203247, "learning_rate": 8.724168912848159e-05}
+{"ts": "2025-12-23T19:02:44", "event": "train_log", "step": 974, "epoch": 0.4202804746494067, "progress_pct": 7.0, "epoch_pct": 7.0, "eta": "61:49:44", "max_grad_norm": 0.8, "loss": 0.7396363019943237, "grad_norm": 1.3770837783813477, "learning_rate": 8.742138364779875e-05}
+{"ts": "2025-12-23T19:03:03", "event": "train_log", "step": 976, "epoch": 0.421143473570658, "progress_pct": 7.02, "epoch_pct": 7.02, "eta": "61:45:38", "max_grad_norm": 0.8, "loss": 0.8390982151031494, "grad_norm": 1.5534236431121826, "learning_rate": 8.76010781671159e-05}
+{"ts": "2025-12-23T19:03:21", "event": "train_log", "step": 978, "epoch": 0.42200647249190937, "progress_pct": 7.03, "epoch_pct": 7.03, "eta": "61:41:38", "max_grad_norm": 0.8, "loss": 0.8368987441062927, "grad_norm": 1.5428354740142822, "learning_rate": 8.778077268643307e-05}
+{"ts": "2025-12-23T19:03:40", "event": "train_log", "step": 980, "epoch": 0.4228694714131607, "progress_pct": 7.05, "epoch_pct": 7.05, "eta": "61:37:36", "max_grad_norm": 0.8, "loss": 0.8203011155128479, "grad_norm": 1.3579497337341309, "learning_rate": 8.796046720575023e-05}
+{"ts": "2025-12-23T19:03:59", "event": "train_log", "step": 982, "epoch": 0.4237324703344121, "progress_pct": 7.06, "epoch_pct": 7.06, "eta": "61:33:36", "max_grad_norm": 0.8, "loss": 0.7622520923614502, "grad_norm": 1.5543184280395508, "learning_rate": 8.81401617250674e-05}
+{"ts": "2025-12-23T19:04:18", "event": "train_log", "step": 984, "epoch": 0.42459546925566344, "progress_pct": 7.08, "epoch_pct": 7.08, "eta": "61:29:47", "max_grad_norm": 0.8, "loss": 0.7764845490455627, "grad_norm": 1.5728850364685059, "learning_rate": 8.831985624438455e-05}
+{"ts": "2025-12-23T19:04:37", "event": "train_log", "step": 986, "epoch": 0.4254584681769148, "progress_pct": 7.09, "epoch_pct": 7.09, "eta": "61:25:55", "max_grad_norm": 0.8, "loss": 0.8136081099510193, "grad_norm": 1.5842933654785156, "learning_rate": 8.849955076370172e-05}
+{"ts": "2025-12-23T19:04:56", "event": "train_log", "step": 988, "epoch": 0.42632146709816615, "progress_pct": 7.1, "epoch_pct": 7.11, "eta": "61:21:54", "max_grad_norm": 0.8, "loss": 0.8388259410858154, "grad_norm": 1.5721858739852905, "learning_rate": 8.867924528301888e-05}
+{"ts": "2025-12-23T19:05:13", "event": "train_log", "step": 990, "epoch": 0.42718446601941745, "progress_pct": 7.12, "epoch_pct": 7.12, "eta": "61:17:38", "max_grad_norm": 0.8, "loss": 0.8829372525215149, "grad_norm": 1.6682584285736084, "learning_rate": 8.885893980233603e-05}
+{"ts": "2025-12-23T19:05:31", "event": "train_log", "step": 992, "epoch": 0.4280474649406688, "progress_pct": 7.13, "epoch_pct": 7.13, "eta": "61:13:31", "max_grad_norm": 0.8, "loss": 0.8545100688934326, "grad_norm": 1.7098870277404785, "learning_rate": 8.90386343216532e-05}
+{"ts": "2025-12-23T19:05:50", "event": "train_log", "step": 994, "epoch": 0.42891046386192017, "progress_pct": 7.15, "epoch_pct": 7.15, "eta": "61:09:47", "max_grad_norm": 0.8, "loss": 0.8233623504638672, "grad_norm": 1.5161751508712769, "learning_rate": 8.921832884097035e-05}
+{"ts": "2025-12-23T19:06:10", "event": "train_log", "step": 996, "epoch": 0.4297734627831715, "progress_pct": 7.16, "epoch_pct": 7.16, "eta": "61:06:07", "max_grad_norm": 0.8, "loss": 0.7576998472213745, "grad_norm": 1.457794189453125, "learning_rate": 8.939802336028751e-05}
+{"ts": "2025-12-23T19:06:29", "event": "train_log", "step": 998, "epoch": 0.4306364617044229, "progress_pct": 7.18, "epoch_pct": 7.18, "eta": "61:02:21", "max_grad_norm": 0.8, "loss": 0.8089564442634583, "grad_norm": 1.6383825540542603, "learning_rate": 8.957771787960468e-05}
+{"ts": "2025-12-23T19:06:49", "event": "train_log", "step": 1000, "epoch": 0.43149946062567424, "progress_pct": 7.19, "epoch_pct": 7.19, "eta": "60:58:38", "max_grad_norm": 0.8, "loss": 0.7791779637336731, "grad_norm": 1.4149754047393799, "learning_rate": 8.975741239892183e-05}
+{"ts": "2025-12-23T19:20:44", "event": "train_log", "step": 1000, "epoch": 0.43149946062567424, "progress_pct": 7.19, "epoch_pct": 7.19, "eta": "63:58:16", "max_grad_norm": 0.8, "eval_loss": 0.8266019821166992, "eval_runtime": 835.0044, "eval_samples_per_second": 2.467, "eval_steps_per_second": 2.467}
+{"ts": "2025-12-23T19:21:04", "event": "train_log", "step": 1002, "epoch": 0.4323624595469256, "progress_pct": 7.2, "epoch_pct": 7.21, "eta": "63:54:23", "max_grad_norm": 0.8, "loss": 0.7443642616271973, "grad_norm": 1.5475730895996094, "learning_rate": 8.9937106918239e-05}
+{"ts": "2025-12-23T19:21:24", "event": "train_log", "step": 1004, "epoch": 0.4332254584681769, "progress_pct": 7.22, "epoch_pct": 7.22, "eta": "63:50:22", "max_grad_norm": 0.8, "loss": 0.8210510015487671, "grad_norm": 1.4621447324752808, "learning_rate": 9.011680143755616e-05}
+{"ts": "2025-12-23T19:21:43", "event": "train_log", "step": 1006, "epoch": 0.43408845738942825, "progress_pct": 7.23, "epoch_pct": 7.23, "eta": "63:46:12", "max_grad_norm": 0.8, "loss": 0.8130766153335571, "grad_norm": 1.3651922941207886, "learning_rate": 9.029649595687331e-05}
+{"ts": "2025-12-23T19:22:04", "event": "train_log", "step": 1008, "epoch": 0.4349514563106796, "progress_pct": 7.25, "epoch_pct": 7.25, "eta": "63:42:31", "max_grad_norm": 0.8, "loss": 0.79145747423172, "grad_norm": 1.3587357997894287, "learning_rate": 9.047619047619048e-05}
+{"ts": "2025-12-23T19:22:22", "event": "train_log", "step": 1010, "epoch": 0.43581445523193096, "progress_pct": 7.26, "epoch_pct": 7.26, "eta": "63:38:21", "max_grad_norm": 0.8, "loss": 0.8286946415901184, "grad_norm": 1.464652180671692, "learning_rate": 9.065588499550764e-05}
+{"ts": "2025-12-23T19:22:42", "event": "train_log", "step": 1012, "epoch": 0.4366774541531823, "progress_pct": 7.28, "epoch_pct": 7.28, "eta": "63:34:19", "max_grad_norm": 0.8, "loss": 0.717319667339325, "grad_norm": 1.3912625312805176, "learning_rate": 9.083557951482479e-05}
+{"ts": "2025-12-23T19:23:04", "event": "train_log", "step": 1014, "epoch": 0.4375404530744337, "progress_pct": 7.29, "epoch_pct": 7.29, "eta": "63:30:53", "max_grad_norm": 0.8, "loss": 0.7294418215751648, "grad_norm": 1.2045491933822632, "learning_rate": 9.101527403414196e-05}
+{"ts": "2025-12-23T19:23:25", "event": "train_log", "step": 1016, "epoch": 0.438403451995685, "progress_pct": 7.31, "epoch_pct": 7.31, "eta": "63:27:18", "max_grad_norm": 0.8, "loss": 0.7366700768470764, "grad_norm": 1.3449069261550903, "learning_rate": 9.119496855345912e-05}
+{"ts": "2025-12-23T19:23:44", "event": "train_log", "step": 1018, "epoch": 0.43926645091693634, "progress_pct": 7.32, "epoch_pct": 7.32, "eta": "63:23:11", "max_grad_norm": 0.8, "loss": 0.8551293611526489, "grad_norm": 2.901384115219116, "learning_rate": 9.137466307277629e-05}
+{"ts": "2025-12-23T19:24:02", "event": "train_log", "step": 1020, "epoch": 0.4401294498381877, "progress_pct": 7.33, "epoch_pct": 7.34, "eta": "63:18:54", "max_grad_norm": 0.8, "loss": 0.8944112658500671, "grad_norm": 1.6648688316345215, "learning_rate": 9.155435759209344e-05}
+{"ts": "2025-12-23T19:24:21", "event": "train_log", "step": 1022, "epoch": 0.44099244875943905, "progress_pct": 7.35, "epoch_pct": 7.35, "eta": "63:15:01", "max_grad_norm": 0.8, "loss": 0.7644214630126953, "grad_norm": 1.3050304651260376, "learning_rate": 9.17340521114106e-05}
+{"ts": "2025-12-23T19:24:40", "event": "train_log", "step": 1024, "epoch": 0.4418554476806904, "progress_pct": 7.36, "epoch_pct": 7.36, "eta": "63:10:50", "max_grad_norm": 0.8, "loss": 0.8644799590110779, "grad_norm": 1.4643231630325317, "learning_rate": 9.191374663072777e-05}
+{"ts": "2025-12-23T19:24:58", "event": "train_log", "step": 1026, "epoch": 0.44271844660194176, "progress_pct": 7.38, "epoch_pct": 7.38, "eta": "63:06:42", "max_grad_norm": 0.8, "loss": 0.8281022906303406, "grad_norm": 1.4690274000167847, "learning_rate": 9.209344115004492e-05}
+{"ts": "2025-12-23T19:25:18", "event": "train_log", "step": 1028, "epoch": 0.4435814455231931, "progress_pct": 7.39, "epoch_pct": 7.39, "eta": "63:02:50", "max_grad_norm": 0.8, "loss": 0.7432141900062561, "grad_norm": 1.3321986198425293, "learning_rate": 9.227313566936209e-05}
+{"ts": "2025-12-23T19:25:37", "event": "train_log", "step": 1030, "epoch": 0.4444444444444444, "progress_pct": 7.41, "epoch_pct": 7.41, "eta": "62:58:58", "max_grad_norm": 0.8, "loss": 0.8143638372421265, "grad_norm": 1.4342927932739258, "learning_rate": 9.245283018867925e-05}
+{"ts": "2025-12-23T19:25:55", "event": "train_log", "step": 1032, "epoch": 0.4453074433656958, "progress_pct": 7.42, "epoch_pct": 7.42, "eta": "62:54:50", "max_grad_norm": 0.8, "loss": 0.8102239370346069, "grad_norm": 1.747295618057251, "learning_rate": 9.263252470799642e-05}
+{"ts": "2025-12-23T19:26:14", "event": "train_log", "step": 1034, "epoch": 0.44617044228694713, "progress_pct": 7.43, "epoch_pct": 7.44, "eta": "62:50:54", "max_grad_norm": 0.8, "loss": 0.7768088579177856, "grad_norm": 1.3273073434829712, "learning_rate": 9.281221922731357e-05}
+{"ts": "2025-12-23T19:26:34", "event": "train_log", "step": 1036, "epoch": 0.4470334412081985, "progress_pct": 7.45, "epoch_pct": 7.45, "eta": "62:47:03", "max_grad_norm": 0.8, "loss": 0.7886938452720642, "grad_norm": 1.5019903182983398, "learning_rate": 9.299191374663073e-05}
+{"ts": "2025-12-23T19:26:51", "event": "train_log", "step": 1038, "epoch": 0.44789644012944985, "progress_pct": 7.46, "epoch_pct": 7.46, "eta": "62:42:49", "max_grad_norm": 0.8, "loss": 0.8152258396148682, "grad_norm": 1.6711527109146118, "learning_rate": 9.31716082659479e-05}
+{"ts": "2025-12-23T19:27:09", "event": "train_log", "step": 1040, "epoch": 0.4487594390507012, "progress_pct": 7.48, "epoch_pct": 7.48, "eta": "62:38:41", "max_grad_norm": 0.8, "loss": 0.8339349627494812, "grad_norm": 1.8504793643951416, "learning_rate": 9.335130278526505e-05}
+{"ts": "2025-12-23T19:27:27", "event": "train_log", "step": 1042, "epoch": 0.44962243797195256, "progress_pct": 7.49, "epoch_pct": 7.49, "eta": "62:34:38", "max_grad_norm": 0.8, "loss": 0.7798911929130554, "grad_norm": 1.4333337545394897, "learning_rate": 9.353099730458221e-05}
+{"ts": "2025-12-23T19:27:46", "event": "train_log", "step": 1044, "epoch": 0.45048543689320386, "progress_pct": 7.51, "epoch_pct": 7.51, "eta": "62:30:44", "max_grad_norm": 0.8, "loss": 0.7676728963851929, "grad_norm": 1.3361700773239136, "learning_rate": 9.371069182389938e-05}
+{"ts": "2025-12-23T19:28:06", "event": "train_log", "step": 1046, "epoch": 0.4513484358144552, "progress_pct": 7.52, "epoch_pct": 7.52, "eta": "62:26:57", "max_grad_norm": 0.8, "loss": 0.7729206085205078, "grad_norm": 1.338934063911438, "learning_rate": 9.389038634321654e-05}
+{"ts": "2025-12-23T19:28:25", "event": "train_log", "step": 1048, "epoch": 0.4522114347357066, "progress_pct": 7.54, "epoch_pct": 7.54, "eta": "62:23:16", "max_grad_norm": 0.8, "loss": 0.7923640608787537, "grad_norm": 1.4347442388534546, "learning_rate": 9.40700808625337e-05}
+{"ts": "2025-12-23T19:28:44", "event": "train_log", "step": 1050, "epoch": 0.45307443365695793, "progress_pct": 7.55, "epoch_pct": 7.55, "eta": "62:19:17", "max_grad_norm": 0.8, "loss": 0.8068996071815491, "grad_norm": 1.4451628923416138, "learning_rate": 9.424977538185086e-05}
+{"ts": "2025-12-23T19:29:04", "event": "train_log", "step": 1052, "epoch": 0.4539374325782093, "progress_pct": 7.56, "epoch_pct": 7.57, "eta": "62:15:44", "max_grad_norm": 0.8, "loss": 0.7532910704612732, "grad_norm": 1.1904500722885132, "learning_rate": 9.442946990116803e-05}
+{"ts": "2025-12-23T19:29:22", "event": "train_log", "step": 1054, "epoch": 0.45480043149946064, "progress_pct": 7.58, "epoch_pct": 7.58, "eta": "62:11:45", "max_grad_norm": 0.8, "loss": 0.8236854672431946, "grad_norm": 1.4103976488113403, "learning_rate": 9.460916442048518e-05}
+{"ts": "2025-12-23T19:29:41", "event": "train_log", "step": 1056, "epoch": 0.455663430420712, "progress_pct": 7.59, "epoch_pct": 7.59, "eta": "62:07:51", "max_grad_norm": 0.8, "loss": 0.7532716393470764, "grad_norm": 1.3762754201889038, "learning_rate": 9.478885893980234e-05}
+{"ts": "2025-12-23T19:30:00", "event": "train_log", "step": 1058, "epoch": 0.4565264293419633, "progress_pct": 7.61, "epoch_pct": 7.61, "eta": "62:04:06", "max_grad_norm": 0.8, "loss": 0.8506031632423401, "grad_norm": 1.461421012878418, "learning_rate": 9.496855345911951e-05}
+{"ts": "2025-12-23T19:30:17", "event": "train_log", "step": 1060, "epoch": 0.45738942826321466, "progress_pct": 7.62, "epoch_pct": 7.62, "eta": "62:00:01", "max_grad_norm": 0.8, "loss": 0.7824664115905762, "grad_norm": 1.7207789421081543, "learning_rate": 9.514824797843667e-05}
+{"ts": "2025-12-23T19:30:35", "event": "train_log", "step": 1062, "epoch": 0.458252427184466, "progress_pct": 7.64, "epoch_pct": 7.64, "eta": "61:56:01", "max_grad_norm": 0.8, "loss": 0.8674250245094299, "grad_norm": 1.5946974754333496, "learning_rate": 9.532794249775382e-05}
+{"ts": "2025-12-23T19:30:55", "event": "train_log", "step": 1064, "epoch": 0.4591154261057174, "progress_pct": 7.65, "epoch_pct": 7.65, "eta": "61:52:28", "max_grad_norm": 0.8, "loss": 0.7851845622062683, "grad_norm": 1.3552565574645996, "learning_rate": 9.550763701707099e-05}
+{"ts": "2025-12-23T19:31:13", "event": "train_log", "step": 1066, "epoch": 0.45997842502696873, "progress_pct": 7.66, "epoch_pct": 7.67, "eta": "61:48:34", "max_grad_norm": 0.8, "loss": 0.8890700936317444, "grad_norm": 1.414617657661438, "learning_rate": 9.568733153638815e-05}
+{"ts": "2025-12-23T19:31:31", "event": "train_log", "step": 1068, "epoch": 0.4608414239482201, "progress_pct": 7.68, "epoch_pct": 7.68, "eta": "61:44:43", "max_grad_norm": 0.8, "loss": 0.8226593732833862, "grad_norm": 1.3322749137878418, "learning_rate": 9.58670260557053e-05}
+{"ts": "2025-12-23T19:31:50", "event": "train_log", "step": 1070, "epoch": 0.4617044228694714, "progress_pct": 7.69, "epoch_pct": 7.7, "eta": "61:40:56", "max_grad_norm": 0.8, "loss": 0.7663303017616272, "grad_norm": 1.41420316696167, "learning_rate": 9.604672057502247e-05}
+{"ts": "2025-12-23T19:32:10", "event": "train_log", "step": 1072, "epoch": 0.46256742179072274, "progress_pct": 7.71, "epoch_pct": 7.71, "eta": "61:37:33", "max_grad_norm": 0.8, "loss": 0.7122516632080078, "grad_norm": 1.324967384338379, "learning_rate": 9.622641509433963e-05}
+{"ts": "2025-12-23T19:32:30", "event": "train_log", "step": 1074, "epoch": 0.4634304207119741, "progress_pct": 7.72, "epoch_pct": 7.72, "eta": "61:34:05", "max_grad_norm": 0.8, "loss": 0.7177618145942688, "grad_norm": 1.2046154737472534, "learning_rate": 9.640610961365679e-05}
+{"ts": "2025-12-23T19:32:49", "event": "train_log", "step": 1076, "epoch": 0.46429341963322546, "progress_pct": 7.74, "epoch_pct": 7.74, "eta": "61:30:24", "max_grad_norm": 0.8, "loss": 0.7618930339813232, "grad_norm": 1.5221766233444214, "learning_rate": 9.658580413297395e-05}
+{"ts": "2025-12-23T19:33:10", "event": "train_log", "step": 1078, "epoch": 0.4651564185544768, "progress_pct": 7.75, "epoch_pct": 7.75, "eta": "61:27:00", "max_grad_norm": 0.8, "loss": 0.7483780384063721, "grad_norm": 1.3517179489135742, "learning_rate": 9.67654986522911e-05}
+{"ts": "2025-12-23T19:33:29", "event": "train_log", "step": 1080, "epoch": 0.46601941747572817, "progress_pct": 7.77, "epoch_pct": 7.77, "eta": "61:23:23", "max_grad_norm": 0.8, "loss": 0.8507670164108276, "grad_norm": 1.747865080833435, "learning_rate": 9.694519317160827e-05}
+{"ts": "2025-12-23T19:33:48", "event": "train_log", "step": 1082, "epoch": 0.4668824163969795, "progress_pct": 7.78, "epoch_pct": 7.78, "eta": "61:19:48", "max_grad_norm": 0.8, "loss": 0.7893540859222412, "grad_norm": 1.4183019399642944, "learning_rate": 9.712488769092543e-05}
+{"ts": "2025-12-23T19:34:07", "event": "train_log", "step": 1084, "epoch": 0.46774541531823083, "progress_pct": 7.79, "epoch_pct": 7.8, "eta": "61:16:15", "max_grad_norm": 0.8, "loss": 0.7891272306442261, "grad_norm": 1.6290462017059326, "learning_rate": 9.730458221024258e-05}
+{"ts": "2025-12-23T19:34:27", "event": "train_log", "step": 1086, "epoch": 0.4686084142394822, "progress_pct": 7.81, "epoch_pct": 7.81, "eta": "61:12:41", "max_grad_norm": 0.8, "loss": 0.841018795967102, "grad_norm": 1.6605802774429321, "learning_rate": 9.748427672955975e-05}
+{"ts": "2025-12-23T19:34:46", "event": "train_log", "step": 1088, "epoch": 0.46947141316073354, "progress_pct": 7.82, "epoch_pct": 7.82, "eta": "61:09:08", "max_grad_norm": 0.8, "loss": 0.8099003434181213, "grad_norm": 1.388964295387268, "learning_rate": 9.766397124887691e-05}
+{"ts": "2025-12-23T19:35:06", "event": "train_log", "step": 1090, "epoch": 0.4703344120819849, "progress_pct": 7.84, "epoch_pct": 7.84, "eta": "61:05:49", "max_grad_norm": 0.8, "loss": 0.7177658677101135, "grad_norm": 1.3901978731155396, "learning_rate": 9.784366576819407e-05}
+{"ts": "2025-12-23T19:35:26", "event": "train_log", "step": 1092, "epoch": 0.47119741100323626, "progress_pct": 7.85, "epoch_pct": 7.85, "eta": "61:02:21", "max_grad_norm": 0.8, "loss": 0.8285676836967468, "grad_norm": 1.3341423273086548, "learning_rate": 9.802336028751123e-05}
+{"ts": "2025-12-23T19:35:45", "event": "train_log", "step": 1094, "epoch": 0.4720604099244876, "progress_pct": 7.87, "epoch_pct": 7.87, "eta": "60:58:47", "max_grad_norm": 0.8, "loss": 0.7582666873931885, "grad_norm": 1.5451914072036743, "learning_rate": 9.82030548068284e-05}
+{"ts": "2025-12-23T19:36:04", "event": "train_log", "step": 1096, "epoch": 0.47292340884573897, "progress_pct": 7.88, "epoch_pct": 7.88, "eta": "60:55:18", "max_grad_norm": 0.8, "loss": 0.7293475270271301, "grad_norm": 1.5395640134811401, "learning_rate": 9.838274932614556e-05}
+{"ts": "2025-12-23T19:36:23", "event": "train_log", "step": 1098, "epoch": 0.47378640776699027, "progress_pct": 7.89, "epoch_pct": 7.9, "eta": "60:51:48", "max_grad_norm": 0.8, "loss": 0.8858493566513062, "grad_norm": 1.5005141496658325, "learning_rate": 9.856244384546271e-05}
+{"ts": "2025-12-23T19:36:42", "event": "train_log", "step": 1100, "epoch": 0.4746494066882416, "progress_pct": 7.91, "epoch_pct": 7.91, "eta": "60:48:16", "max_grad_norm": 0.8, "loss": 0.7176282405853271, "grad_norm": 1.3562248945236206, "learning_rate": 9.874213836477988e-05}
+{"ts": "2025-12-23T19:50:37", "event": "train_log", "step": 1100, "epoch": 0.4746494066882416, "progress_pct": 7.91, "epoch_pct": 7.91, "eta": "63:30:19", "max_grad_norm": 0.8, "eval_loss": 0.8152187466621399, "eval_runtime": 835.0394, "eval_samples_per_second": 2.467, "eval_steps_per_second": 2.467}
+{"ts": "2025-12-23T19:50:56", "event": "train_log", "step": 1102, "epoch": 0.475512405609493, "progress_pct": 7.92, "epoch_pct": 7.93, "eta": "63:26:35", "max_grad_norm": 0.8, "loss": 0.8048524856567383, "grad_norm": 1.5226716995239258, "learning_rate": 9.892183288409704e-05}
+{"ts": "2025-12-23T19:51:16", "event": "train_log", "step": 1104, "epoch": 0.47637540453074434, "progress_pct": 7.94, "epoch_pct": 7.94, "eta": "63:22:55", "max_grad_norm": 0.8, "loss": 0.7954329252243042, "grad_norm": 1.3544842004776, "learning_rate": 9.91015274034142e-05}
+{"ts": "2025-12-23T19:51:35", "event": "train_log", "step": 1106, "epoch": 0.4772384034519957, "progress_pct": 7.95, "epoch_pct": 7.95, "eta": "63:19:03", "max_grad_norm": 0.8, "loss": 0.804015040397644, "grad_norm": 1.5169771909713745, "learning_rate": 9.928122192273136e-05}
+{"ts": "2025-12-23T19:51:54", "event": "train_log", "step": 1108, "epoch": 0.47810140237324705, "progress_pct": 7.97, "epoch_pct": 7.97, "eta": "63:15:16", "max_grad_norm": 0.8, "loss": 0.7652344703674316, "grad_norm": 1.4011081457138062, "learning_rate": 9.946091644204852e-05}
+{"ts": "2025-12-23T19:52:14", "event": "train_log", "step": 1110, "epoch": 0.47896440129449835, "progress_pct": 7.98, "epoch_pct": 7.98, "eta": "63:11:36", "max_grad_norm": 0.8, "loss": 0.8501335382461548, "grad_norm": 1.4163414239883423, "learning_rate": 9.964061096136569e-05}
+{"ts": "2025-12-23T19:52:31", "event": "train_log", "step": 1112, "epoch": 0.4798274002157497, "progress_pct": 8.0, "epoch_pct": 8.0, "eta": "63:07:31", "max_grad_norm": 0.8, "loss": 0.9024044871330261, "grad_norm": 1.7406005859375, "learning_rate": 9.982030548068284e-05}
+{"ts": "2025-12-23T19:52:51", "event": "train_log", "step": 1114, "epoch": 0.48069039913700107, "progress_pct": 8.01, "epoch_pct": 8.01, "eta": "63:03:56", "max_grad_norm": 0.8, "loss": 0.7308287024497986, "grad_norm": 1.1490557193756104, "learning_rate": 0.0001}
+{"ts": "2025-12-23T19:53:11", "event": "train_log", "step": 1116, "epoch": 0.4815533980582524, "progress_pct": 8.02, "epoch_pct": 8.03, "eta": "63:00:26", "max_grad_norm": 0.8, "loss": 0.7273974418640137, "grad_norm": 1.2512775659561157, "learning_rate": 9.99999939713627e-05}
+{"ts": "2025-12-23T19:53:30", "event": "train_log", "step": 1118, "epoch": 0.4824163969795038, "progress_pct": 8.04, "epoch_pct": 8.04, "eta": "62:56:38", "max_grad_norm": 0.8, "loss": 0.8071790337562561, "grad_norm": 1.3944242000579834, "learning_rate": 9.999997588545227e-05}
+{"ts": "2025-12-23T19:53:50", "event": "train_log", "step": 1120, "epoch": 0.48327939590075514, "progress_pct": 8.05, "epoch_pct": 8.05, "eta": "62:53:09", "max_grad_norm": 0.8, "loss": 0.6586678624153137, "grad_norm": 1.2209112644195557, "learning_rate": 9.999994574227303e-05}
+{"ts": "2025-12-23T19:54:07", "event": "train_log", "step": 1122, "epoch": 0.4841423948220065, "progress_pct": 8.07, "epoch_pct": 8.07, "eta": "62:49:08", "max_grad_norm": 0.8, "loss": 0.8529233336448669, "grad_norm": 1.7328095436096191, "learning_rate": 9.999990354183228e-05}
+{"ts": "2025-12-23T19:54:27", "event": "train_log", "step": 1124, "epoch": 0.4850053937432578, "progress_pct": 8.08, "epoch_pct": 8.08, "eta": "62:45:35", "max_grad_norm": 0.8, "loss": 0.7529556155204773, "grad_norm": 1.3309327363967896, "learning_rate": 9.999984928414021e-05}
+{"ts": "2025-12-23T19:54:48", "event": "train_log", "step": 1126, "epoch": 0.48586839266450915, "progress_pct": 8.1, "epoch_pct": 8.1, "eta": "62:42:20", "max_grad_norm": 0.8, "loss": 0.6780388355255127, "grad_norm": 1.22034752368927, "learning_rate": 9.999978296920987e-05}
+{"ts": "2025-12-23T19:55:07", "event": "train_log", "step": 1128, "epoch": 0.4867313915857605, "progress_pct": 8.11, "epoch_pct": 8.11, "eta": "62:38:38", "max_grad_norm": 0.8, "loss": 0.7941693067550659, "grad_norm": 1.975162386894226, "learning_rate": 9.999970459705727e-05}
+{"ts": "2025-12-23T19:55:26", "event": "train_log", "step": 1130, "epoch": 0.48759439050701187, "progress_pct": 8.12, "epoch_pct": 8.13, "eta": "62:34:55", "max_grad_norm": 0.8, "loss": 0.7895156145095825, "grad_norm": 1.4939875602722168, "learning_rate": 9.999961416770132e-05}
+{"ts": "2025-12-23T19:55:45", "event": "train_log", "step": 1132, "epoch": 0.4884573894282632, "progress_pct": 8.14, "epoch_pct": 8.14, "eta": "62:31:12", "max_grad_norm": 0.8, "loss": 0.7662149667739868, "grad_norm": 1.4237034320831299, "learning_rate": 9.999951168116379e-05}
+{"ts": "2025-12-23T19:56:04", "event": "train_log", "step": 1134, "epoch": 0.4893203883495146, "progress_pct": 8.15, "epoch_pct": 8.16, "eta": "62:27:44", "max_grad_norm": 0.8, "loss": 0.8079231381416321, "grad_norm": 1.24979567527771, "learning_rate": 9.999939713746942e-05}
+{"ts": "2025-12-23T19:56:25", "event": "train_log", "step": 1136, "epoch": 0.49018338727076594, "progress_pct": 8.17, "epoch_pct": 8.17, "eta": "62:24:20", "max_grad_norm": 0.8, "loss": 0.7616434693336487, "grad_norm": 1.2252353429794312, "learning_rate": 9.999927053664585e-05}
+{"ts": "2025-12-23T19:56:44", "event": "train_log", "step": 1138, "epoch": 0.49104638619201724, "progress_pct": 8.18, "epoch_pct": 8.18, "eta": "62:20:48", "max_grad_norm": 0.8, "loss": 0.8576793074607849, "grad_norm": 1.2176212072372437, "learning_rate": 9.999913187872357e-05}
+{"ts": "2025-12-23T19:57:03", "event": "train_log", "step": 1140, "epoch": 0.4919093851132686, "progress_pct": 8.2, "epoch_pct": 8.2, "eta": "62:17:16", "max_grad_norm": 0.8, "loss": 0.8060292601585388, "grad_norm": 1.2108005285263062, "learning_rate": 9.999898116373605e-05}
+{"ts": "2025-12-23T19:57:23", "event": "train_log", "step": 1142, "epoch": 0.49277238403451995, "progress_pct": 8.21, "epoch_pct": 8.21, "eta": "62:13:45", "max_grad_norm": 0.8, "loss": 0.7532875537872314, "grad_norm": 1.6340605020523071, "learning_rate": 9.99988183917196e-05}
+{"ts": "2025-12-23T19:57:42", "event": "train_log", "step": 1144, "epoch": 0.4936353829557713, "progress_pct": 8.23, "epoch_pct": 8.23, "eta": "62:10:10", "max_grad_norm": 0.8, "loss": 0.7241997718811035, "grad_norm": 1.261606216430664, "learning_rate": 9.99986435627135e-05}
+{"ts": "2025-12-23T19:58:01", "event": "train_log", "step": 1146, "epoch": 0.49449838187702266, "progress_pct": 8.24, "epoch_pct": 8.24, "eta": "62:06:39", "max_grad_norm": 0.8, "loss": 0.78656005859375, "grad_norm": 1.2678134441375732, "learning_rate": 9.99984566767599e-05}
+{"ts": "2025-12-23T19:58:20", "event": "train_log", "step": 1148, "epoch": 0.495361380798274, "progress_pct": 8.25, "epoch_pct": 8.26, "eta": "62:03:03", "max_grad_norm": 0.8, "loss": 0.7423107624053955, "grad_norm": 1.4185810089111328, "learning_rate": 9.999825773390387e-05}
+{"ts": "2025-12-23T19:58:39", "event": "train_log", "step": 1150, "epoch": 0.4962243797195254, "progress_pct": 8.27, "epoch_pct": 8.27, "eta": "61:59:28", "max_grad_norm": 0.8, "loss": 0.8178372383117676, "grad_norm": 1.3218237161636353, "learning_rate": 9.999804673419338e-05}
+{"ts": "2025-12-23T19:58:58", "event": "train_log", "step": 1152, "epoch": 0.4970873786407767, "progress_pct": 8.28, "epoch_pct": 8.28, "eta": "61:56:00", "max_grad_norm": 0.8, "loss": 0.8016347885131836, "grad_norm": 1.9698240756988525, "learning_rate": 9.99978236776793e-05}
+{"ts": "2025-12-23T19:59:20", "event": "train_log", "step": 1154, "epoch": 0.49795037756202803, "progress_pct": 8.3, "epoch_pct": 8.3, "eta": "61:52:58", "max_grad_norm": 0.8, "loss": 0.782421886920929, "grad_norm": 1.2228590250015259, "learning_rate": 9.999758856441544e-05}
+{"ts": "2025-12-23T19:59:40", "event": "train_log", "step": 1156, "epoch": 0.4988133764832794, "progress_pct": 8.31, "epoch_pct": 8.31, "eta": "61:49:48", "max_grad_norm": 0.8, "loss": 0.7235615849494934, "grad_norm": 1.2753938436508179, "learning_rate": 9.999734139445849e-05}
+{"ts": "2025-12-23T19:59:59", "event": "train_log", "step": 1158, "epoch": 0.49967637540453075, "progress_pct": 8.33, "epoch_pct": 8.33, "eta": "61:46:12", "max_grad_norm": 0.8, "loss": 0.7686187028884888, "grad_norm": 1.353155493736267, "learning_rate": 9.999708216786805e-05}
+{"ts": "2025-12-23T20:00:19", "event": "train_log", "step": 1160, "epoch": 0.5005393743257821, "progress_pct": 8.34, "epoch_pct": 8.34, "eta": "61:42:59", "max_grad_norm": 0.8, "loss": 0.7842718362808228, "grad_norm": 1.4242770671844482, "learning_rate": 9.999681088470663e-05}
+{"ts": "2025-12-23T20:00:39", "event": "train_log", "step": 1162, "epoch": 0.5014023732470334, "progress_pct": 8.35, "epoch_pct": 8.36, "eta": "61:39:38", "max_grad_norm": 0.8, "loss": 0.8324156999588013, "grad_norm": 1.501827359199524, "learning_rate": 9.999652754503966e-05}
+{"ts": "2025-12-23T20:00:59", "event": "train_log", "step": 1164, "epoch": 0.5022653721682848, "progress_pct": 8.37, "epoch_pct": 8.37, "eta": "61:36:21", "max_grad_norm": 0.8, "loss": 0.7797855138778687, "grad_norm": 1.5403931140899658, "learning_rate": 9.999623214893545e-05}
+{"ts": "2025-12-23T20:01:20", "event": "train_log", "step": 1166, "epoch": 0.5031283710895361, "progress_pct": 8.38, "epoch_pct": 8.39, "eta": "61:33:12", "max_grad_norm": 0.8, "loss": 0.7667348384857178, "grad_norm": 1.4075721502304077, "learning_rate": 9.999592469646525e-05}
+{"ts": "2025-12-23T20:01:39", "event": "train_log", "step": 1168, "epoch": 0.5039913700107875, "progress_pct": 8.4, "epoch_pct": 8.4, "eta": "61:29:48", "max_grad_norm": 0.8, "loss": 0.8153020739555359, "grad_norm": 1.4411870241165161, "learning_rate": 9.99956051877032e-05}
+{"ts": "2025-12-23T20:01:58", "event": "train_log", "step": 1170, "epoch": 0.5048543689320388, "progress_pct": 8.41, "epoch_pct": 8.41, "eta": "61:26:16", "max_grad_norm": 0.8, "loss": 0.8548861742019653, "grad_norm": 1.4197871685028076, "learning_rate": 9.999527362272632e-05}
+{"ts": "2025-12-23T20:02:16", "event": "train_log", "step": 1172, "epoch": 0.5057173678532901, "progress_pct": 8.43, "epoch_pct": 8.43, "eta": "61:22:40", "max_grad_norm": 0.8, "loss": 0.7978097796440125, "grad_norm": 1.5030468702316284, "learning_rate": 9.999493000161458e-05}
+{"ts": "2025-12-23T20:02:35", "event": "train_log", "step": 1174, "epoch": 0.5065803667745415, "progress_pct": 8.44, "epoch_pct": 8.44, "eta": "61:19:16", "max_grad_norm": 0.8, "loss": 0.7990087866783142, "grad_norm": 1.483534574508667, "learning_rate": 9.999457432445088e-05}
+{"ts": "2025-12-23T20:02:53", "event": "train_log", "step": 1176, "epoch": 0.5074433656957928, "progress_pct": 8.46, "epoch_pct": 8.46, "eta": "61:15:39", "max_grad_norm": 0.8, "loss": 0.8072807788848877, "grad_norm": 1.5424920320510864, "learning_rate": 9.999420659132093e-05}
+{"ts": "2025-12-23T20:03:12", "event": "train_log", "step": 1178, "epoch": 0.5083063646170443, "progress_pct": 8.47, "epoch_pct": 8.47, "eta": "61:12:23", "max_grad_norm": 0.8, "loss": 0.7216097712516785, "grad_norm": 1.129503846168518, "learning_rate": 9.999382680231344e-05}
+{"ts": "2025-12-23T20:03:31", "event": "train_log", "step": 1180, "epoch": 0.5091693635382956, "progress_pct": 8.48, "epoch_pct": 8.49, "eta": "61:09:00", "max_grad_norm": 0.8, "loss": 0.8637414574623108, "grad_norm": 1.3208131790161133, "learning_rate": 9.999343495752001e-05}
+{"ts": "2025-12-23T20:03:49", "event": "train_log", "step": 1182, "epoch": 0.510032362459547, "progress_pct": 8.5, "epoch_pct": 8.5, "eta": "61:05:24", "max_grad_norm": 0.8, "loss": 0.8561362028121948, "grad_norm": 1.5239129066467285, "learning_rate": 9.999303105703508e-05}
+{"ts": "2025-12-23T20:04:08", "event": "train_log", "step": 1184, "epoch": 0.5108953613807983, "progress_pct": 8.51, "epoch_pct": 8.51, "eta": "61:02:04", "max_grad_norm": 0.8, "loss": 0.7514780759811401, "grad_norm": 1.3193504810333252, "learning_rate": 9.99926151009561e-05}
+{"ts": "2025-12-23T20:04:28", "event": "train_log", "step": 1186, "epoch": 0.5117583603020496, "progress_pct": 8.53, "epoch_pct": 8.53, "eta": "60:58:47", "max_grad_norm": 0.8, "loss": 0.8634516596794128, "grad_norm": 1.312649130821228, "learning_rate": 9.999218708938336e-05}
+{"ts": "2025-12-23T20:04:48", "event": "train_log", "step": 1188, "epoch": 0.512621359223301, "progress_pct": 8.54, "epoch_pct": 8.54, "eta": "60:55:41", "max_grad_norm": 0.8, "loss": 0.7493832111358643, "grad_norm": 1.2677429914474487, "learning_rate": 9.999174702242006e-05}
+{"ts": "2025-12-23T20:05:07", "event": "train_log", "step": 1190, "epoch": 0.5134843581445523, "progress_pct": 8.56, "epoch_pct": 8.56, "eta": "60:52:23", "max_grad_norm": 0.8, "loss": 0.8981389999389648, "grad_norm": 1.3901816606521606, "learning_rate": 9.999129490017234e-05}
+{"ts": "2025-12-23T20:05:27", "event": "train_log", "step": 1192, "epoch": 0.5143473570658037, "progress_pct": 8.57, "epoch_pct": 8.57, "eta": "60:49:18", "max_grad_norm": 0.8, "loss": 0.6771019697189331, "grad_norm": 1.232013463973999, "learning_rate": 9.999083072274921e-05}
+{"ts": "2025-12-23T20:05:48", "event": "train_log", "step": 1194, "epoch": 0.515210355987055, "progress_pct": 8.58, "epoch_pct": 8.59, "eta": "60:46:15", "max_grad_norm": 0.8, "loss": 0.789888858795166, "grad_norm": 1.1299076080322266, "learning_rate": 9.999035449026261e-05}
+{"ts": "2025-12-23T20:06:07", "event": "train_log", "step": 1196, "epoch": 0.5160733549083064, "progress_pct": 8.6, "epoch_pct": 8.6, "eta": "60:43:00", "max_grad_norm": 0.8, "loss": 0.6656325459480286, "grad_norm": 1.582215666770935, "learning_rate": 9.998986620282738e-05}
+{"ts": "2025-12-23T20:06:27", "event": "train_log", "step": 1198, "epoch": 0.5169363538295577, "progress_pct": 8.61, "epoch_pct": 8.62, "eta": "60:39:53", "max_grad_norm": 0.8, "loss": 0.795595645904541, "grad_norm": 1.3438260555267334, "learning_rate": 9.998936586056127e-05}
+{"ts": "2025-12-23T20:06:48", "event": "train_log", "step": 1200, "epoch": 0.517799352750809, "progress_pct": 8.63, "epoch_pct": 8.63, "eta": "60:36:56", "max_grad_norm": 0.8, "loss": 0.7769836187362671, "grad_norm": 1.45097017288208, "learning_rate": 9.998885346358495e-05}
+{"ts": "2025-12-23T20:20:32", "event": "train_log", "step": 1200, "epoch": 0.517799352750809, "progress_pct": 8.63, "epoch_pct": 8.63, "eta": "63:02:21", "max_grad_norm": 0.8, "eval_loss": 0.8070361018180847, "eval_runtime": 823.9296, "eval_samples_per_second": 2.5, "eval_steps_per_second": 2.5}
+{"ts": "2025-12-23T20:20:51", "event": "train_log", "step": 1202, "epoch": 0.5186623516720604, "progress_pct": 8.64, "epoch_pct": 8.64, "eta": "62:58:46", "max_grad_norm": 0.8, "loss": 0.7527318000793457, "grad_norm": 1.3791370391845703, "learning_rate": 9.998832901202195e-05}
+{"ts": "2025-12-23T20:21:09", "event": "train_log", "step": 1204, "epoch": 0.5195253505933117, "progress_pct": 8.66, "epoch_pct": 8.66, "eta": "62:55:10", "max_grad_norm": 0.8, "loss": 0.8021673560142517, "grad_norm": 1.2574659585952759, "learning_rate": 9.998779250599876e-05}
+{"ts": "2025-12-23T20:21:29", "event": "train_log", "step": 1206, "epoch": 0.5203883495145631, "progress_pct": 8.67, "epoch_pct": 8.67, "eta": "62:51:50", "max_grad_norm": 0.8, "loss": 0.776639461517334, "grad_norm": 1.3305809497833252, "learning_rate": 9.998724394564476e-05}
+{"ts": "2025-12-23T20:21:49", "event": "train_log", "step": 1208, "epoch": 0.5212513484358144, "progress_pct": 8.69, "epoch_pct": 8.69, "eta": "62:48:25", "max_grad_norm": 0.8, "loss": 0.7381181120872498, "grad_norm": 1.2330507040023804, "learning_rate": 9.998668333109223e-05}
+{"ts": "2025-12-23T20:22:11", "event": "train_log", "step": 1210, "epoch": 0.5221143473570659, "progress_pct": 8.7, "epoch_pct": 8.7, "eta": "62:45:24", "max_grad_norm": 0.8, "loss": 0.6595721244812012, "grad_norm": 1.0482324361801147, "learning_rate": 9.998611066247635e-05}
+{"ts": "2025-12-23T20:22:29", "event": "train_log", "step": 1212, "epoch": 0.5229773462783172, "progress_pct": 8.71, "epoch_pct": 8.72, "eta": "62:41:45", "max_grad_norm": 0.8, "loss": 0.8388534784317017, "grad_norm": 1.5456492900848389, "learning_rate": 9.998552593993522e-05}
+{"ts": "2025-12-23T20:22:48", "event": "train_log", "step": 1214, "epoch": 0.5238403451995685, "progress_pct": 8.73, "epoch_pct": 8.73, "eta": "62:38:21", "max_grad_norm": 0.8, "loss": 0.7593658566474915, "grad_norm": 1.3730167150497437, "learning_rate": 9.998492916360986e-05}
+{"ts": "2025-12-23T20:23:08", "event": "train_log", "step": 1216, "epoch": 0.5247033441208199, "progress_pct": 8.74, "epoch_pct": 8.75, "eta": "62:35:04", "max_grad_norm": 0.8, "loss": 0.7796505093574524, "grad_norm": 1.2668508291244507, "learning_rate": 9.998432033364416e-05}
+{"ts": "2025-12-23T20:23:29", "event": "train_log", "step": 1218, "epoch": 0.5255663430420712, "progress_pct": 8.76, "epoch_pct": 8.76, "eta": "62:31:54", "max_grad_norm": 0.8, "loss": 0.7493594288825989, "grad_norm": 1.1175068616867065, "learning_rate": 9.998369945018493e-05}
+{"ts": "2025-12-23T20:23:50", "event": "train_log", "step": 1220, "epoch": 0.5264293419633226, "progress_pct": 8.77, "epoch_pct": 8.77, "eta": "62:28:52", "max_grad_norm": 0.8, "loss": 0.7791322469711304, "grad_norm": 1.121848464012146, "learning_rate": 9.99830665133819e-05}
+{"ts": "2025-12-23T20:24:10", "event": "train_log", "step": 1222, "epoch": 0.5272923408845739, "progress_pct": 8.79, "epoch_pct": 8.79, "eta": "62:25:27", "max_grad_norm": 0.8, "loss": 0.8299469947814941, "grad_norm": 1.49788236618042, "learning_rate": 9.998242152338772e-05}
+{"ts": "2025-12-23T20:24:30", "event": "train_log", "step": 1224, "epoch": 0.5281553398058253, "progress_pct": 8.8, "epoch_pct": 8.8, "eta": "62:22:18", "max_grad_norm": 0.8, "loss": 0.7911439538002014, "grad_norm": 1.227125644683838, "learning_rate": 9.998176448035791e-05}
+{"ts": "2025-12-23T20:24:52", "event": "train_log", "step": 1226, "epoch": 0.5290183387270766, "progress_pct": 8.82, "epoch_pct": 8.82, "eta": "62:19:22", "max_grad_norm": 0.8, "loss": 0.6487553119659424, "grad_norm": 1.176081895828247, "learning_rate": 9.998109538445091e-05}
+{"ts": "2025-12-23T20:25:12", "event": "train_log", "step": 1228, "epoch": 0.5298813376483279, "progress_pct": 8.83, "epoch_pct": 8.83, "eta": "62:16:10", "max_grad_norm": 0.8, "loss": 0.7331781983375549, "grad_norm": 1.169835090637207, "learning_rate": 9.99804142358281e-05}
+{"ts": "2025-12-23T20:25:33", "event": "train_log", "step": 1230, "epoch": 0.5307443365695793, "progress_pct": 8.84, "epoch_pct": 8.85, "eta": "62:13:01", "max_grad_norm": 0.8, "loss": 0.8198121190071106, "grad_norm": 1.4085001945495605, "learning_rate": 9.997972103465367e-05}
+{"ts": "2025-12-23T20:25:52", "event": "train_log", "step": 1232, "epoch": 0.5316073354908306, "progress_pct": 8.86, "epoch_pct": 8.86, "eta": "62:09:43", "max_grad_norm": 0.8, "loss": 0.7099791169166565, "grad_norm": 1.8634135723114014, "learning_rate": 9.997901578109484e-05}
+{"ts": "2025-12-23T20:26:11", "event": "train_log", "step": 1234, "epoch": 0.532470334412082, "progress_pct": 8.87, "epoch_pct": 8.87, "eta": "62:06:23", "max_grad_norm": 0.8, "loss": 0.7416951656341553, "grad_norm": 1.4719511270523071, "learning_rate": 9.997829847532166e-05}
+{"ts": "2025-12-23T20:26:31", "event": "train_log", "step": 1236, "epoch": 0.5333333333333333, "progress_pct": 8.89, "epoch_pct": 8.89, "eta": "62:03:08", "max_grad_norm": 0.8, "loss": 0.7583408355712891, "grad_norm": 1.2430769205093384, "learning_rate": 9.99775691175071e-05}
+{"ts": "2025-12-23T20:26:50", "event": "train_log", "step": 1238, "epoch": 0.5341963322545846, "progress_pct": 8.9, "epoch_pct": 8.9, "eta": "61:59:48", "max_grad_norm": 0.8, "loss": 0.7639391422271729, "grad_norm": 1.2502130270004272, "learning_rate": 9.997682770782706e-05}
+{"ts": "2025-12-23T20:27:10", "event": "train_log", "step": 1240, "epoch": 0.535059331175836, "progress_pct": 8.92, "epoch_pct": 8.92, "eta": "61:56:31", "max_grad_norm": 0.8, "loss": 0.8304449319839478, "grad_norm": 1.3577396869659424, "learning_rate": 9.997607424646032e-05}
+{"ts": "2025-12-23T20:27:29", "event": "train_log", "step": 1242, "epoch": 0.5359223300970873, "progress_pct": 8.93, "epoch_pct": 8.93, "eta": "61:53:16", "max_grad_norm": 0.8, "loss": 0.7331952452659607, "grad_norm": 1.0986218452453613, "learning_rate": 9.997530873358854e-05}
+{"ts": "2025-12-23T20:27:48", "event": "train_log", "step": 1244, "epoch": 0.5367853290183388, "progress_pct": 8.94, "epoch_pct": 8.95, "eta": "61:49:50", "max_grad_norm": 0.8, "loss": 0.8173373341560364, "grad_norm": 1.4607514142990112, "learning_rate": 9.997453116939637e-05}
+{"ts": "2025-12-23T20:28:07", "event": "train_log", "step": 1246, "epoch": 0.53764832793959, "progress_pct": 8.96, "epoch_pct": 8.96, "eta": "61:46:31", "max_grad_norm": 0.8, "loss": 0.7677372694015503, "grad_norm": 1.2897552251815796, "learning_rate": 9.997374155407128e-05}
+{"ts": "2025-12-23T20:28:25", "event": "train_log", "step": 1248, "epoch": 0.5385113268608415, "progress_pct": 8.97, "epoch_pct": 8.98, "eta": "61:43:05", "max_grad_norm": 0.8, "loss": 0.8524296283721924, "grad_norm": 1.4581117630004883, "learning_rate": 9.997293988780372e-05}
+{"ts": "2025-12-23T20:28:44", "event": "train_log", "step": 1250, "epoch": 0.5393743257820928, "progress_pct": 8.99, "epoch_pct": 8.99, "eta": "61:39:44", "max_grad_norm": 0.8, "loss": 0.7920866012573242, "grad_norm": 1.2029366493225098, "learning_rate": 9.997212617078698e-05}
+{"ts": "2025-12-23T20:29:04", "event": "train_log", "step": 1252, "epoch": 0.5402373247033441, "progress_pct": 9.0, "epoch_pct": 9.0, "eta": "61:36:37", "max_grad_norm": 0.8, "loss": 0.7889736294746399, "grad_norm": 1.109483242034912, "learning_rate": 9.997130040321727e-05}
+{"ts": "2025-12-23T20:29:27", "event": "train_log", "step": 1254, "epoch": 0.5411003236245955, "progress_pct": 9.02, "epoch_pct": 9.02, "eta": "61:34:07", "max_grad_norm": 0.8, "loss": 0.6593763828277588, "grad_norm": 1.1930333375930786, "learning_rate": 9.997046258529377e-05}
+{"ts": "2025-12-23T20:29:45", "event": "train_log", "step": 1256, "epoch": 0.5419633225458468, "progress_pct": 9.03, "epoch_pct": 9.03, "eta": "61:30:35", "max_grad_norm": 0.8, "loss": 0.7768212556838989, "grad_norm": 1.3668162822723389, "learning_rate": 9.996961271721845e-05}
+{"ts": "2025-12-23T20:30:04", "event": "train_log", "step": 1258, "epoch": 0.5428263214670982, "progress_pct": 9.05, "epoch_pct": 9.05, "eta": "61:27:20", "max_grad_norm": 0.8, "loss": 0.7392467260360718, "grad_norm": 1.2901886701583862, "learning_rate": 9.996875079919632e-05}
+{"ts": "2025-12-23T20:30:23", "event": "train_log", "step": 1260, "epoch": 0.5436893203883495, "progress_pct": 9.06, "epoch_pct": 9.06, "eta": "61:24:07", "max_grad_norm": 0.8, "loss": 0.6961479187011719, "grad_norm": 1.366720199584961, "learning_rate": 9.996787683143518e-05}
+{"ts": "2025-12-23T20:30:42", "event": "train_log", "step": 1262, "epoch": 0.5445523193096009, "progress_pct": 9.07, "epoch_pct": 9.08, "eta": "61:20:53", "max_grad_norm": 0.8, "loss": 0.801517128944397, "grad_norm": 1.2925949096679688, "learning_rate": 9.99669908141458e-05}
+{"ts": "2025-12-23T20:31:03", "event": "train_log", "step": 1264, "epoch": 0.5454153182308522, "progress_pct": 9.09, "epoch_pct": 9.09, "eta": "61:17:52", "max_grad_norm": 0.8, "loss": 0.7287030816078186, "grad_norm": 1.1967244148254395, "learning_rate": 9.996609274754182e-05}
+{"ts": "2025-12-23T20:31:21", "event": "train_log", "step": 1266, "epoch": 0.5462783171521035, "progress_pct": 9.1, "epoch_pct": 9.1, "eta": "61:14:33", "max_grad_norm": 0.8, "loss": 0.7791255116462708, "grad_norm": 1.419996738433838, "learning_rate": 9.996518263183985e-05}
+{"ts": "2025-12-23T20:31:40", "event": "train_log", "step": 1268, "epoch": 0.5471413160733549, "progress_pct": 9.12, "epoch_pct": 9.12, "eta": "61:11:24", "max_grad_norm": 0.8, "loss": 0.6665297150611877, "grad_norm": 1.3544845581054688, "learning_rate": 9.996426046725933e-05}
+{"ts": "2025-12-23T20:32:00", "event": "train_log", "step": 1270, "epoch": 0.5480043149946062, "progress_pct": 9.13, "epoch_pct": 9.13, "eta": "61:08:17", "max_grad_norm": 0.8, "loss": 0.7902728915214539, "grad_norm": 1.6272345781326294, "learning_rate": 9.996332625402262e-05}
+{"ts": "2025-12-23T20:32:20", "event": "train_log", "step": 1272, "epoch": 0.5488673139158576, "progress_pct": 9.15, "epoch_pct": 9.15, "eta": "61:05:16", "max_grad_norm": 0.8, "loss": 0.8483795523643494, "grad_norm": 1.4264246225357056, "learning_rate": 9.996237999235502e-05}
+{"ts": "2025-12-23T20:32:41", "event": "train_log", "step": 1274, "epoch": 0.5497303128371089, "progress_pct": 9.16, "epoch_pct": 9.16, "eta": "61:02:19", "max_grad_norm": 0.8, "loss": 0.7432113289833069, "grad_norm": 1.3180512189865112, "learning_rate": 9.996142168248473e-05}
+{"ts": "2025-12-23T20:33:01", "event": "train_log", "step": 1276, "epoch": 0.5505933117583603, "progress_pct": 9.17, "epoch_pct": 9.18, "eta": "60:59:20", "max_grad_norm": 0.8, "loss": 0.6721545457839966, "grad_norm": 0.9958710074424744, "learning_rate": 9.996045132464283e-05}
+{"ts": "2025-12-23T20:33:21", "event": "train_log", "step": 1278, "epoch": 0.5514563106796116, "progress_pct": 9.19, "epoch_pct": 9.19, "eta": "60:56:18", "max_grad_norm": 0.8, "loss": 0.7770468592643738, "grad_norm": 1.371719241142273, "learning_rate": 9.995946891906331e-05}
+{"ts": "2025-12-23T20:33:40", "event": "train_log", "step": 1280, "epoch": 0.552319309600863, "progress_pct": 9.2, "epoch_pct": 9.21, "eta": "60:53:15", "max_grad_norm": 0.8, "loss": 0.816495418548584, "grad_norm": 1.1906969547271729, "learning_rate": 9.995847446598307e-05}
+{"ts": "2025-12-23T20:34:02", "event": "train_log", "step": 1282, "epoch": 0.5531823085221144, "progress_pct": 9.22, "epoch_pct": 9.22, "eta": "60:50:28", "max_grad_norm": 0.8, "loss": 0.7122345566749573, "grad_norm": 1.262184977531433, "learning_rate": 9.995746796564195e-05}
+{"ts": "2025-12-23T20:34:20", "event": "train_log", "step": 1284, "epoch": 0.5540453074433657, "progress_pct": 9.23, "epoch_pct": 9.23, "eta": "60:47:12", "max_grad_norm": 0.8, "loss": 0.8476479649543762, "grad_norm": 1.454692006111145, "learning_rate": 9.995644941828262e-05}
+{"ts": "2025-12-23T20:34:39", "event": "train_log", "step": 1286, "epoch": 0.5549083063646171, "progress_pct": 9.25, "epoch_pct": 9.25, "eta": "60:44:09", "max_grad_norm": 0.8, "loss": 0.7595248818397522, "grad_norm": 1.44277024269104, "learning_rate": 9.995541882415072e-05}
+{"ts": "2025-12-23T20:35:00", "event": "train_log", "step": 1288, "epoch": 0.5557713052858684, "progress_pct": 9.26, "epoch_pct": 9.26, "eta": "60:41:12", "max_grad_norm": 0.8, "loss": 0.7413201928138733, "grad_norm": 1.151465654373169, "learning_rate": 9.995437618349478e-05}
+{"ts": "2025-12-23T20:35:20", "event": "train_log", "step": 1290, "epoch": 0.5566343042071198, "progress_pct": 9.28, "epoch_pct": 9.28, "eta": "60:38:14", "max_grad_norm": 0.8, "loss": 0.7242997884750366, "grad_norm": 1.3160182237625122, "learning_rate": 9.995332149656621e-05}
+{"ts": "2025-12-23T20:35:39", "event": "train_log", "step": 1292, "epoch": 0.5574973031283711, "progress_pct": 9.29, "epoch_pct": 9.29, "eta": "60:35:12", "max_grad_norm": 0.8, "loss": 0.7779543995857239, "grad_norm": 1.2748576402664185, "learning_rate": 9.995225476361936e-05}
+{"ts": "2025-12-23T20:35:58", "event": "train_log", "step": 1294, "epoch": 0.5583603020496224, "progress_pct": 9.3, "epoch_pct": 9.31, "eta": "60:32:07", "max_grad_norm": 0.8, "loss": 0.8075780868530273, "grad_norm": 1.3639990091323853, "learning_rate": 9.995117598491145e-05}
+{"ts": "2025-12-23T20:36:19", "event": "train_log", "step": 1296, "epoch": 0.5592233009708738, "progress_pct": 9.32, "epoch_pct": 9.32, "eta": "60:29:22", "max_grad_norm": 0.8, "loss": 0.6940524578094482, "grad_norm": 1.0247999429702759, "learning_rate": 9.995008516070263e-05}
+{"ts": "2025-12-23T20:36:38", "event": "train_log", "step": 1298, "epoch": 0.5600862998921251, "progress_pct": 9.33, "epoch_pct": 9.33, "eta": "60:26:10", "max_grad_norm": 0.8, "loss": 0.799100399017334, "grad_norm": 1.3383198976516724, "learning_rate": 9.994898229125597e-05}
+{"ts": "2025-12-23T20:36:57", "event": "train_log", "step": 1300, "epoch": 0.5609492988133765, "progress_pct": 9.35, "epoch_pct": 9.35, "eta": "60:23:07", "max_grad_norm": 0.8, "loss": 0.8038942217826843, "grad_norm": 1.37660551071167, "learning_rate": 9.994786737683738e-05}
+{"ts": "2025-12-23T20:50:45", "event": "train_log", "step": 1300, "epoch": 0.5609492988133765, "progress_pct": 9.35, "epoch_pct": 9.35, "eta": "62:36:56", "max_grad_norm": 0.8, "eval_loss": 0.7962777018547058, "eval_runtime": 827.8798, "eval_samples_per_second": 2.488, "eval_steps_per_second": 2.488}
+{"ts": "2025-12-23T20:51:04", "event": "train_log", "step": 1302, "epoch": 0.5618122977346278, "progress_pct": 9.36, "epoch_pct": 9.36, "eta": "62:33:40", "max_grad_norm": 0.8, "loss": 0.8553245067596436, "grad_norm": 1.5424575805664062, "learning_rate": 9.994674041771574e-05}
+{"ts": "2025-12-23T20:51:22", "event": "train_log", "step": 1304, "epoch": 0.5626752966558792, "progress_pct": 9.38, "epoch_pct": 9.38, "eta": "62:30:18", "max_grad_norm": 0.8, "loss": 0.6650274991989136, "grad_norm": 1.1047813892364502, "learning_rate": 9.994560141416282e-05}
+{"ts": "2025-12-23T20:51:42", "event": "train_log", "step": 1306, "epoch": 0.5635382955771305, "progress_pct": 9.39, "epoch_pct": 9.39, "eta": "62:27:02", "max_grad_norm": 0.8, "loss": 0.794471025466919, "grad_norm": 1.4532393217086792, "learning_rate": 9.994445036645328e-05}
+{"ts": "2025-12-23T20:52:01", "event": "train_log", "step": 1308, "epoch": 0.5644012944983818, "progress_pct": 9.4, "epoch_pct": 9.41, "eta": "62:23:52", "max_grad_norm": 0.8, "loss": 0.7241487503051758, "grad_norm": 1.2588927745819092, "learning_rate": 9.994328727486468e-05}
+{"ts": "2025-12-23T20:52:21", "event": "train_log", "step": 1310, "epoch": 0.5652642934196332, "progress_pct": 9.42, "epoch_pct": 9.42, "eta": "62:20:39", "max_grad_norm": 0.8, "loss": 0.7781044244766235, "grad_norm": 1.4981781244277954, "learning_rate": 9.994211213967748e-05}
+{"ts": "2025-12-23T20:52:38", "event": "train_log", "step": 1312, "epoch": 0.5661272923408845, "progress_pct": 9.43, "epoch_pct": 9.44, "eta": "62:17:11", "max_grad_norm": 0.8, "loss": 0.7772427201271057, "grad_norm": 1.426277995109558, "learning_rate": 9.99409249611751e-05}
+{"ts": "2025-12-23T20:52:58", "event": "train_log", "step": 1314, "epoch": 0.566990291262136, "progress_pct": 9.45, "epoch_pct": 9.45, "eta": "62:14:06", "max_grad_norm": 0.8, "loss": 0.7443774342536926, "grad_norm": 1.1987197399139404, "learning_rate": 9.993972573964379e-05}
+{"ts": "2025-12-23T20:53:18", "event": "train_log", "step": 1316, "epoch": 0.5678532901833873, "progress_pct": 9.46, "epoch_pct": 9.46, "eta": "62:10:57", "max_grad_norm": 0.8, "loss": 0.7203393578529358, "grad_norm": 1.1490470170974731, "learning_rate": 9.993851447537275e-05}
+{"ts": "2025-12-23T20:53:36", "event": "train_log", "step": 1318, "epoch": 0.5687162891046387, "progress_pct": 9.48, "epoch_pct": 9.48, "eta": "62:07:39", "max_grad_norm": 0.8, "loss": 0.8038793206214905, "grad_norm": 1.3586851358413696, "learning_rate": 9.993729116865406e-05}
+{"ts": "2025-12-23T20:53:55", "event": "train_log", "step": 1320, "epoch": 0.56957928802589, "progress_pct": 9.49, "epoch_pct": 9.49, "eta": "62:04:22", "max_grad_norm": 0.8, "loss": 0.827308714389801, "grad_norm": 1.4160653352737427, "learning_rate": 9.993605581978274e-05}
+{"ts": "2025-12-23T20:54:14", "event": "train_log", "step": 1322, "epoch": 0.5704422869471413, "progress_pct": 9.51, "epoch_pct": 9.51, "eta": "62:01:05", "max_grad_norm": 0.8, "loss": 0.8183196783065796, "grad_norm": 1.2725251913070679, "learning_rate": 9.993480842905667e-05}
+{"ts": "2025-12-23T20:54:31", "event": "train_log", "step": 1324, "epoch": 0.5713052858683927, "progress_pct": 9.52, "epoch_pct": 9.52, "eta": "61:57:43", "max_grad_norm": 0.8, "loss": 0.8268020153045654, "grad_norm": 1.3789756298065186, "learning_rate": 9.993354899677665e-05}
+{"ts": "2025-12-23T20:54:51", "event": "train_log", "step": 1326, "epoch": 0.572168284789644, "progress_pct": 9.53, "epoch_pct": 9.54, "eta": "61:54:41", "max_grad_norm": 0.8, "loss": 0.7434715628623962, "grad_norm": 1.2641910314559937, "learning_rate": 9.993227752324639e-05}
+{"ts": "2025-12-23T20:55:12", "event": "train_log", "step": 1328, "epoch": 0.5730312837108954, "progress_pct": 9.55, "epoch_pct": 9.55, "eta": "61:51:47", "max_grad_norm": 0.8, "loss": 0.7636955976486206, "grad_norm": 1.2429786920547485, "learning_rate": 9.99309940087725e-05}
+{"ts": "2025-12-23T20:55:30", "event": "train_log", "step": 1330, "epoch": 0.5738942826321467, "progress_pct": 9.56, "epoch_pct": 9.56, "eta": "61:48:27", "max_grad_norm": 0.8, "loss": 0.794617772102356, "grad_norm": 1.3281534910202026, "learning_rate": 9.992969845366449e-05}
+{"ts": "2025-12-23T20:55:50", "event": "train_log", "step": 1332, "epoch": 0.574757281553398, "progress_pct": 9.58, "epoch_pct": 9.58, "eta": "61:45:29", "max_grad_norm": 0.8, "loss": 0.7658806443214417, "grad_norm": 1.2402182817459106, "learning_rate": 9.992839085823477e-05}
+{"ts": "2025-12-23T20:56:08", "event": "train_log", "step": 1334, "epoch": 0.5756202804746494, "progress_pct": 9.59, "epoch_pct": 9.59, "eta": "61:42:02", "max_grad_norm": 0.8, "loss": 0.8483052849769592, "grad_norm": 1.4143259525299072, "learning_rate": 9.99270712227987e-05}
+{"ts": "2025-12-23T20:56:27", "event": "train_log", "step": 1336, "epoch": 0.5764832793959007, "progress_pct": 9.61, "epoch_pct": 9.61, "eta": "61:38:54", "max_grad_norm": 0.8, "loss": 0.8610434532165527, "grad_norm": 1.2387818098068237, "learning_rate": 9.992573954767449e-05}
+{"ts": "2025-12-23T20:56:47", "event": "train_log", "step": 1338, "epoch": 0.5773462783171521, "progress_pct": 9.62, "epoch_pct": 9.62, "eta": "61:35:52", "max_grad_norm": 0.8, "loss": 0.8021210432052612, "grad_norm": 1.3307154178619385, "learning_rate": 9.992439583318322e-05}
+{"ts": "2025-12-23T20:57:05", "event": "train_log", "step": 1340, "epoch": 0.5782092772384034, "progress_pct": 9.63, "epoch_pct": 9.64, "eta": "61:32:41", "max_grad_norm": 0.8, "loss": 0.7536460757255554, "grad_norm": 1.293459415435791, "learning_rate": 9.992304007964896e-05}
+{"ts": "2025-12-23T20:57:24", "event": "train_log", "step": 1342, "epoch": 0.5790722761596548, "progress_pct": 9.65, "epoch_pct": 9.65, "eta": "61:29:34", "max_grad_norm": 0.8, "loss": 0.7519120573997498, "grad_norm": 1.1508948802947998, "learning_rate": 9.992167228739865e-05}
+{"ts": "2025-12-23T20:57:43", "event": "train_log", "step": 1344, "epoch": 0.5799352750809061, "progress_pct": 9.66, "epoch_pct": 9.67, "eta": "61:26:29", "max_grad_norm": 0.8, "loss": 0.784415066242218, "grad_norm": 1.5906150341033936, "learning_rate": 9.992029245676211e-05}
+{"ts": "2025-12-23T20:58:04", "event": "train_log", "step": 1346, "epoch": 0.5807982740021574, "progress_pct": 9.68, "epoch_pct": 9.68, "eta": "61:23:38", "max_grad_norm": 0.8, "loss": 0.8217549324035645, "grad_norm": 1.1421746015548706, "learning_rate": 9.99189005880721e-05}
+{"ts": "2025-12-23T20:58:25", "event": "train_log", "step": 1348, "epoch": 0.5816612729234089, "progress_pct": 9.69, "epoch_pct": 9.69, "eta": "61:20:47", "max_grad_norm": 0.8, "loss": 0.7446820735931396, "grad_norm": 1.0767277479171753, "learning_rate": 9.991749668166422e-05}
+{"ts": "2025-12-23T20:58:43", "event": "train_log", "step": 1350, "epoch": 0.5825242718446602, "progress_pct": 9.71, "epoch_pct": 9.71, "eta": "61:17:32", "max_grad_norm": 0.8, "loss": 0.8426011800765991, "grad_norm": 1.147594928741455, "learning_rate": 9.991608073787706e-05}
+{"ts": "2025-12-23T20:59:01", "event": "train_log", "step": 1352, "epoch": 0.5833872707659116, "progress_pct": 9.72, "epoch_pct": 9.72, "eta": "61:14:22", "max_grad_norm": 0.8, "loss": 0.8185623288154602, "grad_norm": 1.6618242263793945, "learning_rate": 9.991465275705206e-05}
+{"ts": "2025-12-23T20:59:19", "event": "train_log", "step": 1354, "epoch": 0.5842502696871629, "progress_pct": 9.74, "epoch_pct": 9.74, "eta": "61:11:07", "max_grad_norm": 0.8, "loss": 0.7521756887435913, "grad_norm": 1.414926290512085, "learning_rate": 9.991321273953355e-05}
+{"ts": "2025-12-23T20:59:39", "event": "train_log", "step": 1356, "epoch": 0.5851132686084143, "progress_pct": 9.75, "epoch_pct": 9.75, "eta": "61:08:11", "max_grad_norm": 0.8, "loss": 0.7517471313476562, "grad_norm": 1.298245906829834, "learning_rate": 9.991176068566882e-05}
+{"ts": "2025-12-23T20:59:59", "event": "train_log", "step": 1358, "epoch": 0.5859762675296656, "progress_pct": 9.76, "epoch_pct": 9.77, "eta": "61:05:19", "max_grad_norm": 0.8, "loss": 0.749305248260498, "grad_norm": 1.3243110179901123, "learning_rate": 9.9910296595808e-05}
+{"ts": "2025-12-23T21:00:19", "event": "train_log", "step": 1360, "epoch": 0.5868392664509169, "progress_pct": 9.78, "epoch_pct": 9.78, "eta": "61:02:21", "max_grad_norm": 0.8, "loss": 0.6883649230003357, "grad_norm": 1.1953799724578857, "learning_rate": 9.990882047030413e-05}
+{"ts": "2025-12-23T21:00:36", "event": "train_log", "step": 1362, "epoch": 0.5877022653721683, "progress_pct": 9.79, "epoch_pct": 9.8, "eta": "60:59:05", "max_grad_norm": 0.8, "loss": 0.7859632968902588, "grad_norm": 1.3934388160705566, "learning_rate": 9.990733230951322e-05}
+{"ts": "2025-12-23T21:00:56", "event": "train_log", "step": 1364, "epoch": 0.5885652642934196, "progress_pct": 9.81, "epoch_pct": 9.81, "eta": "60:56:10", "max_grad_norm": 0.8, "loss": 0.7313099503517151, "grad_norm": 1.2212340831756592, "learning_rate": 9.990583211379409e-05}
+{"ts": "2025-12-23T21:01:16", "event": "train_log", "step": 1366, "epoch": 0.589428263214671, "progress_pct": 9.82, "epoch_pct": 9.82, "eta": "60:53:15", "max_grad_norm": 0.8, "loss": 0.8007960319519043, "grad_norm": 1.5374075174331665, "learning_rate": 9.990431988350853e-05}
+{"ts": "2025-12-23T21:01:35", "event": "train_log", "step": 1368, "epoch": 0.5902912621359223, "progress_pct": 9.84, "epoch_pct": 9.84, "eta": "60:50:13", "max_grad_norm": 0.8, "loss": 0.773026168346405, "grad_norm": 1.1826236248016357, "learning_rate": 9.990279561902118e-05}
+{"ts": "2025-12-23T21:01:54", "event": "train_log", "step": 1370, "epoch": 0.5911542610571737, "progress_pct": 9.85, "epoch_pct": 9.85, "eta": "60:47:19", "max_grad_norm": 0.8, "loss": 0.7469213604927063, "grad_norm": 1.1181561946868896, "learning_rate": 9.990125932069964e-05}
+{"ts": "2025-12-23T21:02:13", "event": "train_log", "step": 1372, "epoch": 0.592017259978425, "progress_pct": 9.86, "epoch_pct": 9.87, "eta": "60:44:17", "max_grad_norm": 0.8, "loss": 0.7788950204849243, "grad_norm": 1.3735432624816895, "learning_rate": 9.989971098891439e-05}
+{"ts": "2025-12-23T21:02:32", "event": "train_log", "step": 1374, "epoch": 0.5928802588996763, "progress_pct": 9.88, "epoch_pct": 9.88, "eta": "60:41:18", "max_grad_norm": 0.8, "loss": 0.8117855191230774, "grad_norm": 1.1906198263168335, "learning_rate": 9.989815062403877e-05}
+{"ts": "2025-12-23T21:02:51", "event": "train_log", "step": 1376, "epoch": 0.5937432578209277, "progress_pct": 9.89, "epoch_pct": 9.9, "eta": "60:38:18", "max_grad_norm": 0.8, "loss": 0.7536540627479553, "grad_norm": 1.326940894126892, "learning_rate": 9.989657822644907e-05}
+{"ts": "2025-12-23T21:03:09", "event": "train_log", "step": 1378, "epoch": 0.594606256742179, "progress_pct": 9.91, "epoch_pct": 9.91, "eta": "60:35:08", "max_grad_norm": 0.8, "loss": 0.860320508480072, "grad_norm": 1.4508434534072876, "learning_rate": 9.989499379652447e-05}
+{"ts": "2025-12-23T21:03:30", "event": "train_log", "step": 1380, "epoch": 0.5954692556634305, "progress_pct": 9.92, "epoch_pct": 9.92, "eta": "60:32:28", "max_grad_norm": 0.8, "loss": 0.749508798122406, "grad_norm": 1.114499568939209, "learning_rate": 9.989339733464704e-05}
+{"ts": "2025-12-23T21:03:49", "event": "train_log", "step": 1382, "epoch": 0.5963322545846818, "progress_pct": 9.94, "epoch_pct": 9.94, "eta": "60:29:30", "max_grad_norm": 0.8, "loss": 0.7294815182685852, "grad_norm": 1.2856234312057495, "learning_rate": 9.989178884120176e-05}
+{"ts": "2025-12-23T21:04:07", "event": "train_log", "step": 1384, "epoch": 0.5971952535059332, "progress_pct": 9.95, "epoch_pct": 9.95, "eta": "60:26:27", "max_grad_norm": 0.8, "loss": 0.7797688245773315, "grad_norm": 1.291913628578186, "learning_rate": 9.989016831657652e-05}
+{"ts": "2025-12-23T21:04:26", "event": "train_log", "step": 1386, "epoch": 0.5980582524271845, "progress_pct": 9.97, "epoch_pct": 9.97, "eta": "60:23:26", "max_grad_norm": 0.8, "loss": 0.7824407815933228, "grad_norm": 1.1245452165603638, "learning_rate": 9.988853576116208e-05}
+{"ts": "2025-12-23T21:04:45", "event": "train_log", "step": 1388, "epoch": 0.5989212513484358, "progress_pct": 9.98, "epoch_pct": 9.98, "eta": "60:20:31", "max_grad_norm": 0.8, "loss": 0.714794933795929, "grad_norm": 1.2384979724884033, "learning_rate": 9.988689117535215e-05}
+{"ts": "2025-12-23T21:05:05", "event": "train_log", "step": 1390, "epoch": 0.5997842502696872, "progress_pct": 9.99, "epoch_pct": 10.0, "eta": "60:17:41", "max_grad_norm": 0.8, "loss": 0.7072597742080688, "grad_norm": 1.1945903301239014, "learning_rate": 9.988523455954332e-05}
+{"ts": "2025-12-23T21:05:24", "event": "train_log", "step": 1392, "epoch": 0.6006472491909385, "progress_pct": 10.01, "epoch_pct": 10.01, "eta": "60:14:48", "max_grad_norm": 0.8, "loss": 0.7075814008712769, "grad_norm": 1.3414266109466553, "learning_rate": 9.988356591413504e-05}
+{"ts": "2025-12-23T21:05:43", "event": "train_log", "step": 1394, "epoch": 0.6015102481121899, "progress_pct": 10.02, "epoch_pct": 10.03, "eta": "60:11:47", "max_grad_norm": 0.8, "loss": 0.8237126469612122, "grad_norm": 1.4575660228729248, "learning_rate": 9.988188523952973e-05}
+{"ts": "2025-12-23T21:06:01", "event": "train_log", "step": 1396, "epoch": 0.6023732470334412, "progress_pct": 10.04, "epoch_pct": 10.04, "eta": "60:08:47", "max_grad_norm": 0.8, "loss": 0.7991626858711243, "grad_norm": 1.3836476802825928, "learning_rate": 9.988019253613265e-05}
+{"ts": "2025-12-23T21:06:21", "event": "train_log", "step": 1398, "epoch": 0.6032362459546926, "progress_pct": 10.05, "epoch_pct": 10.05, "eta": "60:05:59", "max_grad_norm": 0.8, "loss": 0.6936330199241638, "grad_norm": 1.2176212072372437, "learning_rate": 9.987848780435201e-05}
+{"ts": "2025-12-23T21:06:41", "event": "train_log", "step": 1400, "epoch": 0.6040992448759439, "progress_pct": 10.07, "epoch_pct": 10.07, "eta": "60:03:17", "max_grad_norm": 0.8, "loss": 0.7834921479225159, "grad_norm": 1.171294093132019, "learning_rate": 9.98767710445989e-05}
+{"ts": "2025-12-23T21:20:37", "event": "train_log", "step": 1400, "epoch": 0.6040992448759439, "progress_pct": 10.07, "epoch_pct": 10.07, "eta": "62:07:47", "max_grad_norm": 0.8, "eval_loss": 0.7866972088813782, "eval_runtime": 836.0759, "eval_samples_per_second": 2.464, "eval_steps_per_second": 2.464}
+{"ts": "2025-12-23T21:20:55", "event": "train_log", "step": 1402, "epoch": 0.6049622437971952, "progress_pct": 10.08, "epoch_pct": 10.08, "eta": "62:04:37", "max_grad_norm": 0.8, "loss": 0.765119194984436, "grad_norm": 1.5934456586837769, "learning_rate": 9.987504225728727e-05}
+{"ts": "2025-12-23T21:21:14", "event": "train_log", "step": 1404, "epoch": 0.6058252427184466, "progress_pct": 10.09, "epoch_pct": 10.1, "eta": "62:01:27", "max_grad_norm": 0.8, "loss": 0.752061665058136, "grad_norm": 1.231372356414795, "learning_rate": 9.987330144283407e-05}
+{"ts": "2025-12-23T21:21:34", "event": "train_log", "step": 1406, "epoch": 0.6066882416396979, "progress_pct": 10.11, "epoch_pct": 10.11, "eta": "61:58:29", "max_grad_norm": 0.8, "loss": 0.7475255727767944, "grad_norm": 1.315215826034546, "learning_rate": 9.987154860165903e-05}
+{"ts": "2025-12-23T21:21:54", "event": "train_log", "step": 1408, "epoch": 0.6075512405609493, "progress_pct": 10.12, "epoch_pct": 10.13, "eta": "61:55:37", "max_grad_norm": 0.8, "loss": 0.777019202709198, "grad_norm": 1.2640653848648071, "learning_rate": 9.986978373418488e-05}
+{"ts": "2025-12-23T21:22:15", "event": "train_log", "step": 1410, "epoch": 0.6084142394822006, "progress_pct": 10.14, "epoch_pct": 10.14, "eta": "61:52:55", "max_grad_norm": 0.8, "loss": 0.6682190299034119, "grad_norm": 1.085493803024292, "learning_rate": 9.98680068408372e-05}
+{"ts": "2025-12-23T21:22:34", "event": "train_log", "step": 1412, "epoch": 0.609277238403452, "progress_pct": 10.15, "epoch_pct": 10.15, "eta": "61:49:49", "max_grad_norm": 0.8, "loss": 0.7765674591064453, "grad_norm": 1.4247251749038696, "learning_rate": 9.986621792204447e-05}
+{"ts": "2025-12-23T21:22:53", "event": "train_log", "step": 1414, "epoch": 0.6101402373247033, "progress_pct": 10.17, "epoch_pct": 10.17, "eta": "61:46:47", "max_grad_norm": 0.8, "loss": 0.7429938316345215, "grad_norm": 1.2993831634521484, "learning_rate": 9.986441697823808e-05}
+{"ts": "2025-12-23T21:23:12", "event": "train_log", "step": 1416, "epoch": 0.6110032362459547, "progress_pct": 10.18, "epoch_pct": 10.18, "eta": "61:43:46", "max_grad_norm": 0.8, "loss": 0.7838386297225952, "grad_norm": 2.614980936050415, "learning_rate": 9.986260400985233e-05}
+{"ts": "2025-12-23T21:23:32", "event": "train_log", "step": 1418, "epoch": 0.6118662351672061, "progress_pct": 10.2, "epoch_pct": 10.2, "eta": "61:40:47", "max_grad_norm": 0.8, "loss": 0.7679475545883179, "grad_norm": 1.4305615425109863, "learning_rate": 9.98607790173244e-05}
+{"ts": "2025-12-23T21:23:50", "event": "train_log", "step": 1420, "epoch": 0.6127292340884574, "progress_pct": 10.21, "epoch_pct": 10.21, "eta": "61:37:44", "max_grad_norm": 0.8, "loss": 0.6889669895172119, "grad_norm": 1.1412850618362427, "learning_rate": 9.98589420010944e-05}
+{"ts": "2025-12-23T21:24:08", "event": "train_log", "step": 1422, "epoch": 0.6135922330097088, "progress_pct": 10.22, "epoch_pct": 10.23, "eta": "61:34:32", "max_grad_norm": 0.8, "loss": 0.7659094333648682, "grad_norm": 1.372313141822815, "learning_rate": 9.98570929616053e-05}
+{"ts": "2025-12-23T21:24:28", "event": "train_log", "step": 1424, "epoch": 0.6144552319309601, "progress_pct": 10.24, "epoch_pct": 10.24, "eta": "61:31:37", "max_grad_norm": 0.8, "loss": 0.7404800653457642, "grad_norm": 1.2629581689834595, "learning_rate": 9.985523189930297e-05}
+{"ts": "2025-12-23T21:24:46", "event": "train_log", "step": 1426, "epoch": 0.6153182308522115, "progress_pct": 10.25, "epoch_pct": 10.26, "eta": "61:28:31", "max_grad_norm": 0.8, "loss": 0.8082308769226074, "grad_norm": 1.5652254819869995, "learning_rate": 9.985335881463623e-05}
+{"ts": "2025-12-23T21:25:04", "event": "train_log", "step": 1428, "epoch": 0.6161812297734628, "progress_pct": 10.27, "epoch_pct": 10.27, "eta": "61:25:24", "max_grad_norm": 0.8, "loss": 0.7773997187614441, "grad_norm": 1.2268856763839722, "learning_rate": 9.985147370805675e-05}
+{"ts": "2025-12-23T21:25:23", "event": "train_log", "step": 1430, "epoch": 0.6170442286947141, "progress_pct": 10.28, "epoch_pct": 10.28, "eta": "61:22:28", "max_grad_norm": 0.8, "loss": 0.7412725687026978, "grad_norm": 1.2190957069396973, "learning_rate": 9.984957658001912e-05}
+{"ts": "2025-12-23T21:25:43", "event": "train_log", "step": 1432, "epoch": 0.6179072276159655, "progress_pct": 10.3, "epoch_pct": 10.3, "eta": "61:19:33", "max_grad_norm": 0.8, "loss": 0.7257611751556396, "grad_norm": 1.481990098953247, "learning_rate": 9.984766743098081e-05}
+{"ts": "2025-12-23T21:26:02", "event": "train_log", "step": 1434, "epoch": 0.6187702265372168, "progress_pct": 10.31, "epoch_pct": 10.31, "eta": "61:16:41", "max_grad_norm": 0.8, "loss": 0.7151464819908142, "grad_norm": 1.0297027826309204, "learning_rate": 9.984574626140223e-05}
+{"ts": "2025-12-23T21:26:22", "event": "train_log", "step": 1436, "epoch": 0.6196332254584682, "progress_pct": 10.32, "epoch_pct": 10.33, "eta": "61:13:46", "max_grad_norm": 0.8, "loss": 0.7855883240699768, "grad_norm": 1.2404571771621704, "learning_rate": 9.984381307174663e-05}
+{"ts": "2025-12-23T21:26:40", "event": "train_log", "step": 1438, "epoch": 0.6204962243797195, "progress_pct": 10.34, "epoch_pct": 10.34, "eta": "61:10:46", "max_grad_norm": 0.8, "loss": 0.6360295414924622, "grad_norm": 1.2952802181243896, "learning_rate": 9.984186786248023e-05}
+{"ts": "2025-12-23T21:26:59", "event": "train_log", "step": 1440, "epoch": 0.6213592233009708, "progress_pct": 10.35, "epoch_pct": 10.36, "eta": "61:07:44", "max_grad_norm": 0.8, "loss": 0.7280036807060242, "grad_norm": 1.5405724048614502, "learning_rate": 9.983991063407206e-05}
+{"ts": "2025-12-23T21:27:18", "event": "train_log", "step": 1442, "epoch": 0.6222222222222222, "progress_pct": 10.37, "epoch_pct": 10.37, "eta": "61:04:46", "max_grad_norm": 0.8, "loss": 0.7553068995475769, "grad_norm": 1.303699254989624, "learning_rate": 9.983794138699411e-05}
+{"ts": "2025-12-23T21:27:37", "event": "train_log", "step": 1444, "epoch": 0.6230852211434735, "progress_pct": 10.38, "epoch_pct": 10.38, "eta": "61:01:51", "max_grad_norm": 0.8, "loss": 0.7329047918319702, "grad_norm": 1.1452997922897339, "learning_rate": 9.983596012172127e-05}
+{"ts": "2025-12-23T21:27:56", "event": "train_log", "step": 1446, "epoch": 0.623948220064725, "progress_pct": 10.4, "epoch_pct": 10.4, "eta": "60:58:56", "max_grad_norm": 0.8, "loss": 0.753168523311615, "grad_norm": 1.2648253440856934, "learning_rate": 9.983396683873133e-05}
+{"ts": "2025-12-23T21:28:16", "event": "train_log", "step": 1448, "epoch": 0.6248112189859762, "progress_pct": 10.41, "epoch_pct": 10.41, "eta": "60:56:13", "max_grad_norm": 0.8, "loss": 0.7378705739974976, "grad_norm": 1.2095013856887817, "learning_rate": 9.983196153850493e-05}
+{"ts": "2025-12-23T21:28:35", "event": "train_log", "step": 1450, "epoch": 0.6256742179072277, "progress_pct": 10.43, "epoch_pct": 10.43, "eta": "60:53:14", "max_grad_norm": 0.8, "loss": 0.7910448908805847, "grad_norm": 1.2855647802352905, "learning_rate": 9.982994422152564e-05}
+{"ts": "2025-12-23T21:28:53", "event": "train_log", "step": 1452, "epoch": 0.626537216828479, "progress_pct": 10.44, "epoch_pct": 10.44, "eta": "60:50:18", "max_grad_norm": 0.8, "loss": 0.785302996635437, "grad_norm": 1.3895999193191528, "learning_rate": 9.982791488827994e-05}
+{"ts": "2025-12-23T21:29:11", "event": "train_log", "step": 1454, "epoch": 0.6274002157497303, "progress_pct": 10.45, "epoch_pct": 10.46, "eta": "60:47:16", "max_grad_norm": 0.8, "loss": 0.7542029023170471, "grad_norm": 1.2398446798324585, "learning_rate": 9.982587353925719e-05}
+{"ts": "2025-12-23T21:29:30", "event": "train_log", "step": 1456, "epoch": 0.6282632146709817, "progress_pct": 10.47, "epoch_pct": 10.47, "eta": "60:44:22", "max_grad_norm": 0.8, "loss": 0.8081737756729126, "grad_norm": 1.2903637886047363, "learning_rate": 9.982382017494965e-05}
+{"ts": "2025-12-23T21:29:49", "event": "train_log", "step": 1458, "epoch": 0.629126213592233, "progress_pct": 10.48, "epoch_pct": 10.49, "eta": "60:41:30", "max_grad_norm": 0.8, "loss": 0.7886892557144165, "grad_norm": 1.223611831665039, "learning_rate": 9.98217547958525e-05}
+{"ts": "2025-12-23T21:30:07", "event": "train_log", "step": 1460, "epoch": 0.6299892125134844, "progress_pct": 10.5, "epoch_pct": 10.5, "eta": "60:38:30", "max_grad_norm": 0.8, "loss": 0.844115138053894, "grad_norm": 1.3036459684371948, "learning_rate": 9.981967740246377e-05}
+{"ts": "2025-12-23T21:30:26", "event": "train_log", "step": 1462, "epoch": 0.6308522114347357, "progress_pct": 10.51, "epoch_pct": 10.51, "eta": "60:35:37", "max_grad_norm": 0.8, "loss": 0.7784068584442139, "grad_norm": 1.6421090364456177, "learning_rate": 9.98175879952844e-05}
+{"ts": "2025-12-23T21:30:46", "event": "train_log", "step": 1464, "epoch": 0.6317152103559871, "progress_pct": 10.53, "epoch_pct": 10.53, "eta": "60:32:48", "max_grad_norm": 0.8, "loss": 0.746979832649231, "grad_norm": 1.244330644607544, "learning_rate": 9.981548657481828e-05}
+{"ts": "2025-12-23T21:31:05", "event": "train_log", "step": 1466, "epoch": 0.6325782092772384, "progress_pct": 10.54, "epoch_pct": 10.54, "eta": "60:30:04", "max_grad_norm": 0.8, "loss": 0.6680153608322144, "grad_norm": 1.0323492288589478, "learning_rate": 9.981337314157215e-05}
+{"ts": "2025-12-23T21:31:25", "event": "train_log", "step": 1468, "epoch": 0.6334412081984897, "progress_pct": 10.56, "epoch_pct": 10.56, "eta": "60:27:19", "max_grad_norm": 0.8, "loss": 0.7103193402290344, "grad_norm": 1.2238314151763916, "learning_rate": 9.981124769605563e-05}
+{"ts": "2025-12-23T21:31:45", "event": "train_log", "step": 1470, "epoch": 0.6343042071197411, "progress_pct": 10.57, "epoch_pct": 10.57, "eta": "60:24:37", "max_grad_norm": 0.8, "loss": 0.7426751255989075, "grad_norm": 1.2647801637649536, "learning_rate": 9.98091102387813e-05}
+{"ts": "2025-12-23T21:32:03", "event": "train_log", "step": 1472, "epoch": 0.6351672060409924, "progress_pct": 10.58, "epoch_pct": 10.59, "eta": "60:21:41", "max_grad_norm": 0.8, "loss": 0.7664520144462585, "grad_norm": 1.1923259496688843, "learning_rate": 9.980696077026456e-05}
+{"ts": "2025-12-23T21:32:24", "event": "train_log", "step": 1474, "epoch": 0.6360302049622438, "progress_pct": 10.6, "epoch_pct": 10.6, "eta": "60:19:02", "max_grad_norm": 0.8, "loss": 0.7348257899284363, "grad_norm": 1.0503894090652466, "learning_rate": 9.980479929102377e-05}
+{"ts": "2025-12-23T21:32:43", "event": "train_log", "step": 1476, "epoch": 0.6368932038834951, "progress_pct": 10.61, "epoch_pct": 10.61, "eta": "60:16:15", "max_grad_norm": 0.8, "loss": 0.687703013420105, "grad_norm": 1.2223190069198608, "learning_rate": 9.980262580158014e-05}
+{"ts": "2025-12-23T21:33:02", "event": "train_log", "step": 1478, "epoch": 0.6377562028047465, "progress_pct": 10.63, "epoch_pct": 10.63, "eta": "60:13:30", "max_grad_norm": 0.8, "loss": 0.7903051972389221, "grad_norm": 1.1469025611877441, "learning_rate": 9.980044030245782e-05}
+{"ts": "2025-12-23T21:33:20", "event": "train_log", "step": 1480, "epoch": 0.6386192017259978, "progress_pct": 10.64, "epoch_pct": 10.64, "eta": "60:10:30", "max_grad_norm": 0.8, "loss": 0.7920540571212769, "grad_norm": 1.2581169605255127, "learning_rate": 9.979824279418381e-05}
+{"ts": "2025-12-23T21:33:40", "event": "train_log", "step": 1482, "epoch": 0.6394822006472491, "progress_pct": 10.66, "epoch_pct": 10.66, "eta": "60:07:50", "max_grad_norm": 0.8, "loss": 0.7030006647109985, "grad_norm": 1.2616536617279053, "learning_rate": 9.979603327728806e-05}
+{"ts": "2025-12-23T21:34:00", "event": "train_log", "step": 1484, "epoch": 0.6403451995685006, "progress_pct": 10.67, "epoch_pct": 10.67, "eta": "60:05:11", "max_grad_norm": 0.8, "loss": 0.7489339709281921, "grad_norm": 1.3970606327056885, "learning_rate": 9.979381175230335e-05}
+{"ts": "2025-12-23T21:34:21", "event": "train_log", "step": 1486, "epoch": 0.6412081984897519, "progress_pct": 10.68, "epoch_pct": 10.69, "eta": "60:02:45", "max_grad_norm": 0.8, "loss": 0.6733010411262512, "grad_norm": 1.0060384273529053, "learning_rate": 9.979157821976542e-05}
+{"ts": "2025-12-23T21:34:40", "event": "train_log", "step": 1488, "epoch": 0.6420711974110033, "progress_pct": 10.7, "epoch_pct": 10.7, "eta": "59:59:55", "max_grad_norm": 0.8, "loss": 0.8221960663795471, "grad_norm": 1.4987449645996094, "learning_rate": 9.978933268021286e-05}
+{"ts": "2025-12-23T21:35:00", "event": "train_log", "step": 1490, "epoch": 0.6429341963322546, "progress_pct": 10.71, "epoch_pct": 10.72, "eta": "59:57:14", "max_grad_norm": 0.8, "loss": 0.8829001188278198, "grad_norm": 1.2843966484069824, "learning_rate": 9.978707513418719e-05}
+{"ts": "2025-12-23T21:35:21", "event": "train_log", "step": 1492, "epoch": 0.643797195253506, "progress_pct": 10.73, "epoch_pct": 10.73, "eta": "59:54:46", "max_grad_norm": 0.8, "loss": 0.7605764269828796, "grad_norm": 1.175879955291748, "learning_rate": 9.978480558223278e-05}
+{"ts": "2025-12-23T21:35:41", "event": "train_log", "step": 1494, "epoch": 0.6446601941747573, "progress_pct": 10.74, "epoch_pct": 10.74, "eta": "59:52:07", "max_grad_norm": 0.8, "loss": 0.7618141174316406, "grad_norm": 1.1991835832595825, "learning_rate": 9.978252402489692e-05}
+{"ts": "2025-12-23T21:36:00", "event": "train_log", "step": 1496, "epoch": 0.6455231930960086, "progress_pct": 10.76, "epoch_pct": 10.76, "eta": "59:49:25", "max_grad_norm": 0.8, "loss": 0.7071146368980408, "grad_norm": 1.167635202407837, "learning_rate": 9.978023046272983e-05}
+{"ts": "2025-12-23T21:36:20", "event": "train_log", "step": 1498, "epoch": 0.64638619201726, "progress_pct": 10.77, "epoch_pct": 10.77, "eta": "59:46:49", "max_grad_norm": 0.8, "loss": 0.7559816837310791, "grad_norm": 1.263371229171753, "learning_rate": 9.977792489628456e-05}
+{"ts": "2025-12-23T21:36:39", "event": "train_log", "step": 1500, "epoch": 0.6472491909385113, "progress_pct": 10.79, "epoch_pct": 10.79, "eta": "59:44:06", "max_grad_norm": 0.8, "loss": 0.7253944873809814, "grad_norm": 1.2108519077301025, "learning_rate": 9.977560732611713e-05}
+{"ts": "2025-12-23T21:50:35", "event": "train_log", "step": 1500, "epoch": 0.6472491909385113, "progress_pct": 10.79, "epoch_pct": 10.79, "eta": "61:39:24", "max_grad_norm": 0.8, "eval_loss": 0.778505802154541, "eval_runtime": 836.2317, "eval_samples_per_second": 2.463, "eval_steps_per_second": 2.463}
+{"ts": "2025-12-23T21:50:54", "event": "train_log", "step": 1502, "epoch": 0.6481121898597627, "progress_pct": 10.8, "epoch_pct": 10.8, "eta": "61:36:28", "max_grad_norm": 0.8, "loss": 0.8212620615959167, "grad_norm": 1.4794813394546509, "learning_rate": 9.977327775278636e-05}
+{"ts": "2025-12-23T21:51:14", "event": "train_log", "step": 1504, "epoch": 0.648975188781014, "progress_pct": 10.81, "epoch_pct": 10.82, "eta": "61:33:37", "max_grad_norm": 0.8, "loss": 0.7220476865768433, "grad_norm": 1.2120516300201416, "learning_rate": 9.977093617685404e-05}
+{"ts": "2025-12-23T21:51:32", "event": "train_log", "step": 1506, "epoch": 0.6498381877022654, "progress_pct": 10.83, "epoch_pct": 10.83, "eta": "61:30:39", "max_grad_norm": 0.8, "loss": 0.7791266441345215, "grad_norm": 1.3129044771194458, "learning_rate": 9.976858259888484e-05}
+{"ts": "2025-12-23T21:51:51", "event": "train_log", "step": 1508, "epoch": 0.6507011866235167, "progress_pct": 10.84, "epoch_pct": 10.85, "eta": "61:27:49", "max_grad_norm": 0.8, "loss": 0.7052001357078552, "grad_norm": 1.146490454673767, "learning_rate": 9.976621701944628e-05}
+{"ts": "2025-12-23T21:52:10", "event": "train_log", "step": 1510, "epoch": 0.651564185544768, "progress_pct": 10.86, "epoch_pct": 10.86, "eta": "61:24:52", "max_grad_norm": 0.8, "loss": 0.7135450839996338, "grad_norm": 1.1903722286224365, "learning_rate": 9.976383943910885e-05}
+{"ts": "2025-12-23T21:52:28", "event": "train_log", "step": 1512, "epoch": 0.6524271844660194, "progress_pct": 10.87, "epoch_pct": 10.87, "eta": "61:21:55", "max_grad_norm": 0.8, "loss": 0.8114889860153198, "grad_norm": 1.4106701612472534, "learning_rate": 9.976144985844589e-05}
+{"ts": "2025-12-23T21:52:47", "event": "train_log", "step": 1514, "epoch": 0.6532901833872707, "progress_pct": 10.89, "epoch_pct": 10.89, "eta": "61:19:03", "max_grad_norm": 0.8, "loss": 0.7501562237739563, "grad_norm": 1.245661973953247, "learning_rate": 9.97590482780336e-05}
+{"ts": "2025-12-23T21:53:07", "event": "train_log", "step": 1516, "epoch": 0.6541531823085222, "progress_pct": 10.9, "epoch_pct": 10.9, "eta": "61:16:17", "max_grad_norm": 0.8, "loss": 0.7697334289550781, "grad_norm": 1.2918543815612793, "learning_rate": 9.975663469845114e-05}
+{"ts": "2025-12-23T21:53:26", "event": "train_log", "step": 1518, "epoch": 0.6550161812297735, "progress_pct": 10.91, "epoch_pct": 10.92, "eta": "61:13:25", "max_grad_norm": 0.8, "loss": 0.7457168102264404, "grad_norm": 1.2699558734893799, "learning_rate": 9.975420912028053e-05}
+{"ts": "2025-12-23T21:53:44", "event": "train_log", "step": 1520, "epoch": 0.6558791801510249, "progress_pct": 10.93, "epoch_pct": 10.93, "eta": "61:10:26", "max_grad_norm": 0.8, "loss": 0.8368204236030579, "grad_norm": 1.3968263864517212, "learning_rate": 9.975177154410668e-05}
+{"ts": "2025-12-23T21:54:03", "event": "train_log", "step": 1522, "epoch": 0.6567421790722762, "progress_pct": 10.94, "epoch_pct": 10.95, "eta": "61:07:38", "max_grad_norm": 0.8, "loss": 0.7355788350105286, "grad_norm": 1.2161848545074463, "learning_rate": 9.974932197051741e-05}
+{"ts": "2025-12-23T21:54:23", "event": "train_log", "step": 1524, "epoch": 0.6576051779935275, "progress_pct": 10.96, "epoch_pct": 10.96, "eta": "61:04:54", "max_grad_norm": 0.8, "loss": 0.8023449182510376, "grad_norm": 1.129886507987976, "learning_rate": 9.974686040010341e-05}
+{"ts": "2025-12-23T21:54:41", "event": "train_log", "step": 1526, "epoch": 0.6584681769147789, "progress_pct": 10.97, "epoch_pct": 10.97, "eta": "61:01:57", "max_grad_norm": 0.8, "loss": 0.7965474724769592, "grad_norm": 1.1114557981491089, "learning_rate": 9.974438683345829e-05}
+{"ts": "2025-12-23T21:55:00", "event": "train_log", "step": 1528, "epoch": 0.6593311758360302, "progress_pct": 10.99, "epoch_pct": 10.99, "eta": "60:59:06", "max_grad_norm": 0.8, "loss": 0.7176838517189026, "grad_norm": 1.5741950273513794, "learning_rate": 9.974190127117853e-05}
+{"ts": "2025-12-23T21:55:20", "event": "train_log", "step": 1530, "epoch": 0.6601941747572816, "progress_pct": 11.0, "epoch_pct": 11.0, "eta": "60:56:25", "max_grad_norm": 0.8, "loss": 0.7114462852478027, "grad_norm": 1.1439597606658936, "learning_rate": 9.973940371386352e-05}
+{"ts": "2025-12-23T21:55:38", "event": "train_log", "step": 1532, "epoch": 0.6610571736785329, "progress_pct": 11.02, "epoch_pct": 11.02, "eta": "60:53:33", "max_grad_norm": 0.8, "loss": 0.754712700843811, "grad_norm": 1.273659348487854, "learning_rate": 9.973689416211553e-05}
+{"ts": "2025-12-23T21:55:58", "event": "train_log", "step": 1534, "epoch": 0.6619201725997842, "progress_pct": 11.03, "epoch_pct": 11.03, "eta": "60:50:51", "max_grad_norm": 0.8, "loss": 0.6720641255378723, "grad_norm": 3.9874372482299805, "learning_rate": 9.973437261653973e-05}
+{"ts": "2025-12-23T21:56:17", "event": "train_log", "step": 1536, "epoch": 0.6627831715210356, "progress_pct": 11.04, "epoch_pct": 11.05, "eta": "60:48:03", "max_grad_norm": 0.8, "loss": 0.793043851852417, "grad_norm": 1.2506502866744995, "learning_rate": 9.973183907774417e-05}
+{"ts": "2025-12-23T21:56:35", "event": "train_log", "step": 1538, "epoch": 0.6636461704422869, "progress_pct": 11.06, "epoch_pct": 11.06, "eta": "60:45:10", "max_grad_norm": 0.8, "loss": 0.7806380987167358, "grad_norm": 1.43257737159729, "learning_rate": 9.972929354633982e-05}
+{"ts": "2025-12-23T21:56:55", "event": "train_log", "step": 1540, "epoch": 0.6645091693635383, "progress_pct": 11.07, "epoch_pct": 11.08, "eta": "60:42:27", "max_grad_norm": 0.8, "loss": 0.7176088094711304, "grad_norm": 1.4794938564300537, "learning_rate": 9.97267360229405e-05}
+{"ts": "2025-12-23T21:57:13", "event": "train_log", "step": 1542, "epoch": 0.6653721682847896, "progress_pct": 11.09, "epoch_pct": 11.09, "eta": "60:39:38", "max_grad_norm": 0.8, "loss": 0.7145742774009705, "grad_norm": 1.2398169040679932, "learning_rate": 9.972416650816297e-05}
+{"ts": "2025-12-23T21:57:33", "event": "train_log", "step": 1544, "epoch": 0.666235167206041, "progress_pct": 11.1, "epoch_pct": 11.1, "eta": "60:36:54", "max_grad_norm": 0.8, "loss": 0.8025617003440857, "grad_norm": 1.1309953927993774, "learning_rate": 9.972158500262683e-05}
+{"ts": "2025-12-23T21:57:53", "event": "train_log", "step": 1546, "epoch": 0.6670981661272923, "progress_pct": 11.12, "epoch_pct": 11.12, "eta": "60:34:18", "max_grad_norm": 0.8, "loss": 0.7808589339256287, "grad_norm": 1.2479451894760132, "learning_rate": 9.971899150695463e-05}
+{"ts": "2025-12-23T21:58:12", "event": "train_log", "step": 1548, "epoch": 0.6679611650485436, "progress_pct": 11.13, "epoch_pct": 11.13, "eta": "60:31:36", "max_grad_norm": 0.8, "loss": 0.7096936106681824, "grad_norm": 1.2538350820541382, "learning_rate": 9.971638602177175e-05}
+{"ts": "2025-12-23T21:58:33", "event": "train_log", "step": 1550, "epoch": 0.668824163969795, "progress_pct": 11.14, "epoch_pct": 11.15, "eta": "60:29:06", "max_grad_norm": 0.8, "loss": 0.6529367566108704, "grad_norm": 1.1431525945663452, "learning_rate": 9.97137685477065e-05}
+{"ts": "2025-12-23T21:58:52", "event": "train_log", "step": 1552, "epoch": 0.6696871628910464, "progress_pct": 11.16, "epoch_pct": 11.16, "eta": "60:26:22", "max_grad_norm": 0.8, "loss": 0.7780346870422363, "grad_norm": 1.5966559648513794, "learning_rate": 9.971113908539008e-05}
+{"ts": "2025-12-23T21:59:12", "event": "train_log", "step": 1554, "epoch": 0.6705501618122978, "progress_pct": 11.17, "epoch_pct": 11.18, "eta": "60:23:44", "max_grad_norm": 0.8, "loss": 0.6889166831970215, "grad_norm": 1.186475157737732, "learning_rate": 9.970849763545658e-05}
+{"ts": "2025-12-23T21:59:31", "event": "train_log", "step": 1556, "epoch": 0.6714131607335491, "progress_pct": 11.19, "epoch_pct": 11.19, "eta": "60:21:04", "max_grad_norm": 0.8, "loss": 0.7154867649078369, "grad_norm": 1.1922476291656494, "learning_rate": 9.970584419854295e-05}
+{"ts": "2025-12-23T21:59:51", "event": "train_log", "step": 1558, "epoch": 0.6722761596548005, "progress_pct": 11.2, "epoch_pct": 11.2, "eta": "60:18:25", "max_grad_norm": 0.8, "loss": 0.7419366240501404, "grad_norm": 1.30132257938385, "learning_rate": 9.970317877528906e-05}
+{"ts": "2025-12-23T22:00:11", "event": "train_log", "step": 1560, "epoch": 0.6731391585760518, "progress_pct": 11.22, "epoch_pct": 11.22, "eta": "60:15:47", "max_grad_norm": 0.8, "loss": 0.7856602072715759, "grad_norm": 1.1539888381958008, "learning_rate": 9.970050136633769e-05}
+{"ts": "2025-12-23T22:00:30", "event": "train_log", "step": 1562, "epoch": 0.6740021574973031, "progress_pct": 11.23, "epoch_pct": 11.23, "eta": "60:13:07", "max_grad_norm": 0.8, "loss": 0.6834115386009216, "grad_norm": 1.3056161403656006, "learning_rate": 9.969781197233445e-05}
+{"ts": "2025-12-23T22:00:49", "event": "train_log", "step": 1564, "epoch": 0.6748651564185545, "progress_pct": 11.25, "epoch_pct": 11.25, "eta": "60:10:23", "max_grad_norm": 0.8, "loss": 0.8490727543830872, "grad_norm": 1.2899249792099, "learning_rate": 9.96951105939279e-05}
+{"ts": "2025-12-23T22:01:09", "event": "train_log", "step": 1566, "epoch": 0.6757281553398058, "progress_pct": 11.26, "epoch_pct": 11.26, "eta": "60:07:54", "max_grad_norm": 0.8, "loss": 0.7462526559829712, "grad_norm": 1.2182374000549316, "learning_rate": 9.969239723176946e-05}
+{"ts": "2025-12-23T22:01:29", "event": "train_log", "step": 1568, "epoch": 0.6765911542610572, "progress_pct": 11.27, "epoch_pct": 11.28, "eta": "60:05:18", "max_grad_norm": 0.8, "loss": 0.7145209908485413, "grad_norm": 1.1231273412704468, "learning_rate": 9.968967188651344e-05}
+{"ts": "2025-12-23T22:01:47", "event": "train_log", "step": 1570, "epoch": 0.6774541531823085, "progress_pct": 11.29, "epoch_pct": 11.29, "eta": "60:02:31", "max_grad_norm": 0.8, "loss": 0.7911707162857056, "grad_norm": 1.1776760816574097, "learning_rate": 9.968693455881703e-05}
+{"ts": "2025-12-23T22:02:08", "event": "train_log", "step": 1572, "epoch": 0.6783171521035599, "progress_pct": 11.3, "epoch_pct": 11.31, "eta": "60:00:03", "max_grad_norm": 0.8, "loss": 0.6946635246276855, "grad_norm": 1.3542319536209106, "learning_rate": 9.968418524934035e-05}
+{"ts": "2025-12-23T22:02:28", "event": "train_log", "step": 1574, "epoch": 0.6791801510248112, "progress_pct": 11.32, "epoch_pct": 11.32, "eta": "59:57:31", "max_grad_norm": 0.8, "loss": 0.7597488164901733, "grad_norm": 1.3073060512542725, "learning_rate": 9.96814239587464e-05}
+{"ts": "2025-12-23T22:02:49", "event": "train_log", "step": 1576, "epoch": 0.6800431499460625, "progress_pct": 11.33, "epoch_pct": 11.33, "eta": "59:55:07", "max_grad_norm": 0.8, "loss": 0.6763246655464172, "grad_norm": 1.3205567598342896, "learning_rate": 9.967865068770099e-05}
+{"ts": "2025-12-23T22:03:10", "event": "train_log", "step": 1578, "epoch": 0.6809061488673139, "progress_pct": 11.35, "epoch_pct": 11.35, "eta": "59:52:42", "max_grad_norm": 0.8, "loss": 0.7560655474662781, "grad_norm": 1.1734344959259033, "learning_rate": 9.967586543687295e-05}
+{"ts": "2025-12-23T22:03:26", "event": "train_log", "step": 1580, "epoch": 0.6817691477885652, "progress_pct": 11.36, "epoch_pct": 11.36, "eta": "59:49:42", "max_grad_norm": 0.8, "loss": 0.8050205707550049, "grad_norm": 1.445770025253296, "learning_rate": 9.967306820693388e-05}
+{"ts": "2025-12-23T22:03:48", "event": "train_log", "step": 1582, "epoch": 0.6826321467098166, "progress_pct": 11.37, "epoch_pct": 11.38, "eta": "59:47:19", "max_grad_norm": 0.8, "loss": 0.6628211140632629, "grad_norm": 1.1185686588287354, "learning_rate": 9.967025899855836e-05}
+{"ts": "2025-12-23T22:04:07", "event": "train_log", "step": 1584, "epoch": 0.683495145631068, "progress_pct": 11.39, "epoch_pct": 11.39, "eta": "59:44:44", "max_grad_norm": 0.8, "loss": 0.7684114575386047, "grad_norm": 1.3831934928894043, "learning_rate": 9.966743781242379e-05}
+{"ts": "2025-12-23T22:04:28", "event": "train_log", "step": 1586, "epoch": 0.6843581445523194, "progress_pct": 11.4, "epoch_pct": 11.41, "eta": "59:42:22", "max_grad_norm": 0.8, "loss": 0.7084140777587891, "grad_norm": 1.2417019605636597, "learning_rate": 9.966460464921047e-05}
+{"ts": "2025-12-23T22:04:47", "event": "train_log", "step": 1588, "epoch": 0.6852211434735707, "progress_pct": 11.42, "epoch_pct": 11.42, "eta": "59:39:46", "max_grad_norm": 0.8, "loss": 0.6817399263381958, "grad_norm": 1.2138086557388306, "learning_rate": 9.966175950960163e-05}
+{"ts": "2025-12-23T22:05:07", "event": "train_log", "step": 1590, "epoch": 0.686084142394822, "progress_pct": 11.43, "epoch_pct": 11.43, "eta": "59:37:13", "max_grad_norm": 0.8, "loss": 0.8422354459762573, "grad_norm": 1.399778962135315, "learning_rate": 9.965890239428339e-05}
+{"ts": "2025-12-23T22:05:28", "event": "train_log", "step": 1592, "epoch": 0.6869471413160734, "progress_pct": 11.45, "epoch_pct": 11.45, "eta": "59:34:50", "max_grad_norm": 0.8, "loss": 0.7432681322097778, "grad_norm": 1.4348714351654053, "learning_rate": 9.965603330394468e-05}
+{"ts": "2025-12-23T22:05:48", "event": "train_log", "step": 1594, "epoch": 0.6878101402373247, "progress_pct": 11.46, "epoch_pct": 11.46, "eta": "59:32:20", "max_grad_norm": 0.8, "loss": 0.7571184635162354, "grad_norm": 1.192425012588501, "learning_rate": 9.965315223927741e-05}
+{"ts": "2025-12-23T22:06:07", "event": "train_log", "step": 1596, "epoch": 0.6886731391585761, "progress_pct": 11.48, "epoch_pct": 11.48, "eta": "59:29:46", "max_grad_norm": 0.8, "loss": 0.802443265914917, "grad_norm": 1.2834038734436035, "learning_rate": 9.96502592009763e-05}
+{"ts": "2025-12-23T22:06:26", "event": "train_log", "step": 1598, "epoch": 0.6895361380798274, "progress_pct": 11.49, "epoch_pct": 11.49, "eta": "59:27:09", "max_grad_norm": 0.8, "loss": 0.8113405704498291, "grad_norm": 1.1650444269180298, "learning_rate": 9.964735418973901e-05}
+{"ts": "2025-12-23T22:06:47", "event": "train_log", "step": 1600, "epoch": 0.6903991370010788, "progress_pct": 11.5, "epoch_pct": 11.51, "eta": "59:24:45", "max_grad_norm": 0.8, "loss": 0.7507118582725525, "grad_norm": 1.1714438199996948, "learning_rate": 9.964443720626606e-05}
+{"ts": "2025-12-23T22:20:28", "event": "train_log", "step": 1600, "epoch": 0.6903991370010788, "progress_pct": 11.5, "epoch_pct": 11.51, "eta": "61:10:04", "max_grad_norm": 0.8, "eval_loss": 0.7706249952316284, "eval_runtime": 821.4894, "eval_samples_per_second": 2.508, "eval_steps_per_second": 2.508}
+{"ts": "2025-12-23T22:20:48", "event": "train_log", "step": 1602, "epoch": 0.6912621359223301, "progress_pct": 11.52, "epoch_pct": 11.52, "eta": "61:07:27", "max_grad_norm": 0.8, "loss": 0.7422796487808228, "grad_norm": 1.175459861755371, "learning_rate": 9.964150825126088e-05}
+{"ts": "2025-12-23T22:21:08", "event": "train_log", "step": 1604, "epoch": 0.6921251348435814, "progress_pct": 11.53, "epoch_pct": 11.54, "eta": "61:04:49", "max_grad_norm": 0.8, "loss": 0.7125853300094604, "grad_norm": 1.6418036222457886, "learning_rate": 9.963856732542978e-05}
+{"ts": "2025-12-23T22:21:27", "event": "train_log", "step": 1606, "epoch": 0.6929881337648328, "progress_pct": 11.55, "epoch_pct": 11.55, "eta": "61:02:02", "max_grad_norm": 0.8, "loss": 0.757859468460083, "grad_norm": 1.3256275653839111, "learning_rate": 9.963561442948193e-05}
+{"ts": "2025-12-23T22:21:47", "event": "train_log", "step": 1608, "epoch": 0.6938511326860841, "progress_pct": 11.56, "epoch_pct": 11.56, "eta": "60:59:32", "max_grad_norm": 0.8, "loss": 0.6849292516708374, "grad_norm": 1.19981050491333, "learning_rate": 9.963264956412944e-05}
+{"ts": "2025-12-23T22:22:08", "event": "train_log", "step": 1610, "epoch": 0.6947141316073355, "progress_pct": 11.58, "epoch_pct": 11.58, "eta": "60:57:00", "max_grad_norm": 0.8, "loss": 0.7636215090751648, "grad_norm": 1.0392169952392578, "learning_rate": 9.962967273008723e-05}
+{"ts": "2025-12-23T22:22:29", "event": "train_log", "step": 1612, "epoch": 0.6955771305285868, "progress_pct": 11.59, "epoch_pct": 11.59, "eta": "60:54:30", "max_grad_norm": 0.8, "loss": 0.7208617329597473, "grad_norm": 1.1106646060943604, "learning_rate": 9.962668392807318e-05}
+{"ts": "2025-12-23T22:22:49", "event": "train_log", "step": 1614, "epoch": 0.6964401294498382, "progress_pct": 11.6, "epoch_pct": 11.61, "eta": "60:51:57", "max_grad_norm": 0.8, "loss": 0.671805739402771, "grad_norm": 1.1569621562957764, "learning_rate": 9.962368315880801e-05}
+{"ts": "2025-12-23T22:23:08", "event": "train_log", "step": 1616, "epoch": 0.6973031283710895, "progress_pct": 11.62, "epoch_pct": 11.62, "eta": "60:49:19", "max_grad_norm": 0.8, "loss": 0.7648762464523315, "grad_norm": 1.2157769203186035, "learning_rate": 9.962067042301537e-05}
+{"ts": "2025-12-23T22:23:27", "event": "train_log", "step": 1618, "epoch": 0.6981661272923408, "progress_pct": 11.63, "epoch_pct": 11.64, "eta": "60:46:35", "max_grad_norm": 0.8, "loss": 0.8490509390830994, "grad_norm": 1.4804017543792725, "learning_rate": 9.961764572142173e-05}
+{"ts": "2025-12-23T22:23:48", "event": "train_log", "step": 1620, "epoch": 0.6990291262135923, "progress_pct": 11.65, "epoch_pct": 11.65, "eta": "60:44:07", "max_grad_norm": 0.8, "loss": 0.5950891971588135, "grad_norm": 1.0344431400299072, "learning_rate": 9.961460905475651e-05}
+{"ts": "2025-12-23T22:24:08", "event": "train_log", "step": 1622, "epoch": 0.6998921251348436, "progress_pct": 11.66, "epoch_pct": 11.66, "eta": "60:41:35", "max_grad_norm": 0.8, "loss": 0.7288329005241394, "grad_norm": 1.1109060049057007, "learning_rate": 9.9611560423752e-05}
+{"ts": "2025-12-23T22:24:28", "event": "train_log", "step": 1624, "epoch": 0.700755124056095, "progress_pct": 11.68, "epoch_pct": 11.68, "eta": "60:39:00", "max_grad_norm": 0.8, "loss": 0.7074542045593262, "grad_norm": 1.096996545791626, "learning_rate": 9.960849982914331e-05}
+{"ts": "2025-12-23T22:24:46", "event": "train_log", "step": 1626, "epoch": 0.7016181229773463, "progress_pct": 11.69, "epoch_pct": 11.69, "eta": "60:36:15", "max_grad_norm": 0.8, "loss": 0.7348865270614624, "grad_norm": 1.3109341859817505, "learning_rate": 9.960542727166853e-05}
+{"ts": "2025-12-23T22:25:06", "event": "train_log", "step": 1628, "epoch": 0.7024811218985976, "progress_pct": 11.71, "epoch_pct": 11.71, "eta": "60:33:41", "max_grad_norm": 0.8, "loss": 0.7658320665359497, "grad_norm": 1.3402032852172852, "learning_rate": 9.96023427520686e-05}
+{"ts": "2025-12-23T22:25:25", "event": "train_log", "step": 1630, "epoch": 0.703344120819849, "progress_pct": 11.72, "epoch_pct": 11.72, "eta": "60:30:58", "max_grad_norm": 0.8, "loss": 0.809951901435852, "grad_norm": 1.3712561130523682, "learning_rate": 9.959924627108732e-05}
+{"ts": "2025-12-23T22:25:44", "event": "train_log", "step": 1632, "epoch": 0.7042071197411003, "progress_pct": 11.73, "epoch_pct": 11.74, "eta": "60:28:22", "max_grad_norm": 0.8, "loss": 0.6834598779678345, "grad_norm": 1.2278292179107666, "learning_rate": 9.959613782947139e-05}
+{"ts": "2025-12-23T22:26:04", "event": "train_log", "step": 1634, "epoch": 0.7050701186623517, "progress_pct": 11.75, "epoch_pct": 11.75, "eta": "60:25:51", "max_grad_norm": 0.8, "loss": 0.7416030764579773, "grad_norm": 1.0510832071304321, "learning_rate": 9.95930174279704e-05}
+{"ts": "2025-12-23T22:26:24", "event": "train_log", "step": 1636, "epoch": 0.705933117583603, "progress_pct": 11.76, "epoch_pct": 11.77, "eta": "60:23:15", "max_grad_norm": 0.8, "loss": 0.7304092645645142, "grad_norm": 1.130991816520691, "learning_rate": 9.958988506733684e-05}
+{"ts": "2025-12-23T22:26:43", "event": "train_log", "step": 1638, "epoch": 0.7067961165048544, "progress_pct": 11.78, "epoch_pct": 11.78, "eta": "60:20:36", "max_grad_norm": 0.8, "loss": 0.8230003118515015, "grad_norm": 1.3328672647476196, "learning_rate": 9.958674074832603e-05}
+{"ts": "2025-12-23T22:27:01", "event": "train_log", "step": 1640, "epoch": 0.7076591154261057, "progress_pct": 11.79, "epoch_pct": 11.79, "eta": "60:17:54", "max_grad_norm": 0.8, "loss": 0.8413107991218567, "grad_norm": 1.3842110633850098, "learning_rate": 9.958358447169625e-05}
+{"ts": "2025-12-23T22:27:19", "event": "train_log", "step": 1642, "epoch": 0.708522114347357, "progress_pct": 11.81, "epoch_pct": 11.81, "eta": "60:15:11", "max_grad_norm": 0.8, "loss": 0.77432781457901, "grad_norm": 1.3143812417984009, "learning_rate": 9.958041623820858e-05}
+{"ts": "2025-12-23T22:27:38", "event": "train_log", "step": 1644, "epoch": 0.7093851132686084, "progress_pct": 11.82, "epoch_pct": 11.82, "eta": "60:12:34", "max_grad_norm": 0.8, "loss": 0.7563844323158264, "grad_norm": 1.4727009534835815, "learning_rate": 9.957723604862704e-05}
+{"ts": "2025-12-23T22:27:58", "event": "train_log", "step": 1646, "epoch": 0.7102481121898597, "progress_pct": 11.83, "epoch_pct": 11.84, "eta": "60:10:00", "max_grad_norm": 0.8, "loss": 0.7085853219032288, "grad_norm": 1.1892327070236206, "learning_rate": 9.957404390371854e-05}
+{"ts": "2025-12-23T22:28:18", "event": "train_log", "step": 1648, "epoch": 0.7111111111111111, "progress_pct": 11.85, "epoch_pct": 11.85, "eta": "60:07:29", "max_grad_norm": 0.8, "loss": 0.6837730407714844, "grad_norm": 1.124180555343628, "learning_rate": 9.957083980425282e-05}
+{"ts": "2025-12-23T22:28:38", "event": "train_log", "step": 1650, "epoch": 0.7119741100323624, "progress_pct": 11.86, "epoch_pct": 11.87, "eta": "60:05:04", "max_grad_norm": 0.8, "loss": 0.7255740165710449, "grad_norm": 1.2224774360656738, "learning_rate": 9.956762375100255e-05}
+{"ts": "2025-12-23T22:28:58", "event": "train_log", "step": 1652, "epoch": 0.7128371089536139, "progress_pct": 11.88, "epoch_pct": 11.88, "eta": "60:02:31", "max_grad_norm": 0.8, "loss": 0.7849710583686829, "grad_norm": 1.360425591468811, "learning_rate": 9.956439574474326e-05}
+{"ts": "2025-12-23T22:29:17", "event": "train_log", "step": 1654, "epoch": 0.7137001078748652, "progress_pct": 11.89, "epoch_pct": 11.9, "eta": "59:59:55", "max_grad_norm": 0.8, "loss": 0.7669665813446045, "grad_norm": 1.2818151712417603, "learning_rate": 9.956115578625338e-05}
+{"ts": "2025-12-23T22:29:35", "event": "train_log", "step": 1656, "epoch": 0.7145631067961165, "progress_pct": 11.91, "epoch_pct": 11.91, "eta": "59:57:14", "max_grad_norm": 0.8, "loss": 0.7052890658378601, "grad_norm": 1.2415164709091187, "learning_rate": 9.95579038763142e-05}
+{"ts": "2025-12-23T22:29:54", "event": "train_log", "step": 1658, "epoch": 0.7154261057173679, "progress_pct": 11.92, "epoch_pct": 11.92, "eta": "59:54:40", "max_grad_norm": 0.8, "loss": 0.8300023674964905, "grad_norm": 1.3983269929885864, "learning_rate": 9.955464001570992e-05}
+{"ts": "2025-12-23T22:30:14", "event": "train_log", "step": 1660, "epoch": 0.7162891046386192, "progress_pct": 11.94, "epoch_pct": 11.94, "eta": "59:52:11", "max_grad_norm": 0.8, "loss": 0.7742894887924194, "grad_norm": 1.1591373682022095, "learning_rate": 9.955136420522759e-05}
+{"ts": "2025-12-23T22:30:32", "event": "train_log", "step": 1662, "epoch": 0.7171521035598706, "progress_pct": 11.95, "epoch_pct": 11.95, "eta": "59:49:31", "max_grad_norm": 0.8, "loss": 0.8458141684532166, "grad_norm": 1.2503128051757812, "learning_rate": 9.954807644565715e-05}
+{"ts": "2025-12-23T22:30:52", "event": "train_log", "step": 1664, "epoch": 0.7180151024811219, "progress_pct": 11.96, "epoch_pct": 11.97, "eta": "59:47:04", "max_grad_norm": 0.8, "loss": 0.7044742107391357, "grad_norm": 1.2954912185668945, "learning_rate": 9.954477673779146e-05}
+{"ts": "2025-12-23T22:31:13", "event": "train_log", "step": 1666, "epoch": 0.7188781014023733, "progress_pct": 11.98, "epoch_pct": 11.98, "eta": "59:44:43", "max_grad_norm": 0.8, "loss": 0.7917183041572571, "grad_norm": 1.1669788360595703, "learning_rate": 9.954146508242619e-05}
+{"ts": "2025-12-23T22:31:34", "event": "train_log", "step": 1668, "epoch": 0.7197411003236246, "progress_pct": 11.99, "epoch_pct": 12.0, "eta": "59:42:26", "max_grad_norm": 0.8, "loss": 0.7174562811851501, "grad_norm": 1.1466871500015259, "learning_rate": 9.953814148035996e-05}
+{"ts": "2025-12-23T22:31:52", "event": "train_log", "step": 1670, "epoch": 0.7206040992448759, "progress_pct": 12.01, "epoch_pct": 12.01, "eta": "59:39:46", "max_grad_norm": 0.8, "loss": 0.8925880193710327, "grad_norm": 1.367224097251892, "learning_rate": 9.953480593239421e-05}
+{"ts": "2025-12-23T22:32:13", "event": "train_log", "step": 1672, "epoch": 0.7214670981661273, "progress_pct": 12.02, "epoch_pct": 12.02, "eta": "59:37:28", "max_grad_norm": 0.8, "loss": 0.70949387550354, "grad_norm": 1.049497365951538, "learning_rate": 9.953145843933334e-05}
+{"ts": "2025-12-23T22:32:35", "event": "train_log", "step": 1674, "epoch": 0.7223300970873786, "progress_pct": 12.04, "epoch_pct": 12.04, "eta": "59:35:16", "max_grad_norm": 0.8, "loss": 0.6731935143470764, "grad_norm": 0.9995084404945374, "learning_rate": 9.952809900198455e-05}
+{"ts": "2025-12-23T22:32:56", "event": "train_log", "step": 1676, "epoch": 0.72319309600863, "progress_pct": 12.05, "epoch_pct": 12.05, "eta": "59:32:55", "max_grad_norm": 0.8, "loss": 0.7316275835037231, "grad_norm": 1.100899577140808, "learning_rate": 9.952472762115794e-05}
+{"ts": "2025-12-23T22:33:14", "event": "train_log", "step": 1678, "epoch": 0.7240560949298813, "progress_pct": 12.06, "epoch_pct": 12.07, "eta": "59:30:22", "max_grad_norm": 0.8, "loss": 0.7379520535469055, "grad_norm": 1.2594135999679565, "learning_rate": 9.952134429766655e-05}
+{"ts": "2025-12-23T22:33:33", "event": "train_log", "step": 1680, "epoch": 0.7249190938511327, "progress_pct": 12.08, "epoch_pct": 12.08, "eta": "59:27:48", "max_grad_norm": 0.8, "loss": 0.7072247266769409, "grad_norm": 1.3124911785125732, "learning_rate": 9.951794903232621e-05}
+{"ts": "2025-12-23T22:33:52", "event": "train_log", "step": 1682, "epoch": 0.725782092772384, "progress_pct": 12.09, "epoch_pct": 12.1, "eta": "59:25:19", "max_grad_norm": 0.8, "loss": 0.745680034160614, "grad_norm": 1.212144374847412, "learning_rate": 9.95145418259557e-05}
+{"ts": "2025-12-23T22:34:10", "event": "train_log", "step": 1684, "epoch": 0.7266450916936353, "progress_pct": 12.11, "epoch_pct": 12.11, "eta": "59:22:38", "max_grad_norm": 0.8, "loss": 0.796852707862854, "grad_norm": 1.3913586139678955, "learning_rate": 9.951112267937663e-05}
+{"ts": "2025-12-23T22:34:30", "event": "train_log", "step": 1686, "epoch": 0.7275080906148867, "progress_pct": 12.12, "epoch_pct": 12.13, "eta": "59:20:14", "max_grad_norm": 0.8, "loss": 0.700892984867096, "grad_norm": 1.3227289915084839, "learning_rate": 9.950769159341353e-05}
+{"ts": "2025-12-23T22:34:48", "event": "train_log", "step": 1688, "epoch": 0.728371089536138, "progress_pct": 12.14, "epoch_pct": 12.14, "eta": "59:17:35", "max_grad_norm": 0.8, "loss": 0.7973095178604126, "grad_norm": 1.4656381607055664, "learning_rate": 9.950424856889378e-05}
+{"ts": "2025-12-23T22:35:08", "event": "train_log", "step": 1690, "epoch": 0.7292340884573895, "progress_pct": 12.15, "epoch_pct": 12.15, "eta": "59:15:14", "max_grad_norm": 0.8, "loss": 0.7811378836631775, "grad_norm": 1.1698973178863525, "learning_rate": 9.950079360664766e-05}
+{"ts": "2025-12-23T22:35:26", "event": "train_log", "step": 1692, "epoch": 0.7300970873786408, "progress_pct": 12.17, "epoch_pct": 12.17, "eta": "59:12:38", "max_grad_norm": 0.8, "loss": 0.753012478351593, "grad_norm": 1.201695203781128, "learning_rate": 9.949732670750831e-05}
+{"ts": "2025-12-23T22:35:44", "event": "train_log", "step": 1694, "epoch": 0.7309600862998922, "progress_pct": 12.18, "epoch_pct": 12.18, "eta": "59:10:00", "max_grad_norm": 0.8, "loss": 0.7620642781257629, "grad_norm": 1.3465325832366943, "learning_rate": 9.949384787231175e-05}
+{"ts": "2025-12-23T22:36:03", "event": "train_log", "step": 1696, "epoch": 0.7318230852211435, "progress_pct": 12.19, "epoch_pct": 12.2, "eta": "59:07:28", "max_grad_norm": 0.8, "loss": 0.7479873299598694, "grad_norm": 1.2191137075424194, "learning_rate": 9.94903571018969e-05}
+{"ts": "2025-12-23T22:36:22", "event": "train_log", "step": 1698, "epoch": 0.7326860841423948, "progress_pct": 12.21, "epoch_pct": 12.21, "eta": "59:05:03", "max_grad_norm": 0.8, "loss": 0.7327720522880554, "grad_norm": 1.1412187814712524, "learning_rate": 9.948685439710554e-05}
+{"ts": "2025-12-23T22:36:41", "event": "train_log", "step": 1700, "epoch": 0.7335490830636462, "progress_pct": 12.22, "epoch_pct": 12.23, "eta": "59:02:36", "max_grad_norm": 0.8, "loss": 0.761475682258606, "grad_norm": 1.310327410697937, "learning_rate": 9.948333975878233e-05}
+{"ts": "2025-12-23T22:50:30", "event": "train_log", "step": 1700, "epoch": 0.7335490830636462, "progress_pct": 12.22, "epoch_pct": 12.23, "eta": "60:41:44", "max_grad_norm": 0.8, "eval_loss": 0.7649155259132385, "eval_runtime": 828.3542, "eval_samples_per_second": 2.487, "eval_steps_per_second": 2.487}
+{"ts": "2025-12-23T22:50:48", "event": "train_log", "step": 1702, "epoch": 0.7344120819848975, "progress_pct": 12.24, "epoch_pct": 12.24, "eta": "60:39:04", "max_grad_norm": 0.8, "loss": 0.7564467191696167, "grad_norm": 1.3031405210494995, "learning_rate": 9.947981318777482e-05}
+{"ts": "2025-12-23T22:51:07", "event": "train_log", "step": 1704, "epoch": 0.7352750809061489, "progress_pct": 12.25, "epoch_pct": 12.25, "eta": "60:36:26", "max_grad_norm": 0.8, "loss": 0.7689576745033264, "grad_norm": 1.2672377824783325, "learning_rate": 9.94762746849334e-05}
+{"ts": "2025-12-23T22:51:26", "event": "train_log", "step": 1706, "epoch": 0.7361380798274002, "progress_pct": 12.27, "epoch_pct": 12.27, "eta": "60:33:53", "max_grad_norm": 0.8, "loss": 0.7923810482025146, "grad_norm": 1.1159595251083374, "learning_rate": 9.947272425111138e-05}
+{"ts": "2025-12-23T22:51:47", "event": "train_log", "step": 1708, "epoch": 0.7370010787486516, "progress_pct": 12.28, "epoch_pct": 12.28, "eta": "60:31:28", "max_grad_norm": 0.8, "loss": 0.6540790796279907, "grad_norm": 1.1190921068191528, "learning_rate": 9.946916188716495e-05}
+{"ts": "2025-12-23T22:52:07", "event": "train_log", "step": 1710, "epoch": 0.7378640776699029, "progress_pct": 12.3, "epoch_pct": 12.3, "eta": "60:28:59", "max_grad_norm": 0.8, "loss": 0.7267473936080933, "grad_norm": 1.373832106590271, "learning_rate": 9.946558759395312e-05}
+{"ts": "2025-12-23T22:52:26", "event": "train_log", "step": 1712, "epoch": 0.7387270765911542, "progress_pct": 12.31, "epoch_pct": 12.31, "eta": "60:26:31", "max_grad_norm": 0.8, "loss": 0.7693060636520386, "grad_norm": 1.2346374988555908, "learning_rate": 9.946200137233785e-05}
+{"ts": "2025-12-23T22:52:46", "event": "train_log", "step": 1714, "epoch": 0.7395900755124056, "progress_pct": 12.32, "epoch_pct": 12.33, "eta": "60:23:58", "max_grad_norm": 0.8, "loss": 0.7751407027244568, "grad_norm": 1.1629668474197388, "learning_rate": 9.945840322318391e-05}
+{"ts": "2025-12-23T22:53:05", "event": "train_log", "step": 1716, "epoch": 0.7404530744336569, "progress_pct": 12.34, "epoch_pct": 12.34, "eta": "60:21:28", "max_grad_norm": 0.8, "loss": 0.7798545360565186, "grad_norm": 1.2268708944320679, "learning_rate": 9.945479314735899e-05}
+{"ts": "2025-12-23T22:53:27", "event": "train_log", "step": 1718, "epoch": 0.7413160733549083, "progress_pct": 12.35, "epoch_pct": 12.36, "eta": "60:19:15", "max_grad_norm": 0.8, "loss": 0.6128867268562317, "grad_norm": 0.9968738555908203, "learning_rate": 9.945117114573366e-05}
+{"ts": "2025-12-23T22:53:47", "event": "train_log", "step": 1720, "epoch": 0.7421790722761596, "progress_pct": 12.37, "epoch_pct": 12.37, "eta": "60:16:51", "max_grad_norm": 0.8, "loss": 0.7536954283714294, "grad_norm": 1.2135992050170898, "learning_rate": 9.944753721918133e-05}
+{"ts": "2025-12-23T22:54:07", "event": "train_log", "step": 1722, "epoch": 0.7430420711974111, "progress_pct": 12.38, "epoch_pct": 12.38, "eta": "60:14:19", "max_grad_norm": 0.8, "loss": 0.7034479379653931, "grad_norm": 1.1361218690872192, "learning_rate": 9.94438913685783e-05}
+{"ts": "2025-12-23T22:54:28", "event": "train_log", "step": 1724, "epoch": 0.7439050701186624, "progress_pct": 12.4, "epoch_pct": 12.4, "eta": "60:12:04", "max_grad_norm": 0.8, "loss": 0.7285382747650146, "grad_norm": 1.0728967189788818, "learning_rate": 9.944023359480378e-05}
+{"ts": "2025-12-23T22:54:47", "event": "train_log", "step": 1726, "epoch": 0.7447680690399137, "progress_pct": 12.41, "epoch_pct": 12.41, "eta": "60:09:30", "max_grad_norm": 0.8, "loss": 0.7865195274353027, "grad_norm": 1.3613945245742798, "learning_rate": 9.94365638987398e-05}
+{"ts": "2025-12-23T22:55:07", "event": "train_log", "step": 1728, "epoch": 0.7456310679611651, "progress_pct": 12.42, "epoch_pct": 12.43, "eta": "60:07:03", "max_grad_norm": 0.8, "loss": 0.7511224746704102, "grad_norm": 1.0985898971557617, "learning_rate": 9.94328822812713e-05}
+{"ts": "2025-12-23T22:55:27", "event": "train_log", "step": 1730, "epoch": 0.7464940668824164, "progress_pct": 12.44, "epoch_pct": 12.44, "eta": "60:04:37", "max_grad_norm": 0.8, "loss": 0.732453465461731, "grad_norm": 1.1506208181381226, "learning_rate": 9.942918874328606e-05}
+{"ts": "2025-12-23T22:55:46", "event": "train_log", "step": 1732, "epoch": 0.7473570658036678, "progress_pct": 12.45, "epoch_pct": 12.46, "eta": "60:02:11", "max_grad_norm": 0.8, "loss": 0.702409029006958, "grad_norm": 1.0995100736618042, "learning_rate": 9.94254832856748e-05}
+{"ts": "2025-12-23T22:56:07", "event": "train_log", "step": 1734, "epoch": 0.7482200647249191, "progress_pct": 12.47, "epoch_pct": 12.47, "eta": "59:59:53", "max_grad_norm": 0.8, "loss": 0.7515703439712524, "grad_norm": 1.0075604915618896, "learning_rate": 9.942176590933106e-05}
+{"ts": "2025-12-23T22:56:27", "event": "train_log", "step": 1736, "epoch": 0.7490830636461704, "progress_pct": 12.48, "epoch_pct": 12.48, "eta": "59:57:27", "max_grad_norm": 0.8, "loss": 0.6910237073898315, "grad_norm": 1.3360812664031982, "learning_rate": 9.941803661515126e-05}
+{"ts": "2025-12-23T22:56:47", "event": "train_log", "step": 1738, "epoch": 0.7499460625674218, "progress_pct": 12.5, "epoch_pct": 12.5, "eta": "59:55:05", "max_grad_norm": 0.8, "loss": 0.7185606956481934, "grad_norm": 1.0172319412231445, "learning_rate": 9.941429540403471e-05}
+{"ts": "2025-12-23T22:57:07", "event": "train_log", "step": 1740, "epoch": 0.7508090614886731, "progress_pct": 12.51, "epoch_pct": 12.51, "eta": "59:52:41", "max_grad_norm": 0.8, "loss": 0.7136536836624146, "grad_norm": 1.2721658945083618, "learning_rate": 9.941054227688358e-05}
+{"ts": "2025-12-23T22:57:27", "event": "train_log", "step": 1742, "epoch": 0.7516720604099245, "progress_pct": 12.53, "epoch_pct": 12.53, "eta": "59:50:14", "max_grad_norm": 0.8, "loss": 0.7595619559288025, "grad_norm": 1.0775413513183594, "learning_rate": 9.940677723460292e-05}
+{"ts": "2025-12-23T22:57:45", "event": "train_log", "step": 1744, "epoch": 0.7525350593311758, "progress_pct": 12.54, "epoch_pct": 12.54, "eta": "59:47:42", "max_grad_norm": 0.8, "loss": 0.7602618336677551, "grad_norm": 1.3436479568481445, "learning_rate": 9.940300027810067e-05}
+{"ts": "2025-12-23T22:58:04", "event": "train_log", "step": 1746, "epoch": 0.7533980582524272, "progress_pct": 12.55, "epoch_pct": 12.56, "eta": "59:45:07", "max_grad_norm": 0.8, "loss": 0.7954604029655457, "grad_norm": 1.2611958980560303, "learning_rate": 9.93992114082876e-05}
+{"ts": "2025-12-23T22:58:22", "event": "train_log", "step": 1748, "epoch": 0.7542610571736785, "progress_pct": 12.57, "epoch_pct": 12.57, "eta": "59:42:36", "max_grad_norm": 0.8, "loss": 0.7661190629005432, "grad_norm": 1.2794841527938843, "learning_rate": 9.93954106260774e-05}
+{"ts": "2025-12-23T22:58:42", "event": "train_log", "step": 1750, "epoch": 0.7551240560949298, "progress_pct": 12.58, "epoch_pct": 12.59, "eta": "59:40:10", "max_grad_norm": 0.8, "loss": 0.6855312585830688, "grad_norm": 1.3042595386505127, "learning_rate": 9.939159793238661e-05}
+{"ts": "2025-12-23T22:59:01", "event": "train_log", "step": 1752, "epoch": 0.7559870550161812, "progress_pct": 12.6, "epoch_pct": 12.6, "eta": "59:37:41", "max_grad_norm": 0.8, "loss": 0.7970905900001526, "grad_norm": 1.2799923419952393, "learning_rate": 9.938777332813463e-05}
+{"ts": "2025-12-23T22:59:21", "event": "train_log", "step": 1754, "epoch": 0.7568500539374325, "progress_pct": 12.61, "epoch_pct": 12.61, "eta": "59:35:20", "max_grad_norm": 0.8, "loss": 0.7072843313217163, "grad_norm": 1.169205904006958, "learning_rate": 9.938393681424375e-05}
+{"ts": "2025-12-23T22:59:40", "event": "train_log", "step": 1756, "epoch": 0.757713052858684, "progress_pct": 12.63, "epoch_pct": 12.63, "eta": "59:32:51", "max_grad_norm": 0.8, "loss": 0.7134629487991333, "grad_norm": 1.3833354711532593, "learning_rate": 9.938008839163913e-05}
+{"ts": "2025-12-23T22:59:58", "event": "train_log", "step": 1758, "epoch": 0.7585760517799353, "progress_pct": 12.64, "epoch_pct": 12.64, "eta": "59:30:22", "max_grad_norm": 0.8, "loss": 0.7873107194900513, "grad_norm": 1.223452091217041, "learning_rate": 9.93762280612488e-05}
+{"ts": "2025-12-23T23:00:18", "event": "train_log", "step": 1760, "epoch": 0.7594390507011867, "progress_pct": 12.65, "epoch_pct": 12.66, "eta": "59:28:01", "max_grad_norm": 0.8, "loss": 0.7680654525756836, "grad_norm": 1.2740365266799927, "learning_rate": 9.937235582400367e-05}
+{"ts": "2025-12-23T23:00:37", "event": "train_log", "step": 1762, "epoch": 0.760302049622438, "progress_pct": 12.67, "epoch_pct": 12.67, "eta": "59:25:32", "max_grad_norm": 0.8, "loss": 0.734126627445221, "grad_norm": 1.2564697265625, "learning_rate": 9.93684716808375e-05}
+{"ts": "2025-12-23T23:00:55", "event": "train_log", "step": 1764, "epoch": 0.7611650485436893, "progress_pct": 12.68, "epoch_pct": 12.69, "eta": "59:22:57", "max_grad_norm": 0.8, "loss": 0.7972320914268494, "grad_norm": 1.3858847618103027, "learning_rate": 9.936457563268694e-05}
+{"ts": "2025-12-23T23:01:14", "event": "train_log", "step": 1766, "epoch": 0.7620280474649407, "progress_pct": 12.7, "epoch_pct": 12.7, "eta": "59:20:30", "max_grad_norm": 0.8, "loss": 0.7904276847839355, "grad_norm": 1.2705003023147583, "learning_rate": 9.93606676804915e-05}
+{"ts": "2025-12-23T23:01:32", "event": "train_log", "step": 1768, "epoch": 0.762891046386192, "progress_pct": 12.71, "epoch_pct": 12.71, "eta": "59:18:00", "max_grad_norm": 0.8, "loss": 0.7542425990104675, "grad_norm": 1.210766077041626, "learning_rate": 9.935674782519358e-05}
+{"ts": "2025-12-23T23:01:51", "event": "train_log", "step": 1770, "epoch": 0.7637540453074434, "progress_pct": 12.73, "epoch_pct": 12.73, "eta": "59:15:30", "max_grad_norm": 0.8, "loss": 0.7975004315376282, "grad_norm": 1.4367541074752808, "learning_rate": 9.93528160677384e-05}
+{"ts": "2025-12-23T23:02:10", "event": "train_log", "step": 1772, "epoch": 0.7646170442286947, "progress_pct": 12.74, "epoch_pct": 12.74, "eta": "59:13:08", "max_grad_norm": 0.8, "loss": 0.7537807822227478, "grad_norm": 1.0994079113006592, "learning_rate": 9.934887240907414e-05}
+{"ts": "2025-12-23T23:02:30", "event": "train_log", "step": 1774, "epoch": 0.7654800431499461, "progress_pct": 12.76, "epoch_pct": 12.76, "eta": "59:10:46", "max_grad_norm": 0.8, "loss": 0.7333855628967285, "grad_norm": 1.111660122871399, "learning_rate": 9.934491685015174e-05}
+{"ts": "2025-12-23T23:02:49", "event": "train_log", "step": 1776, "epoch": 0.7663430420711974, "progress_pct": 12.77, "epoch_pct": 12.77, "eta": "59:08:20", "max_grad_norm": 0.8, "loss": 0.7472840547561646, "grad_norm": 1.2879043817520142, "learning_rate": 9.934094939192508e-05}
+{"ts": "2025-12-23T23:03:07", "event": "train_log", "step": 1778, "epoch": 0.7672060409924487, "progress_pct": 12.78, "epoch_pct": 12.79, "eta": "59:05:52", "max_grad_norm": 0.8, "loss": 0.7644904255867004, "grad_norm": 1.3618478775024414, "learning_rate": 9.933697003535094e-05}
+{"ts": "2025-12-23T23:03:28", "event": "train_log", "step": 1780, "epoch": 0.7680690399137001, "progress_pct": 12.8, "epoch_pct": 12.8, "eta": "59:03:36", "max_grad_norm": 0.8, "loss": 0.7872052192687988, "grad_norm": 1.1235666275024414, "learning_rate": 9.933297878138887e-05}
+{"ts": "2025-12-23T23:03:45", "event": "train_log", "step": 1782, "epoch": 0.7689320388349514, "progress_pct": 12.81, "epoch_pct": 12.82, "eta": "59:01:03", "max_grad_norm": 0.8, "loss": 0.7343124747276306, "grad_norm": 1.3741775751113892, "learning_rate": 9.932897563100136e-05}
+{"ts": "2025-12-23T23:04:04", "event": "train_log", "step": 1784, "epoch": 0.7697950377562028, "progress_pct": 12.83, "epoch_pct": 12.83, "eta": "58:58:34", "max_grad_norm": 0.8, "loss": 0.7486474514007568, "grad_norm": 1.2115039825439453, "learning_rate": 9.932496058515375e-05}
+{"ts": "2025-12-23T23:04:22", "event": "train_log", "step": 1786, "epoch": 0.7706580366774541, "progress_pct": 12.84, "epoch_pct": 12.84, "eta": "58:56:06", "max_grad_norm": 0.8, "loss": 0.7333025932312012, "grad_norm": 1.183457851409912, "learning_rate": 9.932093364481426e-05}
+{"ts": "2025-12-23T23:04:41", "event": "train_log", "step": 1788, "epoch": 0.7715210355987056, "progress_pct": 12.86, "epoch_pct": 12.86, "eta": "58:53:42", "max_grad_norm": 0.8, "loss": 0.7241485118865967, "grad_norm": 1.2841869592666626, "learning_rate": 9.931689481095394e-05}
+{"ts": "2025-12-23T23:04:59", "event": "train_log", "step": 1790, "epoch": 0.7723840345199569, "progress_pct": 12.87, "epoch_pct": 12.87, "eta": "58:51:15", "max_grad_norm": 0.8, "loss": 0.7510396242141724, "grad_norm": 1.2312977313995361, "learning_rate": 9.931284408454678e-05}
+{"ts": "2025-12-23T23:05:19", "event": "train_log", "step": 1792, "epoch": 0.7732470334412082, "progress_pct": 12.88, "epoch_pct": 12.89, "eta": "58:48:54", "max_grad_norm": 0.8, "loss": 0.7108346819877625, "grad_norm": 1.144895315170288, "learning_rate": 9.930878146656956e-05}
+{"ts": "2025-12-23T23:05:37", "event": "train_log", "step": 1794, "epoch": 0.7741100323624596, "progress_pct": 12.9, "epoch_pct": 12.9, "eta": "58:46:28", "max_grad_norm": 0.8, "loss": 0.7687607407569885, "grad_norm": 1.18603515625, "learning_rate": 9.9304706958002e-05}
+{"ts": "2025-12-23T23:05:55", "event": "train_log", "step": 1796, "epoch": 0.7749730312837109, "progress_pct": 12.91, "epoch_pct": 12.92, "eta": "58:44:01", "max_grad_norm": 0.8, "loss": 0.7826682329177856, "grad_norm": 1.2827935218811035, "learning_rate": 9.93006205598266e-05}
+{"ts": "2025-12-23T23:06:14", "event": "train_log", "step": 1798, "epoch": 0.7758360302049623, "progress_pct": 12.93, "epoch_pct": 12.93, "eta": "58:41:35", "max_grad_norm": 0.8, "loss": 0.6970428824424744, "grad_norm": 1.164852261543274, "learning_rate": 9.929652227302881e-05}
+{"ts": "2025-12-23T23:06:33", "event": "train_log", "step": 1800, "epoch": 0.7766990291262136, "progress_pct": 12.94, "epoch_pct": 12.94, "eta": "58:39:15", "max_grad_norm": 0.8, "loss": 0.7238826751708984, "grad_norm": 1.0439956188201904, "learning_rate": 9.92924120985969e-05}
+{"ts": "2025-12-23T23:20:32", "event": "train_log", "step": 1800, "epoch": 0.7766990291262136, "progress_pct": 12.94, "epoch_pct": 12.94, "eta": "60:13:18", "max_grad_norm": 0.8, "eval_loss": 0.7598537802696228, "eval_runtime": 838.9645, "eval_samples_per_second": 2.455, "eval_steps_per_second": 2.455}
+{"ts": "2025-12-23T23:20:50", "event": "train_log", "step": 1802, "epoch": 0.777562028047465, "progress_pct": 12.96, "epoch_pct": 12.96, "eta": "60:10:45", "max_grad_norm": 0.8, "loss": 0.7723641395568848, "grad_norm": 1.2464938163757324, "learning_rate": 9.928829003752202e-05}
+{"ts": "2025-12-23T23:21:10", "event": "train_log", "step": 1804, "epoch": 0.7784250269687163, "progress_pct": 12.97, "epoch_pct": 12.97, "eta": "60:08:21", "max_grad_norm": 0.8, "loss": 0.7154937982559204, "grad_norm": 1.2445050477981567, "learning_rate": 9.92841560907982e-05}
+{"ts": "2025-12-23T23:21:29", "event": "train_log", "step": 1806, "epoch": 0.7792880258899676, "progress_pct": 12.99, "epoch_pct": 12.99, "eta": "60:05:50", "max_grad_norm": 0.8, "loss": 0.8325800895690918, "grad_norm": 1.2235661745071411, "learning_rate": 9.928001025942231e-05}
+{"ts": "2025-12-23T23:21:48", "event": "train_log", "step": 1808, "epoch": 0.780151024811219, "progress_pct": 13.0, "epoch_pct": 13.0, "eta": "60:03:23", "max_grad_norm": 0.8, "loss": 0.7093122601509094, "grad_norm": 1.268066644668579, "learning_rate": 9.927585254439409e-05}
+{"ts": "2025-12-23T23:22:07", "event": "train_log", "step": 1810, "epoch": 0.7810140237324703, "progress_pct": 13.01, "epoch_pct": 13.02, "eta": "60:00:56", "max_grad_norm": 0.8, "loss": 0.7601775527000427, "grad_norm": 1.1731390953063965, "learning_rate": 9.927168294671618e-05}
+{"ts": "2025-12-23T23:22:26", "event": "train_log", "step": 1812, "epoch": 0.7818770226537217, "progress_pct": 13.03, "epoch_pct": 13.03, "eta": "59:58:29", "max_grad_norm": 0.8, "loss": 0.6889236569404602, "grad_norm": 1.0685608386993408, "learning_rate": 9.926750146739403e-05}
+{"ts": "2025-12-23T23:22:45", "event": "train_log", "step": 1814, "epoch": 0.782740021574973, "progress_pct": 13.04, "epoch_pct": 13.05, "eta": "59:56:03", "max_grad_norm": 0.8, "loss": 0.7397196888923645, "grad_norm": 1.1098639965057373, "learning_rate": 9.926330810743602e-05}
+{"ts": "2025-12-23T23:23:04", "event": "train_log", "step": 1816, "epoch": 0.7836030204962244, "progress_pct": 13.06, "epoch_pct": 13.06, "eta": "59:53:36", "max_grad_norm": 0.8, "loss": 0.7180029153823853, "grad_norm": 1.3606436252593994, "learning_rate": 9.925910286785332e-05}
+{"ts": "2025-12-23T23:23:23", "event": "train_log", "step": 1818, "epoch": 0.7844660194174757, "progress_pct": 13.07, "epoch_pct": 13.07, "eta": "59:51:09", "max_grad_norm": 0.8, "loss": 0.7039036750793457, "grad_norm": 1.143418550491333, "learning_rate": 9.925488574966004e-05}
+{"ts": "2025-12-23T23:23:42", "event": "train_log", "step": 1820, "epoch": 0.785329018338727, "progress_pct": 13.09, "epoch_pct": 13.09, "eta": "59:48:45", "max_grad_norm": 0.8, "loss": 0.741872251033783, "grad_norm": 1.180968999862671, "learning_rate": 9.92506567538731e-05}
+{"ts": "2025-12-23T23:24:01", "event": "train_log", "step": 1822, "epoch": 0.7861920172599784, "progress_pct": 13.1, "epoch_pct": 13.1, "eta": "59:46:20", "max_grad_norm": 0.8, "loss": 0.7580198049545288, "grad_norm": 1.1953548192977905, "learning_rate": 9.924641588151231e-05}
+{"ts": "2025-12-23T23:24:19", "event": "train_log", "step": 1824, "epoch": 0.7870550161812297, "progress_pct": 13.11, "epoch_pct": 13.12, "eta": "59:43:47", "max_grad_norm": 0.8, "loss": 0.7479216456413269, "grad_norm": 1.3049067258834839, "learning_rate": 9.924216313360033e-05}
+{"ts": "2025-12-23T23:24:37", "event": "train_log", "step": 1826, "epoch": 0.7879180151024812, "progress_pct": 13.13, "epoch_pct": 13.13, "eta": "59:41:13", "max_grad_norm": 0.8, "loss": 0.7382357716560364, "grad_norm": 1.3962669372558594, "learning_rate": 9.92378985111627e-05}
+{"ts": "2025-12-23T23:24:56", "event": "train_log", "step": 1828, "epoch": 0.7887810140237325, "progress_pct": 13.14, "epoch_pct": 13.15, "eta": "59:38:49", "max_grad_norm": 0.8, "loss": 0.6356467604637146, "grad_norm": 1.1519217491149902, "learning_rate": 9.92336220152278e-05}
+{"ts": "2025-12-23T23:25:15", "event": "train_log", "step": 1830, "epoch": 0.7896440129449838, "progress_pct": 13.16, "epoch_pct": 13.16, "eta": "59:36:26", "max_grad_norm": 0.8, "loss": 0.7674193978309631, "grad_norm": 1.2544089555740356, "learning_rate": 9.922933364682689e-05}
+{"ts": "2025-12-23T23:25:34", "event": "train_log", "step": 1832, "epoch": 0.7905070118662352, "progress_pct": 13.17, "epoch_pct": 13.18, "eta": "59:34:01", "max_grad_norm": 0.8, "loss": 0.7083898782730103, "grad_norm": 1.1358295679092407, "learning_rate": 9.922503340699412e-05}
+{"ts": "2025-12-23T23:25:54", "event": "train_log", "step": 1834, "epoch": 0.7913700107874865, "progress_pct": 13.19, "epoch_pct": 13.19, "eta": "59:31:40", "max_grad_norm": 0.8, "loss": 0.6273066401481628, "grad_norm": 1.1744481325149536, "learning_rate": 9.922072129676644e-05}
+{"ts": "2025-12-23T23:26:12", "event": "train_log", "step": 1836, "epoch": 0.7922330097087379, "progress_pct": 13.2, "epoch_pct": 13.2, "eta": "59:29:08", "max_grad_norm": 0.8, "loss": 0.808181643486023, "grad_norm": 1.20509672164917, "learning_rate": 9.92163973171837e-05}
+{"ts": "2025-12-23T23:26:31", "event": "train_log", "step": 1838, "epoch": 0.7930960086299892, "progress_pct": 13.22, "epoch_pct": 13.22, "eta": "59:26:45", "max_grad_norm": 0.8, "loss": 0.6631553769111633, "grad_norm": 1.2035597562789917, "learning_rate": 9.921206146928863e-05}
+{"ts": "2025-12-23T23:26:52", "event": "train_log", "step": 1840, "epoch": 0.7939590075512406, "progress_pct": 13.23, "epoch_pct": 13.23, "eta": "59:24:35", "max_grad_norm": 0.8, "loss": 0.72016841173172, "grad_norm": 1.058695912361145, "learning_rate": 9.920771375412678e-05}
+{"ts": "2025-12-23T23:27:12", "event": "train_log", "step": 1842, "epoch": 0.7948220064724919, "progress_pct": 13.24, "epoch_pct": 13.25, "eta": "59:22:22", "max_grad_norm": 0.8, "loss": 0.7240060567855835, "grad_norm": 1.035791039466858, "learning_rate": 9.92033541727466e-05}
+{"ts": "2025-12-23T23:27:31", "event": "train_log", "step": 1844, "epoch": 0.7956850053937432, "progress_pct": 13.26, "epoch_pct": 13.26, "eta": "59:19:58", "max_grad_norm": 0.8, "loss": 0.743224561214447, "grad_norm": 1.1845753192901611, "learning_rate": 9.919898272619935e-05}
+{"ts": "2025-12-23T23:27:50", "event": "train_log", "step": 1846, "epoch": 0.7965480043149946, "progress_pct": 13.27, "epoch_pct": 13.28, "eta": "59:17:32", "max_grad_norm": 0.8, "loss": 0.7064419388771057, "grad_norm": 1.2942636013031006, "learning_rate": 9.919459941553922e-05}
+{"ts": "2025-12-23T23:28:08", "event": "train_log", "step": 1848, "epoch": 0.7974110032362459, "progress_pct": 13.29, "epoch_pct": 13.29, "eta": "59:15:08", "max_grad_norm": 0.8, "loss": 0.805932343006134, "grad_norm": 1.2736715078353882, "learning_rate": 9.91902042418232e-05}
+{"ts": "2025-12-23T23:28:27", "event": "train_log", "step": 1850, "epoch": 0.7982740021574973, "progress_pct": 13.3, "epoch_pct": 13.3, "eta": "59:12:43", "max_grad_norm": 0.8, "loss": 0.8050802946090698, "grad_norm": 1.2803614139556885, "learning_rate": 9.918579720611119e-05}
+{"ts": "2025-12-23T23:28:46", "event": "train_log", "step": 1852, "epoch": 0.7991370010787486, "progress_pct": 13.32, "epoch_pct": 13.32, "eta": "59:10:23", "max_grad_norm": 0.8, "loss": 0.7013180255889893, "grad_norm": 1.2225072383880615, "learning_rate": 9.91813783094659e-05}
+{"ts": "2025-12-23T23:29:05", "event": "train_log", "step": 1854, "epoch": 0.8, "progress_pct": 13.33, "epoch_pct": 13.33, "eta": "59:08:01", "max_grad_norm": 0.8, "loss": 0.7415283918380737, "grad_norm": 1.3375571966171265, "learning_rate": 9.917694755295294e-05}
+{"ts": "2025-12-23T23:29:24", "event": "train_log", "step": 1856, "epoch": 0.8008629989212513, "progress_pct": 13.34, "epoch_pct": 13.35, "eta": "59:05:40", "max_grad_norm": 0.8, "loss": 0.7264692783355713, "grad_norm": 1.2209177017211914, "learning_rate": 9.917250493764079e-05}
+{"ts": "2025-12-23T23:29:45", "event": "train_log", "step": 1858, "epoch": 0.8017259978425026, "progress_pct": 13.36, "epoch_pct": 13.36, "eta": "59:03:28", "max_grad_norm": 0.8, "loss": 0.7083719372749329, "grad_norm": 1.1158864498138428, "learning_rate": 9.916805046460074e-05}
+{"ts": "2025-12-23T23:30:03", "event": "train_log", "step": 1860, "epoch": 0.8025889967637541, "progress_pct": 13.37, "epoch_pct": 13.38, "eta": "59:01:04", "max_grad_norm": 0.8, "loss": 0.820396900177002, "grad_norm": 1.171000361442566, "learning_rate": 9.916358413490696e-05}
+{"ts": "2025-12-23T23:30:23", "event": "train_log", "step": 1862, "epoch": 0.8034519956850054, "progress_pct": 13.39, "epoch_pct": 13.39, "eta": "58:58:51", "max_grad_norm": 0.8, "loss": 0.6754014492034912, "grad_norm": 0.9942941069602966, "learning_rate": 9.91591059496365e-05}
+{"ts": "2025-12-23T23:30:43", "event": "train_log", "step": 1864, "epoch": 0.8043149946062568, "progress_pct": 13.4, "epoch_pct": 13.41, "eta": "58:56:33", "max_grad_norm": 0.8, "loss": 0.813154399394989, "grad_norm": 1.239795207977295, "learning_rate": 9.915461590986926e-05}
+{"ts": "2025-12-23T23:31:02", "event": "train_log", "step": 1866, "epoch": 0.8051779935275081, "progress_pct": 13.42, "epoch_pct": 13.42, "eta": "58:54:18", "max_grad_norm": 0.8, "loss": 0.745580792427063, "grad_norm": 1.076019525527954, "learning_rate": 9.915011401668796e-05}
+{"ts": "2025-12-23T23:31:21", "event": "train_log", "step": 1868, "epoch": 0.8060409924487595, "progress_pct": 13.43, "epoch_pct": 13.43, "eta": "58:51:56", "max_grad_norm": 0.8, "loss": 0.7564949989318848, "grad_norm": 1.2134130001068115, "learning_rate": 9.914560027117826e-05}
+{"ts": "2025-12-23T23:31:42", "event": "train_log", "step": 1870, "epoch": 0.8069039913700108, "progress_pct": 13.45, "epoch_pct": 13.45, "eta": "58:49:52", "max_grad_norm": 0.8, "loss": 0.665183424949646, "grad_norm": 1.1774994134902954, "learning_rate": 9.914107467442859e-05}
+{"ts": "2025-12-23T23:32:02", "event": "train_log", "step": 1872, "epoch": 0.8077669902912621, "progress_pct": 13.46, "epoch_pct": 13.46, "eta": "58:47:38", "max_grad_norm": 0.8, "loss": 0.7597744464874268, "grad_norm": 1.0606465339660645, "learning_rate": 9.91365372275303e-05}
+{"ts": "2025-12-23T23:32:21", "event": "train_log", "step": 1874, "epoch": 0.8086299892125135, "progress_pct": 13.47, "epoch_pct": 13.48, "eta": "58:45:15", "max_grad_norm": 0.8, "loss": 0.7453591227531433, "grad_norm": 1.397161841392517, "learning_rate": 9.913198793157757e-05}
+{"ts": "2025-12-23T23:32:41", "event": "train_log", "step": 1876, "epoch": 0.8094929881337648, "progress_pct": 13.49, "epoch_pct": 13.49, "eta": "58:43:07", "max_grad_norm": 0.8, "loss": 0.715792179107666, "grad_norm": 1.250003695487976, "learning_rate": 9.912742678766742e-05}
+{"ts": "2025-12-23T23:33:00", "event": "train_log", "step": 1878, "epoch": 0.8103559870550162, "progress_pct": 13.5, "epoch_pct": 13.51, "eta": "58:40:49", "max_grad_norm": 0.8, "loss": 0.7094268798828125, "grad_norm": 1.2343473434448242, "learning_rate": 9.912285379689979e-05}
+{"ts": "2025-12-23T23:33:20", "event": "train_log", "step": 1880, "epoch": 0.8112189859762675, "progress_pct": 13.52, "epoch_pct": 13.52, "eta": "58:38:38", "max_grad_norm": 0.8, "loss": 0.6652816534042358, "grad_norm": 1.453764796257019, "learning_rate": 9.911826896037739e-05}
+{"ts": "2025-12-23T23:33:41", "event": "train_log", "step": 1882, "epoch": 0.8120819848975189, "progress_pct": 13.53, "epoch_pct": 13.53, "eta": "58:36:32", "max_grad_norm": 0.8, "loss": 0.6764529347419739, "grad_norm": 1.030903935432434, "learning_rate": 9.911367227920587e-05}
+{"ts": "2025-12-23T23:34:00", "event": "train_log", "step": 1884, "epoch": 0.8129449838187702, "progress_pct": 13.55, "epoch_pct": 13.55, "eta": "58:34:11", "max_grad_norm": 0.8, "loss": 0.7043389678001404, "grad_norm": 1.2725969552993774, "learning_rate": 9.91090637544937e-05}
+{"ts": "2025-12-23T23:34:21", "event": "train_log", "step": 1886, "epoch": 0.8138079827400215, "progress_pct": 13.56, "epoch_pct": 13.56, "eta": "58:32:04", "max_grad_norm": 0.8, "loss": 0.6960558295249939, "grad_norm": 1.1610769033432007, "learning_rate": 9.910444338735216e-05}
+{"ts": "2025-12-23T23:34:39", "event": "train_log", "step": 1888, "epoch": 0.8146709816612729, "progress_pct": 13.57, "epoch_pct": 13.58, "eta": "58:29:41", "max_grad_norm": 0.8, "loss": 0.8029319643974304, "grad_norm": 1.2049180269241333, "learning_rate": 9.909981117889548e-05}
+{"ts": "2025-12-23T23:35:00", "event": "train_log", "step": 1890, "epoch": 0.8155339805825242, "progress_pct": 13.59, "epoch_pct": 13.59, "eta": "58:27:41", "max_grad_norm": 0.8, "loss": 0.683114767074585, "grad_norm": 1.1671873331069946, "learning_rate": 9.909516713024068e-05}
+{"ts": "2025-12-23T23:35:21", "event": "train_log", "step": 1892, "epoch": 0.8163969795037757, "progress_pct": 13.6, "epoch_pct": 13.61, "eta": "58:25:32", "max_grad_norm": 0.8, "loss": 0.7193703055381775, "grad_norm": 1.247768759727478, "learning_rate": 9.909051124250764e-05}
+{"ts": "2025-12-23T23:35:41", "event": "train_log", "step": 1894, "epoch": 0.817259978425027, "progress_pct": 13.62, "epoch_pct": 13.62, "eta": "58:23:25", "max_grad_norm": 0.8, "loss": 0.7957084774971008, "grad_norm": 1.1376357078552246, "learning_rate": 9.908584351681911e-05}
+{"ts": "2025-12-23T23:35:59", "event": "train_log", "step": 1896, "epoch": 0.8181229773462784, "progress_pct": 13.63, "epoch_pct": 13.64, "eta": "58:21:05", "max_grad_norm": 0.8, "loss": 0.7168112993240356, "grad_norm": 1.1918323040008545, "learning_rate": 9.90811639543007e-05}
+{"ts": "2025-12-23T23:36:19", "event": "train_log", "step": 1898, "epoch": 0.8189859762675297, "progress_pct": 13.65, "epoch_pct": 13.65, "eta": "58:18:53", "max_grad_norm": 0.8, "loss": 0.7519245147705078, "grad_norm": 1.1249650716781616, "learning_rate": 9.907647255608087e-05}
+{"ts": "2025-12-23T23:36:39", "event": "train_log", "step": 1900, "epoch": 0.819848975188781, "progress_pct": 13.66, "epoch_pct": 13.66, "eta": "58:16:41", "max_grad_norm": 0.8, "loss": 0.7496176362037659, "grad_norm": 1.0688576698303223, "learning_rate": 9.907176932329092e-05}
+{"ts": "2025-12-23T23:50:29", "event": "train_log", "step": 1900, "epoch": 0.819848975188781, "progress_pct": 13.66, "epoch_pct": 13.66, "eta": "59:44:08", "max_grad_norm": 0.8, "eval_loss": 0.7533769607543945, "eval_runtime": 830.2894, "eval_samples_per_second": 2.481, "eval_steps_per_second": 2.481}
+{"ts": "2025-12-23T23:50:49", "event": "train_log", "step": 1902, "epoch": 0.8207119741100324, "progress_pct": 13.68, "epoch_pct": 13.68, "eta": "59:41:50", "max_grad_norm": 0.8, "loss": 0.639653205871582, "grad_norm": 0.9264270663261414, "learning_rate": 9.906705425706499e-05}
+{"ts": "2025-12-23T23:51:07", "event": "train_log", "step": 1904, "epoch": 0.8215749730312837, "progress_pct": 13.69, "epoch_pct": 13.69, "eta": "59:39:26", "max_grad_norm": 0.8, "loss": 0.7555551528930664, "grad_norm": 1.2231628894805908, "learning_rate": 9.906232735854016e-05}
+{"ts": "2025-12-23T23:51:26", "event": "train_log", "step": 1906, "epoch": 0.8224379719525351, "progress_pct": 13.7, "epoch_pct": 13.71, "eta": "59:37:02", "max_grad_norm": 0.8, "loss": 0.7113592028617859, "grad_norm": 1.3369156122207642, "learning_rate": 9.905758862885624e-05}
+{"ts": "2025-12-23T23:51:45", "event": "train_log", "step": 1908, "epoch": 0.8233009708737864, "progress_pct": 13.72, "epoch_pct": 13.72, "eta": "59:34:44", "max_grad_norm": 0.8, "loss": 0.6823481321334839, "grad_norm": 1.217238187789917, "learning_rate": 9.905283806915597e-05}
+{"ts": "2025-12-23T23:52:04", "event": "train_log", "step": 1910, "epoch": 0.8241639697950378, "progress_pct": 13.73, "epoch_pct": 13.74, "eta": "59:32:22", "max_grad_norm": 0.8, "loss": 0.7347100377082825, "grad_norm": 1.2462831735610962, "learning_rate": 9.904807568058495e-05}
+{"ts": "2025-12-23T23:52:22", "event": "train_log", "step": 1912, "epoch": 0.8250269687162891, "progress_pct": 13.75, "epoch_pct": 13.75, "eta": "59:29:56", "max_grad_norm": 0.8, "loss": 0.7981513738632202, "grad_norm": 1.2507221698760986, "learning_rate": 9.904330146429157e-05}
+{"ts": "2025-12-23T23:52:41", "event": "train_log", "step": 1914, "epoch": 0.8258899676375404, "progress_pct": 13.76, "epoch_pct": 13.76, "eta": "59:27:34", "max_grad_norm": 0.8, "loss": 0.759779155254364, "grad_norm": 1.2246814966201782, "learning_rate": 9.903851542142715e-05}
+{"ts": "2025-12-23T23:53:00", "event": "train_log", "step": 1916, "epoch": 0.8267529665587918, "progress_pct": 13.78, "epoch_pct": 13.78, "eta": "59:25:11", "max_grad_norm": 0.8, "loss": 0.741083562374115, "grad_norm": 1.1746087074279785, "learning_rate": 9.903371755314579e-05}
+{"ts": "2025-12-23T23:53:18", "event": "train_log", "step": 1918, "epoch": 0.8276159654800431, "progress_pct": 13.79, "epoch_pct": 13.79, "eta": "59:22:46", "max_grad_norm": 0.8, "loss": 0.7491101622581482, "grad_norm": 1.1913539171218872, "learning_rate": 9.902890786060448e-05}
+{"ts": "2025-12-23T23:53:37", "event": "train_log", "step": 1920, "epoch": 0.8284789644012945, "progress_pct": 13.81, "epoch_pct": 13.81, "eta": "59:20:28", "max_grad_norm": 0.8, "loss": 0.7655978202819824, "grad_norm": 1.2342501878738403, "learning_rate": 9.902408634496307e-05}
+{"ts": "2025-12-23T23:53:56", "event": "train_log", "step": 1922, "epoch": 0.8293419633225458, "progress_pct": 13.82, "epoch_pct": 13.82, "eta": "59:18:10", "max_grad_norm": 0.8, "loss": 0.7199848890304565, "grad_norm": 1.075892686843872, "learning_rate": 9.901925300738424e-05}
+{"ts": "2025-12-23T23:54:14", "event": "train_log", "step": 1924, "epoch": 0.8302049622437971, "progress_pct": 13.83, "epoch_pct": 13.84, "eta": "59:15:45", "max_grad_norm": 0.8, "loss": 0.8325693011283875, "grad_norm": 1.1517611742019653, "learning_rate": 9.901440784903353e-05}
+{"ts": "2025-12-23T23:54:33", "event": "train_log", "step": 1926, "epoch": 0.8310679611650486, "progress_pct": 13.85, "epoch_pct": 13.85, "eta": "59:13:21", "max_grad_norm": 0.8, "loss": 0.7674157023429871, "grad_norm": 1.2084628343582153, "learning_rate": 9.900955087107933e-05}
+{"ts": "2025-12-23T23:54:51", "event": "train_log", "step": 1928, "epoch": 0.8319309600862999, "progress_pct": 13.86, "epoch_pct": 13.87, "eta": "59:11:00", "max_grad_norm": 0.8, "loss": 0.7415181398391724, "grad_norm": 1.334679365158081, "learning_rate": 9.900468207469287e-05}
+{"ts": "2025-12-23T23:55:12", "event": "train_log", "step": 1930, "epoch": 0.8327939590075513, "progress_pct": 13.88, "epoch_pct": 13.88, "eta": "59:08:54", "max_grad_norm": 0.8, "loss": 0.672336220741272, "grad_norm": 1.0505887269973755, "learning_rate": 9.899980146104824e-05}
+{"ts": "2025-12-23T23:55:31", "event": "train_log", "step": 1932, "epoch": 0.8336569579288026, "progress_pct": 13.89, "epoch_pct": 13.89, "eta": "59:06:37", "max_grad_norm": 0.8, "loss": 0.8077969551086426, "grad_norm": 1.415797233581543, "learning_rate": 9.899490903132238e-05}
+{"ts": "2025-12-23T23:55:50", "event": "train_log", "step": 1934, "epoch": 0.834519956850054, "progress_pct": 13.91, "epoch_pct": 13.91, "eta": "59:04:14", "max_grad_norm": 0.8, "loss": 0.8053045272827148, "grad_norm": 1.1793439388275146, "learning_rate": 9.899000478669507e-05}
+{"ts": "2025-12-23T23:56:09", "event": "train_log", "step": 1936, "epoch": 0.8353829557713053, "progress_pct": 13.92, "epoch_pct": 13.92, "eta": "59:02:00", "max_grad_norm": 0.8, "loss": 0.7585914731025696, "grad_norm": 1.196115493774414, "learning_rate": 9.898508872834896e-05}
+{"ts": "2025-12-23T23:56:29", "event": "train_log", "step": 1938, "epoch": 0.8362459546925566, "progress_pct": 13.93, "epoch_pct": 13.94, "eta": "58:59:46", "max_grad_norm": 0.8, "loss": 0.691072940826416, "grad_norm": 1.2181763648986816, "learning_rate": 9.898016085746953e-05}
+{"ts": "2025-12-23T23:56:48", "event": "train_log", "step": 1940, "epoch": 0.837108953613808, "progress_pct": 13.95, "epoch_pct": 13.95, "eta": "58:57:32", "max_grad_norm": 0.8, "loss": 0.7723121643066406, "grad_norm": 1.2929240465164185, "learning_rate": 9.897522117524512e-05}
+{"ts": "2025-12-23T23:57:07", "event": "train_log", "step": 1942, "epoch": 0.8379719525350593, "progress_pct": 13.96, "epoch_pct": 13.97, "eta": "58:55:13", "max_grad_norm": 0.8, "loss": 0.8236626386642456, "grad_norm": 1.2859764099121094, "learning_rate": 9.89702696828669e-05}
+{"ts": "2025-12-23T23:57:27", "event": "train_log", "step": 1944, "epoch": 0.8388349514563107, "progress_pct": 13.98, "epoch_pct": 13.98, "eta": "58:53:03", "max_grad_norm": 0.8, "loss": 0.7315511703491211, "grad_norm": 1.2043784856796265, "learning_rate": 9.89653063815289e-05}
+{"ts": "2025-12-23T23:57:47", "event": "train_log", "step": 1946, "epoch": 0.839697950377562, "progress_pct": 13.99, "epoch_pct": 13.99, "eta": "58:50:51", "max_grad_norm": 0.8, "loss": 0.6707082986831665, "grad_norm": 1.2063398361206055, "learning_rate": 9.896033127242802e-05}
+{"ts": "2025-12-23T23:58:07", "event": "train_log", "step": 1948, "epoch": 0.8405609492988134, "progress_pct": 14.01, "epoch_pct": 14.01, "eta": "58:48:43", "max_grad_norm": 0.8, "loss": 0.6920633912086487, "grad_norm": 1.1906715631484985, "learning_rate": 9.895534435676393e-05}
+{"ts": "2025-12-23T23:58:26", "event": "train_log", "step": 1950, "epoch": 0.8414239482200647, "progress_pct": 14.02, "epoch_pct": 14.02, "eta": "58:46:29", "max_grad_norm": 0.8, "loss": 0.7134768962860107, "grad_norm": 1.122567892074585, "learning_rate": 9.895034563573926e-05}
+{"ts": "2025-12-23T23:58:48", "event": "train_log", "step": 1952, "epoch": 0.842286947141316, "progress_pct": 14.04, "epoch_pct": 14.04, "eta": "58:44:27", "max_grad_norm": 0.8, "loss": 0.6333967447280884, "grad_norm": 1.0681787729263306, "learning_rate": 9.894533511055943e-05}
+{"ts": "2025-12-23T23:59:06", "event": "train_log", "step": 1954, "epoch": 0.8431499460625674, "progress_pct": 14.05, "epoch_pct": 14.05, "eta": "58:42:08", "max_grad_norm": 0.8, "loss": 0.7744026780128479, "grad_norm": 1.5216220617294312, "learning_rate": 9.894031278243266e-05}
+{"ts": "2025-12-23T23:59:27", "event": "train_log", "step": 1956, "epoch": 0.8440129449838187, "progress_pct": 14.06, "epoch_pct": 14.07, "eta": "58:40:06", "max_grad_norm": 0.8, "loss": 0.7678086757659912, "grad_norm": 1.0934535264968872, "learning_rate": 9.89352786525701e-05}
+{"ts": "2025-12-23T23:59:47", "event": "train_log", "step": 1958, "epoch": 0.8448759439050701, "progress_pct": 14.08, "epoch_pct": 14.08, "eta": "58:37:59", "max_grad_norm": 0.8, "loss": 0.7344630360603333, "grad_norm": 1.223285436630249, "learning_rate": 9.89302327221857e-05}
+{"ts": "2025-12-24T00:00:07", "event": "train_log", "step": 1960, "epoch": 0.8457389428263214, "progress_pct": 14.09, "epoch_pct": 14.1, "eta": "58:35:48", "max_grad_norm": 0.8, "loss": 0.76033616065979, "grad_norm": 1.1123322248458862, "learning_rate": 9.892517499249623e-05}
+{"ts": "2025-12-24T00:00:28", "event": "train_log", "step": 1962, "epoch": 0.8466019417475729, "progress_pct": 14.11, "epoch_pct": 14.11, "eta": "58:33:43", "max_grad_norm": 0.8, "loss": 0.7176017761230469, "grad_norm": 1.0472193956375122, "learning_rate": 9.89201054647214e-05}
+{"ts": "2025-12-24T00:00:48", "event": "train_log", "step": 1964, "epoch": 0.8474649406688242, "progress_pct": 14.12, "epoch_pct": 14.12, "eta": "58:31:38", "max_grad_norm": 0.8, "loss": 0.7220107316970825, "grad_norm": 1.8538368940353394, "learning_rate": 9.891502414008365e-05}
+{"ts": "2025-12-24T00:01:08", "event": "train_log", "step": 1966, "epoch": 0.8483279395900755, "progress_pct": 14.14, "epoch_pct": 14.14, "eta": "58:29:30", "max_grad_norm": 0.8, "loss": 0.7399556636810303, "grad_norm": 1.2280668020248413, "learning_rate": 9.890993101980834e-05}
+{"ts": "2025-12-24T00:01:28", "event": "train_log", "step": 1968, "epoch": 0.8491909385113269, "progress_pct": 14.15, "epoch_pct": 14.15, "eta": "58:27:19", "max_grad_norm": 0.8, "loss": 0.7700466513633728, "grad_norm": 1.1482552289962769, "learning_rate": 9.890482610512364e-05}
+{"ts": "2025-12-24T00:01:48", "event": "train_log", "step": 1970, "epoch": 0.8500539374325782, "progress_pct": 14.16, "epoch_pct": 14.17, "eta": "58:25:15", "max_grad_norm": 0.8, "loss": 0.6666419506072998, "grad_norm": 1.1877986192703247, "learning_rate": 9.889970939726061e-05}
+{"ts": "2025-12-24T00:02:06", "event": "train_log", "step": 1972, "epoch": 0.8509169363538296, "progress_pct": 14.18, "epoch_pct": 14.18, "eta": "58:22:55", "max_grad_norm": 0.8, "loss": 0.7883278131484985, "grad_norm": 1.407792329788208, "learning_rate": 9.88945808974531e-05}
+{"ts": "2025-12-24T00:02:25", "event": "train_log", "step": 1974, "epoch": 0.8517799352750809, "progress_pct": 14.19, "epoch_pct": 14.2, "eta": "58:20:38", "max_grad_norm": 0.8, "loss": 0.7441611289978027, "grad_norm": 1.1426323652267456, "learning_rate": 9.888944060693781e-05}
+{"ts": "2025-12-24T00:02:44", "event": "train_log", "step": 1976, "epoch": 0.8526429341963323, "progress_pct": 14.21, "epoch_pct": 14.21, "eta": "58:18:29", "max_grad_norm": 0.8, "loss": 0.6464171409606934, "grad_norm": 1.2660622596740723, "learning_rate": 9.888428852695432e-05}
+{"ts": "2025-12-24T00:03:05", "event": "train_log", "step": 1978, "epoch": 0.8535059331175836, "progress_pct": 14.22, "epoch_pct": 14.23, "eta": "58:16:24", "max_grad_norm": 0.8, "loss": 0.7242478728294373, "grad_norm": 1.1639022827148438, "learning_rate": 9.887912465874502e-05}
+{"ts": "2025-12-24T00:03:24", "event": "train_log", "step": 1980, "epoch": 0.8543689320388349, "progress_pct": 14.24, "epoch_pct": 14.24, "eta": "58:14:16", "max_grad_norm": 0.8, "loss": 0.7431875467300415, "grad_norm": 1.2546422481536865, "learning_rate": 9.887394900355516e-05}
+{"ts": "2025-12-24T00:03:44", "event": "train_log", "step": 1982, "epoch": 0.8552319309600863, "progress_pct": 14.25, "epoch_pct": 14.25, "eta": "58:12:09", "max_grad_norm": 0.8, "loss": 0.7662370204925537, "grad_norm": 1.1057130098342896, "learning_rate": 9.886876156263282e-05}
+{"ts": "2025-12-24T00:04:04", "event": "train_log", "step": 1984, "epoch": 0.8560949298813376, "progress_pct": 14.27, "epoch_pct": 14.27, "eta": "58:10:01", "max_grad_norm": 0.8, "loss": 0.7120938897132874, "grad_norm": 1.1017125844955444, "learning_rate": 9.886356233722894e-05}
+{"ts": "2025-12-24T00:04:23", "event": "train_log", "step": 1986, "epoch": 0.856957928802589, "progress_pct": 14.28, "epoch_pct": 14.28, "eta": "58:07:52", "max_grad_norm": 0.8, "loss": 0.789219856262207, "grad_norm": 1.3548704385757446, "learning_rate": 9.885835132859727e-05}
+{"ts": "2025-12-24T00:04:42", "event": "train_log", "step": 1988, "epoch": 0.8578209277238403, "progress_pct": 14.29, "epoch_pct": 14.3, "eta": "58:05:40", "max_grad_norm": 0.8, "loss": 0.7599530816078186, "grad_norm": 1.1920398473739624, "learning_rate": 9.885312853799443e-05}
+{"ts": "2025-12-24T00:05:00", "event": "train_log", "step": 1990, "epoch": 0.8586839266450917, "progress_pct": 14.31, "epoch_pct": 14.31, "eta": "58:03:22", "max_grad_norm": 0.8, "loss": 0.8315809369087219, "grad_norm": 1.170946478843689, "learning_rate": 9.884789396667988e-05}
+{"ts": "2025-12-24T00:05:20", "event": "train_log", "step": 1992, "epoch": 0.859546925566343, "progress_pct": 14.32, "epoch_pct": 14.33, "eta": "58:01:15", "max_grad_norm": 0.8, "loss": 0.713994026184082, "grad_norm": 1.2530180215835571, "learning_rate": 9.88426476159159e-05}
+{"ts": "2025-12-24T00:05:40", "event": "train_log", "step": 1994, "epoch": 0.8604099244875943, "progress_pct": 14.34, "epoch_pct": 14.34, "eta": "57:59:11", "max_grad_norm": 0.8, "loss": 0.6656507849693298, "grad_norm": 1.1056885719299316, "learning_rate": 9.883738948696765e-05}
+{"ts": "2025-12-24T00:06:00", "event": "train_log", "step": 1996, "epoch": 0.8612729234088458, "progress_pct": 14.35, "epoch_pct": 14.35, "eta": "57:57:08", "max_grad_norm": 0.8, "loss": 0.7164257764816284, "grad_norm": 1.169382929801941, "learning_rate": 9.883211958110308e-05}
+{"ts": "2025-12-24T00:06:19", "event": "train_log", "step": 1998, "epoch": 0.8621359223300971, "progress_pct": 14.37, "epoch_pct": 14.37, "eta": "57:54:57", "max_grad_norm": 0.8, "loss": 0.6885558366775513, "grad_norm": 1.1970325708389282, "learning_rate": 9.882683789959299e-05}
+{"ts": "2025-12-24T00:06:39", "event": "train_log", "step": 2000, "epoch": 0.8629989212513485, "progress_pct": 14.38, "epoch_pct": 14.38, "eta": "57:52:48", "max_grad_norm": 0.8, "loss": 0.7265812754631042, "grad_norm": 1.1700761318206787, "learning_rate": 9.882154444371107e-05}
+{"ts": "2025-12-24T00:20:24", "event": "train_log", "step": 2000, "epoch": 0.8629989212513485, "progress_pct": 14.38, "epoch_pct": 14.38, "eta": "59:14:41", "max_grad_norm": 0.8, "eval_loss": 0.7487027645111084, "eval_runtime": 825.141, "eval_samples_per_second": 2.497, "eval_steps_per_second": 2.497}
+{"ts": "2025-12-24T00:20:44", "event": "train_log", "step": 2002, "epoch": 0.8638619201725998, "progress_pct": 14.39, "epoch_pct": 14.4, "eta": "59:12:32", "max_grad_norm": 0.8, "loss": 0.7371920347213745, "grad_norm": 1.3120243549346924, "learning_rate": 9.88162392147338e-05}
+{"ts": "2025-12-24T00:21:02", "event": "train_log", "step": 2004, "epoch": 0.8647249190938512, "progress_pct": 14.41, "epoch_pct": 14.41, "eta": "59:10:09", "max_grad_norm": 0.8, "loss": 0.7453275918960571, "grad_norm": 1.098697304725647, "learning_rate": 9.881092221394049e-05}
+{"ts": "2025-12-24T00:21:20", "event": "train_log", "step": 2006, "epoch": 0.8655879180151025, "progress_pct": 14.42, "epoch_pct": 14.43, "eta": "59:07:52", "max_grad_norm": 0.8, "loss": 0.7562432289123535, "grad_norm": 1.1961677074432373, "learning_rate": 9.880559344261334e-05}
+{"ts": "2025-12-24T00:21:40", "event": "train_log", "step": 2008, "epoch": 0.8664509169363538, "progress_pct": 14.44, "epoch_pct": 14.44, "eta": "59:05:41", "max_grad_norm": 0.8, "loss": 0.7654905319213867, "grad_norm": 1.1655291318893433, "learning_rate": 9.880025290203735e-05}
+{"ts": "2025-12-24T00:22:00", "event": "train_log", "step": 2010, "epoch": 0.8673139158576052, "progress_pct": 14.45, "epoch_pct": 14.46, "eta": "59:03:31", "max_grad_norm": 0.8, "loss": 0.6807888746261597, "grad_norm": 1.0979351997375488, "learning_rate": 9.879490059350036e-05}
+{"ts": "2025-12-24T00:22:20", "event": "train_log", "step": 2012, "epoch": 0.8681769147788565, "progress_pct": 14.47, "epoch_pct": 14.47, "eta": "59:01:20", "max_grad_norm": 0.8, "loss": 0.6887364387512207, "grad_norm": 1.0378475189208984, "learning_rate": 9.878953651829306e-05}
+{"ts": "2025-12-24T00:22:40", "event": "train_log", "step": 2014, "epoch": 0.8690399137001079, "progress_pct": 14.48, "epoch_pct": 14.48, "eta": "58:59:12", "max_grad_norm": 0.8, "loss": 0.718578577041626, "grad_norm": 2.0228428840637207, "learning_rate": 9.878416067770898e-05}
+{"ts": "2025-12-24T00:22:58", "event": "train_log", "step": 2016, "epoch": 0.8699029126213592, "progress_pct": 14.5, "epoch_pct": 14.5, "eta": "58:56:54", "max_grad_norm": 0.8, "loss": 0.7424600720405579, "grad_norm": 1.1610580682754517, "learning_rate": 9.877877307304446e-05}
+{"ts": "2025-12-24T00:23:18", "event": "train_log", "step": 2018, "epoch": 0.8707659115426106, "progress_pct": 14.51, "epoch_pct": 14.51, "eta": "58:54:45", "max_grad_norm": 0.8, "loss": 0.7450630068778992, "grad_norm": 1.1415833234786987, "learning_rate": 9.877337370559873e-05}
+{"ts": "2025-12-24T00:23:37", "event": "train_log", "step": 2020, "epoch": 0.8716289104638619, "progress_pct": 14.52, "epoch_pct": 14.53, "eta": "58:52:29", "max_grad_norm": 0.8, "loss": 0.7398725748062134, "grad_norm": 1.1008672714233398, "learning_rate": 9.876796257667379e-05}
+{"ts": "2025-12-24T00:23:57", "event": "train_log", "step": 2022, "epoch": 0.8724919093851132, "progress_pct": 14.54, "epoch_pct": 14.54, "eta": "58:50:26", "max_grad_norm": 0.8, "loss": 0.7122271060943604, "grad_norm": 1.1662763357162476, "learning_rate": 9.876253968757452e-05}
+{"ts": "2025-12-24T00:24:17", "event": "train_log", "step": 2024, "epoch": 0.8733549083063646, "progress_pct": 14.55, "epoch_pct": 14.56, "eta": "58:48:17", "max_grad_norm": 0.8, "loss": 0.7026485800743103, "grad_norm": 1.1835166215896606, "learning_rate": 9.875710503960862e-05}
+{"ts": "2025-12-24T00:24:37", "event": "train_log", "step": 2026, "epoch": 0.8742179072276159, "progress_pct": 14.57, "epoch_pct": 14.57, "eta": "58:46:07", "max_grad_norm": 0.8, "loss": 0.7211011052131653, "grad_norm": 1.1602214574813843, "learning_rate": 9.875165863408666e-05}
+{"ts": "2025-12-24T00:24:57", "event": "train_log", "step": 2028, "epoch": 0.8750809061488674, "progress_pct": 14.58, "epoch_pct": 14.58, "eta": "58:43:59", "max_grad_norm": 0.8, "loss": 0.7565322518348694, "grad_norm": 1.222589373588562, "learning_rate": 9.874620047232198e-05}
+{"ts": "2025-12-24T00:25:17", "event": "train_log", "step": 2030, "epoch": 0.8759439050701187, "progress_pct": 14.6, "epoch_pct": 14.6, "eta": "58:41:55", "max_grad_norm": 0.8, "loss": 0.7173075675964355, "grad_norm": 1.203661561012268, "learning_rate": 9.87407305556308e-05}
+{"ts": "2025-12-24T00:25:35", "event": "train_log", "step": 2032, "epoch": 0.87680690399137, "progress_pct": 14.61, "epoch_pct": 14.61, "eta": "58:39:39", "max_grad_norm": 0.8, "loss": 0.820809006690979, "grad_norm": 1.3316307067871094, "learning_rate": 9.87352488853322e-05}
+{"ts": "2025-12-24T00:25:54", "event": "train_log", "step": 2034, "epoch": 0.8776699029126214, "progress_pct": 14.62, "epoch_pct": 14.63, "eta": "58:37:26", "max_grad_norm": 0.8, "loss": 0.8689352869987488, "grad_norm": 1.3071057796478271, "learning_rate": 9.872975546274799e-05}
+{"ts": "2025-12-24T00:26:13", "event": "train_log", "step": 2036, "epoch": 0.8785329018338727, "progress_pct": 14.64, "epoch_pct": 14.64, "eta": "58:35:12", "max_grad_norm": 0.8, "loss": 0.7090175151824951, "grad_norm": 1.3557701110839844, "learning_rate": 9.872425028920294e-05}
+{"ts": "2025-12-24T00:26:32", "event": "train_log", "step": 2038, "epoch": 0.8793959007551241, "progress_pct": 14.65, "epoch_pct": 14.66, "eta": "58:33:02", "max_grad_norm": 0.8, "loss": 0.6883202195167542, "grad_norm": 1.2234264612197876, "learning_rate": 9.871873336602459e-05}
+{"ts": "2025-12-24T00:26:51", "event": "train_log", "step": 2040, "epoch": 0.8802588996763754, "progress_pct": 14.67, "epoch_pct": 14.67, "eta": "58:30:49", "max_grad_norm": 0.8, "loss": 0.7134175896644592, "grad_norm": 1.3064714670181274, "learning_rate": 9.87132046945433e-05}
+{"ts": "2025-12-24T00:27:10", "event": "train_log", "step": 2042, "epoch": 0.8811218985976268, "progress_pct": 14.68, "epoch_pct": 14.69, "eta": "58:28:37", "max_grad_norm": 0.8, "loss": 0.7127227783203125, "grad_norm": 1.2502511739730835, "learning_rate": 9.870766427609232e-05}
+{"ts": "2025-12-24T00:27:30", "event": "train_log", "step": 2044, "epoch": 0.8819848975188781, "progress_pct": 14.7, "epoch_pct": 14.7, "eta": "58:26:32", "max_grad_norm": 0.8, "loss": 0.719437301158905, "grad_norm": 1.1354259252548218, "learning_rate": 9.870211211200766e-05}
+{"ts": "2025-12-24T00:27:49", "event": "train_log", "step": 2046, "epoch": 0.8828478964401294, "progress_pct": 14.71, "epoch_pct": 14.71, "eta": "58:24:22", "max_grad_norm": 0.8, "loss": 0.6933091282844543, "grad_norm": 1.1706442832946777, "learning_rate": 9.869654820362821e-05}
+{"ts": "2025-12-24T00:28:08", "event": "train_log", "step": 2048, "epoch": 0.8837108953613808, "progress_pct": 14.73, "epoch_pct": 14.73, "eta": "58:22:13", "max_grad_norm": 0.8, "loss": 0.6825079321861267, "grad_norm": 1.0227864980697632, "learning_rate": 9.86909725522957e-05}
+{"ts": "2025-12-24T00:28:28", "event": "train_log", "step": 2050, "epoch": 0.8845738942826321, "progress_pct": 14.74, "epoch_pct": 14.74, "eta": "58:20:07", "max_grad_norm": 0.8, "loss": 0.7306133508682251, "grad_norm": 1.103418231010437, "learning_rate": 9.868538515935465e-05}
+{"ts": "2025-12-24T00:28:49", "event": "train_log", "step": 2052, "epoch": 0.8854368932038835, "progress_pct": 14.75, "epoch_pct": 14.76, "eta": "58:18:04", "max_grad_norm": 0.8, "loss": 0.6956495046615601, "grad_norm": 1.0938395261764526, "learning_rate": 9.867978602615245e-05}
+{"ts": "2025-12-24T00:29:09", "event": "train_log", "step": 2054, "epoch": 0.8862998921251348, "progress_pct": 14.77, "epoch_pct": 14.77, "eta": "58:16:02", "max_grad_norm": 0.8, "loss": 0.7740456461906433, "grad_norm": 1.115692377090454, "learning_rate": 9.86741751540393e-05}
+{"ts": "2025-12-24T00:29:29", "event": "train_log", "step": 2056, "epoch": 0.8871628910463862, "progress_pct": 14.78, "epoch_pct": 14.79, "eta": "58:13:57", "max_grad_norm": 0.8, "loss": 0.6996002197265625, "grad_norm": 1.0264382362365723, "learning_rate": 9.866855254436824e-05}
+{"ts": "2025-12-24T00:29:50", "event": "train_log", "step": 2058, "epoch": 0.8880258899676375, "progress_pct": 14.8, "epoch_pct": 14.8, "eta": "58:11:59", "max_grad_norm": 0.8, "loss": 0.760357677936554, "grad_norm": 1.2523163557052612, "learning_rate": 9.866291819849511e-05}
+{"ts": "2025-12-24T00:30:10", "event": "train_log", "step": 2060, "epoch": 0.8888888888888888, "progress_pct": 14.81, "epoch_pct": 14.81, "eta": "58:10:00", "max_grad_norm": 0.8, "loss": 0.6389813423156738, "grad_norm": 1.0595446825027466, "learning_rate": 9.865727211777864e-05}
+{"ts": "2025-12-24T00:30:30", "event": "train_log", "step": 2062, "epoch": 0.8897518878101403, "progress_pct": 14.83, "epoch_pct": 14.83, "eta": "58:07:53", "max_grad_norm": 0.8, "loss": 0.7859060168266296, "grad_norm": 1.198530912399292, "learning_rate": 9.865161430358036e-05}
+{"ts": "2025-12-24T00:30:50", "event": "train_log", "step": 2064, "epoch": 0.8906148867313916, "progress_pct": 14.84, "epoch_pct": 14.84, "eta": "58:05:52", "max_grad_norm": 0.8, "loss": 0.7120063900947571, "grad_norm": 1.0271193981170654, "learning_rate": 9.864594475726461e-05}
+{"ts": "2025-12-24T00:31:10", "event": "train_log", "step": 2066, "epoch": 0.891477885652643, "progress_pct": 14.85, "epoch_pct": 14.86, "eta": "58:03:48", "max_grad_norm": 0.8, "loss": 0.6751878261566162, "grad_norm": 1.1619046926498413, "learning_rate": 9.864026348019857e-05}
+{"ts": "2025-12-24T00:31:30", "event": "train_log", "step": 2068, "epoch": 0.8923408845738943, "progress_pct": 14.87, "epoch_pct": 14.87, "eta": "58:01:42", "max_grad_norm": 0.8, "loss": 0.610360860824585, "grad_norm": 1.2812464237213135, "learning_rate": 9.863457047375227e-05}
+{"ts": "2025-12-24T00:31:49", "event": "train_log", "step": 2070, "epoch": 0.8932038834951457, "progress_pct": 14.88, "epoch_pct": 14.89, "eta": "57:59:33", "max_grad_norm": 0.8, "loss": 0.6540104746818542, "grad_norm": 1.112628698348999, "learning_rate": 9.862886573929855e-05}
+{"ts": "2025-12-24T00:32:07", "event": "train_log", "step": 2072, "epoch": 0.894066882416397, "progress_pct": 14.9, "epoch_pct": 14.9, "eta": "57:57:24", "max_grad_norm": 0.8, "loss": 0.726277232170105, "grad_norm": 1.6257048845291138, "learning_rate": 9.862314927821306e-05}
+{"ts": "2025-12-24T00:32:25", "event": "train_log", "step": 2074, "epoch": 0.8949298813376483, "progress_pct": 14.91, "epoch_pct": 14.92, "eta": "57:55:10", "max_grad_norm": 0.8, "loss": 0.7798765301704407, "grad_norm": 1.227334976196289, "learning_rate": 9.861742109187432e-05}
+{"ts": "2025-12-24T00:32:46", "event": "train_log", "step": 2076, "epoch": 0.8957928802588997, "progress_pct": 14.93, "epoch_pct": 14.93, "eta": "57:53:11", "max_grad_norm": 0.8, "loss": 0.637127161026001, "grad_norm": 1.0728148221969604, "learning_rate": 9.861168118166366e-05}
+{"ts": "2025-12-24T00:33:05", "event": "train_log", "step": 2078, "epoch": 0.896655879180151, "progress_pct": 14.94, "epoch_pct": 14.94, "eta": "57:51:02", "max_grad_norm": 0.8, "loss": 0.7332831025123596, "grad_norm": 1.220282793045044, "learning_rate": 9.860592954896523e-05}
+{"ts": "2025-12-24T00:33:24", "event": "train_log", "step": 2080, "epoch": 0.8975188781014024, "progress_pct": 14.96, "epoch_pct": 14.96, "eta": "57:48:55", "max_grad_norm": 0.8, "loss": 0.7337175607681274, "grad_norm": 1.2287596464157104, "learning_rate": 9.8600166195166e-05}
+{"ts": "2025-12-24T00:33:42", "event": "train_log", "step": 2082, "epoch": 0.8983818770226537, "progress_pct": 14.97, "epoch_pct": 14.97, "eta": "57:46:43", "max_grad_norm": 0.8, "loss": 0.7422571182250977, "grad_norm": 1.0701712369918823, "learning_rate": 9.859439112165579e-05}
+{"ts": "2025-12-24T00:34:00", "event": "train_log", "step": 2084, "epoch": 0.8992448759439051, "progress_pct": 14.98, "epoch_pct": 14.99, "eta": "57:44:34", "max_grad_norm": 0.8, "loss": 0.7010003328323364, "grad_norm": 1.1583118438720703, "learning_rate": 9.858860432982721e-05}
+{"ts": "2025-12-24T00:34:22", "event": "train_log", "step": 2086, "epoch": 0.9001078748651564, "progress_pct": 15.0, "epoch_pct": 15.0, "eta": "57:42:40", "max_grad_norm": 0.8, "loss": 0.6211312413215637, "grad_norm": 0.9644797444343567, "learning_rate": 9.858280582107575e-05}
+{"ts": "2025-12-24T00:34:41", "event": "train_log", "step": 2088, "epoch": 0.9009708737864077, "progress_pct": 15.01, "epoch_pct": 15.02, "eta": "57:40:34", "max_grad_norm": 0.8, "loss": 0.7299860715866089, "grad_norm": 1.2448554039001465, "learning_rate": 9.857699559679968e-05}
+{"ts": "2025-12-24T00:35:00", "event": "train_log", "step": 2090, "epoch": 0.9018338727076591, "progress_pct": 15.03, "epoch_pct": 15.03, "eta": "57:38:26", "max_grad_norm": 0.8, "loss": 0.7461655735969543, "grad_norm": 1.2735580205917358, "learning_rate": 9.85711736584001e-05}
+{"ts": "2025-12-24T00:35:18", "event": "train_log", "step": 2092, "epoch": 0.9026968716289104, "progress_pct": 15.04, "epoch_pct": 15.04, "eta": "57:36:18", "max_grad_norm": 0.8, "loss": 0.7218145132064819, "grad_norm": 1.36841881275177, "learning_rate": 9.856534000728093e-05}
+{"ts": "2025-12-24T00:35:37", "event": "train_log", "step": 2094, "epoch": 0.9035598705501618, "progress_pct": 15.06, "epoch_pct": 15.06, "eta": "57:34:13", "max_grad_norm": 0.8, "loss": 0.7045709490776062, "grad_norm": 1.1471478939056396, "learning_rate": 9.855949464484898e-05}
+{"ts": "2025-12-24T00:35:56", "event": "train_log", "step": 2096, "epoch": 0.9044228694714131, "progress_pct": 15.07, "epoch_pct": 15.07, "eta": "57:32:08", "max_grad_norm": 0.8, "loss": 0.7189744710922241, "grad_norm": 1.058254361152649, "learning_rate": 9.85536375725138e-05}
+{"ts": "2025-12-24T00:36:14", "event": "train_log", "step": 2098, "epoch": 0.9052858683926646, "progress_pct": 15.08, "epoch_pct": 15.09, "eta": "57:29:56", "max_grad_norm": 0.8, "loss": 0.740775465965271, "grad_norm": 1.4070709943771362, "learning_rate": 9.854776879168779e-05}
+{"ts": "2025-12-24T00:36:33", "event": "train_log", "step": 2100, "epoch": 0.9061488673139159, "progress_pct": 15.1, "epoch_pct": 15.1, "eta": "57:27:48", "max_grad_norm": 0.8, "loss": 0.8191262483596802, "grad_norm": 1.2182649374008179, "learning_rate": 9.854188830378622e-05}
+{"ts": "2025-12-24T00:50:25", "event": "train_log", "step": 2100, "epoch": 0.9061488673139159, "progress_pct": 15.1, "epoch_pct": 15.1, "eta": "58:45:49", "max_grad_norm": 0.8, "eval_loss": 0.7449111342430115, "eval_runtime": 832.5135, "eval_samples_per_second": 2.474, "eval_steps_per_second": 2.474}
+{"ts": "2025-12-24T00:50:45", "event": "train_log", "step": 2102, "epoch": 0.9070118662351672, "progress_pct": 15.11, "epoch_pct": 15.12, "eta": "58:43:42", "max_grad_norm": 0.8, "loss": 0.7295873761177063, "grad_norm": 1.139978289604187, "learning_rate": 9.853599611022708e-05}
+{"ts": "2025-12-24T00:51:06", "event": "train_log", "step": 2104, "epoch": 0.9078748651564186, "progress_pct": 15.13, "epoch_pct": 15.13, "eta": "58:41:41", "max_grad_norm": 0.8, "loss": 0.7141753435134888, "grad_norm": 1.1802685260772705, "learning_rate": 9.853009221243127e-05}
+{"ts": "2025-12-24T00:51:26", "event": "train_log", "step": 2106, "epoch": 0.9087378640776699, "progress_pct": 15.14, "epoch_pct": 15.15, "eta": "58:39:38", "max_grad_norm": 0.8, "loss": 0.7557607293128967, "grad_norm": 1.076759934425354, "learning_rate": 9.852417661182253e-05}
+{"ts": "2025-12-24T00:51:46", "event": "train_log", "step": 2108, "epoch": 0.9096008629989213, "progress_pct": 15.16, "epoch_pct": 15.16, "eta": "58:37:33", "max_grad_norm": 0.8, "loss": 0.7025325298309326, "grad_norm": 1.1000943183898926, "learning_rate": 9.851824930982733e-05}
+{"ts": "2025-12-24T00:52:04", "event": "train_log", "step": 2110, "epoch": 0.9104638619201726, "progress_pct": 15.17, "epoch_pct": 15.17, "eta": "58:35:22", "max_grad_norm": 0.8, "loss": 0.6967638731002808, "grad_norm": 1.0235389471054077, "learning_rate": 9.851231030787504e-05}
+{"ts": "2025-12-24T00:52:24", "event": "train_log", "step": 2112, "epoch": 0.911326860841424, "progress_pct": 15.19, "epoch_pct": 15.19, "eta": "58:33:13", "max_grad_norm": 0.8, "loss": 0.7890311479568481, "grad_norm": 1.2311428785324097, "learning_rate": 9.850635960739782e-05}
+{"ts": "2025-12-24T00:52:44", "event": "train_log", "step": 2114, "epoch": 0.9121898597626753, "progress_pct": 15.2, "epoch_pct": 15.2, "eta": "58:31:10", "max_grad_norm": 0.8, "loss": 0.6824729442596436, "grad_norm": 1.173770785331726, "learning_rate": 9.850039720983064e-05}
+{"ts": "2025-12-24T00:53:04", "event": "train_log", "step": 2116, "epoch": 0.9130528586839266, "progress_pct": 15.21, "epoch_pct": 15.22, "eta": "58:29:07", "max_grad_norm": 0.8, "loss": 0.7157722115516663, "grad_norm": 1.2165251970291138, "learning_rate": 9.849442311661131e-05}
+{"ts": "2025-12-24T00:53:23", "event": "train_log", "step": 2118, "epoch": 0.913915857605178, "progress_pct": 15.23, "epoch_pct": 15.23, "eta": "58:27:00", "max_grad_norm": 0.8, "loss": 0.7164111733436584, "grad_norm": 1.251520037651062, "learning_rate": 9.848843732918047e-05}
+{"ts": "2025-12-24T00:53:41", "event": "train_log", "step": 2120, "epoch": 0.9147788565264293, "progress_pct": 15.24, "epoch_pct": 15.25, "eta": "58:24:44", "max_grad_norm": 0.8, "loss": 0.6830796003341675, "grad_norm": 1.1330616474151611, "learning_rate": 9.848243984898153e-05}
+{"ts": "2025-12-24T00:54:01", "event": "train_log", "step": 2122, "epoch": 0.9156418554476807, "progress_pct": 15.26, "epoch_pct": 15.26, "eta": "58:22:40", "max_grad_norm": 0.8, "loss": 0.7149171829223633, "grad_norm": 1.1052451133728027, "learning_rate": 9.847643067746081e-05}
+{"ts": "2025-12-24T00:54:20", "event": "train_log", "step": 2124, "epoch": 0.916504854368932, "progress_pct": 15.27, "epoch_pct": 15.28, "eta": "58:20:36", "max_grad_norm": 0.8, "loss": 0.7213287949562073, "grad_norm": 1.0389703512191772, "learning_rate": 9.847040981606735e-05}
+{"ts": "2025-12-24T00:54:39", "event": "train_log", "step": 2126, "epoch": 0.9173678532901833, "progress_pct": 15.29, "epoch_pct": 15.29, "eta": "58:18:28", "max_grad_norm": 0.8, "loss": 0.703312873840332, "grad_norm": 1.151895523071289, "learning_rate": 9.846437726625308e-05}
+{"ts": "2025-12-24T00:54:57", "event": "train_log", "step": 2128, "epoch": 0.9182308522114347, "progress_pct": 15.3, "epoch_pct": 15.3, "eta": "58:16:16", "max_grad_norm": 0.8, "loss": 0.8059951066970825, "grad_norm": 1.1624029874801636, "learning_rate": 9.845833302947269e-05}
+{"ts": "2025-12-24T00:55:17", "event": "train_log", "step": 2130, "epoch": 0.919093851132686, "progress_pct": 15.31, "epoch_pct": 15.32, "eta": "58:14:10", "max_grad_norm": 0.8, "loss": 0.735476553440094, "grad_norm": 1.2247511148452759, "learning_rate": 9.845227710718377e-05}
+{"ts": "2025-12-24T00:55:36", "event": "train_log", "step": 2132, "epoch": 0.9199568500539375, "progress_pct": 15.33, "epoch_pct": 15.33, "eta": "58:12:03", "max_grad_norm": 0.8, "loss": 0.8130571842193604, "grad_norm": 1.1649967432022095, "learning_rate": 9.844620950084661e-05}
+{"ts": "2025-12-24T00:55:56", "event": "train_log", "step": 2134, "epoch": 0.9208198489751888, "progress_pct": 15.34, "epoch_pct": 15.35, "eta": "58:10:02", "max_grad_norm": 0.8, "loss": 0.6908681392669678, "grad_norm": 1.1915717124938965, "learning_rate": 9.844013021192447e-05}
+{"ts": "2025-12-24T00:56:15", "event": "train_log", "step": 2136, "epoch": 0.9216828478964402, "progress_pct": 15.36, "epoch_pct": 15.36, "eta": "58:07:58", "max_grad_norm": 0.8, "loss": 0.652173638343811, "grad_norm": 1.211199164390564, "learning_rate": 9.843403924188328e-05}
+{"ts": "2025-12-24T00:56:34", "event": "train_log", "step": 2138, "epoch": 0.9225458468176915, "progress_pct": 15.37, "epoch_pct": 15.38, "eta": "58:05:49", "max_grad_norm": 0.8, "loss": 0.6672233939170837, "grad_norm": 1.267242193222046, "learning_rate": 9.842793659219186e-05}
+{"ts": "2025-12-24T00:56:53", "event": "train_log", "step": 2140, "epoch": 0.9234088457389428, "progress_pct": 15.39, "epoch_pct": 15.39, "eta": "58:03:44", "max_grad_norm": 0.8, "loss": 0.7274143695831299, "grad_norm": 1.7038761377334595, "learning_rate": 9.842182226432186e-05}
+{"ts": "2025-12-24T00:57:13", "event": "train_log", "step": 2142, "epoch": 0.9242718446601942, "progress_pct": 15.4, "epoch_pct": 15.4, "eta": "58:01:44", "max_grad_norm": 0.8, "loss": 0.707998514175415, "grad_norm": 1.151426076889038, "learning_rate": 9.841569625974769e-05}
+{"ts": "2025-12-24T00:57:32", "event": "train_log", "step": 2144, "epoch": 0.9251348435814455, "progress_pct": 15.42, "epoch_pct": 15.42, "eta": "57:59:38", "max_grad_norm": 0.8, "loss": 0.6639942526817322, "grad_norm": 1.0414749383926392, "learning_rate": 9.840955857994663e-05}
+{"ts": "2025-12-24T00:57:51", "event": "train_log", "step": 2146, "epoch": 0.9259978425026969, "progress_pct": 15.43, "epoch_pct": 15.43, "eta": "57:57:32", "max_grad_norm": 0.8, "loss": 0.7096070647239685, "grad_norm": 1.1045551300048828, "learning_rate": 9.840340922639875e-05}
+{"ts": "2025-12-24T00:58:10", "event": "train_log", "step": 2148, "epoch": 0.9268608414239482, "progress_pct": 15.44, "epoch_pct": 15.45, "eta": "57:55:27", "max_grad_norm": 0.8, "loss": 0.71246737241745, "grad_norm": 1.1193996667861938, "learning_rate": 9.839724820058695e-05}
+{"ts": "2025-12-24T00:58:30", "event": "train_log", "step": 2150, "epoch": 0.9277238403451996, "progress_pct": 15.46, "epoch_pct": 15.46, "eta": "57:53:23", "max_grad_norm": 0.8, "loss": 0.7351052761077881, "grad_norm": 1.2609117031097412, "learning_rate": 9.839107550399692e-05}
+{"ts": "2025-12-24T00:58:48", "event": "train_log", "step": 2152, "epoch": 0.9285868392664509, "progress_pct": 15.47, "epoch_pct": 15.48, "eta": "57:51:16", "max_grad_norm": 0.8, "loss": 0.7125008702278137, "grad_norm": 1.1127749681472778, "learning_rate": 9.838489113811717e-05}
+{"ts": "2025-12-24T00:59:07", "event": "train_log", "step": 2154, "epoch": 0.9294498381877022, "progress_pct": 15.49, "epoch_pct": 15.49, "eta": "57:49:07", "max_grad_norm": 0.8, "loss": 0.7891085743904114, "grad_norm": 1.4771665334701538, "learning_rate": 9.837869510443905e-05}
+{"ts": "2025-12-24T00:59:25", "event": "train_log", "step": 2156, "epoch": 0.9303128371089536, "progress_pct": 15.5, "epoch_pct": 15.51, "eta": "57:46:59", "max_grad_norm": 0.8, "loss": 0.7370918393135071, "grad_norm": 1.0701470375061035, "learning_rate": 9.83724874044567e-05}
+{"ts": "2025-12-24T00:59:44", "event": "train_log", "step": 2158, "epoch": 0.9311758360302049, "progress_pct": 15.52, "epoch_pct": 15.52, "eta": "57:44:52", "max_grad_norm": 0.8, "loss": 0.7543193697929382, "grad_norm": 1.2045820951461792, "learning_rate": 9.836626803966708e-05}
+{"ts": "2025-12-24T01:00:04", "event": "train_log", "step": 2160, "epoch": 0.9320388349514563, "progress_pct": 15.53, "epoch_pct": 15.53, "eta": "57:42:56", "max_grad_norm": 0.8, "loss": 0.6903626322746277, "grad_norm": 1.1944687366485596, "learning_rate": 9.836003701156996e-05}
+{"ts": "2025-12-24T01:00:24", "event": "train_log", "step": 2162, "epoch": 0.9329018338727076, "progress_pct": 15.55, "epoch_pct": 15.55, "eta": "57:40:53", "max_grad_norm": 0.8, "loss": 0.6923050880432129, "grad_norm": 1.1035394668579102, "learning_rate": 9.83537943216679e-05}
+{"ts": "2025-12-24T01:00:42", "event": "train_log", "step": 2164, "epoch": 0.933764832793959, "progress_pct": 15.56, "epoch_pct": 15.56, "eta": "57:38:43", "max_grad_norm": 0.8, "loss": 0.8099993467330933, "grad_norm": 1.3766103982925415, "learning_rate": 9.834753997146633e-05}
+{"ts": "2025-12-24T01:01:02", "event": "train_log", "step": 2166, "epoch": 0.9346278317152104, "progress_pct": 15.57, "epoch_pct": 15.58, "eta": "57:36:45", "max_grad_norm": 0.8, "loss": 0.7896219491958618, "grad_norm": 1.0874576568603516, "learning_rate": 9.834127396247346e-05}
+{"ts": "2025-12-24T01:01:21", "event": "train_log", "step": 2168, "epoch": 0.9354908306364617, "progress_pct": 15.59, "epoch_pct": 15.59, "eta": "57:34:41", "max_grad_norm": 0.8, "loss": 0.7659367918968201, "grad_norm": 1.1633459329605103, "learning_rate": 9.833499629620028e-05}
+{"ts": "2025-12-24T01:01:39", "event": "train_log", "step": 2170, "epoch": 0.9363538295577131, "progress_pct": 15.6, "epoch_pct": 15.61, "eta": "57:32:36", "max_grad_norm": 0.8, "loss": 0.7474715709686279, "grad_norm": 1.405423641204834, "learning_rate": 9.832870697416066e-05}
+{"ts": "2025-12-24T01:01:59", "event": "train_log", "step": 2172, "epoch": 0.9372168284789644, "progress_pct": 15.62, "epoch_pct": 15.62, "eta": "57:30:35", "max_grad_norm": 0.8, "loss": 0.7018750905990601, "grad_norm": 1.1794270277023315, "learning_rate": 9.83224059978712e-05}
+{"ts": "2025-12-24T01:02:18", "event": "train_log", "step": 2174, "epoch": 0.9380798274002158, "progress_pct": 15.63, "epoch_pct": 15.63, "eta": "57:28:34", "max_grad_norm": 0.8, "loss": 0.6477128267288208, "grad_norm": 1.082466959953308, "learning_rate": 9.831609336885139e-05}
+{"ts": "2025-12-24T01:02:39", "event": "train_log", "step": 2176, "epoch": 0.9389428263214671, "progress_pct": 15.65, "epoch_pct": 15.65, "eta": "57:26:39", "max_grad_norm": 0.8, "loss": 0.6516029238700867, "grad_norm": 0.9830297827720642, "learning_rate": 9.830976908862346e-05}
+{"ts": "2025-12-24T01:02:58", "event": "train_log", "step": 2178, "epoch": 0.9398058252427185, "progress_pct": 15.66, "epoch_pct": 15.66, "eta": "57:24:40", "max_grad_norm": 0.8, "loss": 0.6793080568313599, "grad_norm": 1.112289309501648, "learning_rate": 9.83034331587125e-05}
+{"ts": "2025-12-24T01:03:16", "event": "train_log", "step": 2180, "epoch": 0.9406688241639698, "progress_pct": 15.67, "epoch_pct": 15.68, "eta": "57:22:31", "max_grad_norm": 0.8, "loss": 0.7220560312271118, "grad_norm": 1.1212154626846313, "learning_rate": 9.829708558064638e-05}
+{"ts": "2025-12-24T01:03:34", "event": "train_log", "step": 2182, "epoch": 0.9415318230852211, "progress_pct": 15.69, "epoch_pct": 15.69, "eta": "57:20:23", "max_grad_norm": 0.8, "loss": 0.7621577978134155, "grad_norm": 1.246471643447876, "learning_rate": 9.829072635595582e-05}
+{"ts": "2025-12-24T01:03:53", "event": "train_log", "step": 2184, "epoch": 0.9423948220064725, "progress_pct": 15.7, "epoch_pct": 15.71, "eta": "57:18:20", "max_grad_norm": 0.8, "loss": 0.6988050937652588, "grad_norm": 1.1732288599014282, "learning_rate": 9.828435548617428e-05}
+{"ts": "2025-12-24T01:04:11", "event": "train_log", "step": 2186, "epoch": 0.9432578209277238, "progress_pct": 15.72, "epoch_pct": 15.72, "eta": "57:16:13", "max_grad_norm": 0.8, "loss": 0.7909345626831055, "grad_norm": 1.1767359972000122, "learning_rate": 9.827797297283808e-05}
+{"ts": "2025-12-24T01:04:30", "event": "train_log", "step": 2188, "epoch": 0.9441208198489752, "progress_pct": 15.73, "epoch_pct": 15.74, "eta": "57:14:14", "max_grad_norm": 0.8, "loss": 0.6651493310928345, "grad_norm": 1.007215976715088, "learning_rate": 9.827157881748633e-05}
+{"ts": "2025-12-24T01:04:49", "event": "train_log", "step": 2190, "epoch": 0.9449838187702265, "progress_pct": 15.75, "epoch_pct": 15.75, "eta": "57:12:08", "max_grad_norm": 0.8, "loss": 0.7742198705673218, "grad_norm": 1.184118628501892, "learning_rate": 9.826517302166096e-05}
+{"ts": "2025-12-24T01:05:07", "event": "train_log", "step": 2192, "epoch": 0.9458468176914779, "progress_pct": 15.76, "epoch_pct": 15.76, "eta": "57:10:04", "max_grad_norm": 0.8, "loss": 0.7614272236824036, "grad_norm": 1.1227030754089355, "learning_rate": 9.82587555869067e-05}
+{"ts": "2025-12-24T01:05:26", "event": "train_log", "step": 2194, "epoch": 0.9467098166127292, "progress_pct": 15.78, "epoch_pct": 15.78, "eta": "57:08:01", "max_grad_norm": 0.8, "loss": 0.7302108407020569, "grad_norm": 1.1030738353729248, "learning_rate": 9.825232651477109e-05}
+{"ts": "2025-12-24T01:05:43", "event": "train_log", "step": 2196, "epoch": 0.9475728155339805, "progress_pct": 15.79, "epoch_pct": 15.79, "eta": "57:05:52", "max_grad_norm": 0.8, "loss": 0.7408936023712158, "grad_norm": 1.5209563970565796, "learning_rate": 9.824588580680445e-05}
+{"ts": "2025-12-24T01:06:01", "event": "train_log", "step": 2198, "epoch": 0.948435814455232, "progress_pct": 15.8, "epoch_pct": 15.81, "eta": "57:03:45", "max_grad_norm": 0.8, "loss": 0.8050196766853333, "grad_norm": 1.1853431463241577, "learning_rate": 9.823943346455995e-05}
+{"ts": "2025-12-24T01:06:21", "event": "train_log", "step": 2200, "epoch": 0.9492988133764833, "progress_pct": 15.82, "epoch_pct": 15.82, "eta": "57:01:48", "max_grad_norm": 0.8, "loss": 0.726731538772583, "grad_norm": 1.285551905632019, "learning_rate": 9.823296948959354e-05}
+{"ts": "2025-12-24T01:20:20", "event": "train_log", "step": 2200, "epoch": 0.9492988133764833, "progress_pct": 15.82, "epoch_pct": 15.82, "eta": "58:16:12", "max_grad_norm": 0.8, "eval_loss": 0.7400083541870117, "eval_runtime": 838.7633, "eval_samples_per_second": 2.456, "eval_steps_per_second": 2.456}
+{"ts": "2025-12-24T01:20:39", "event": "train_log", "step": 2202, "epoch": 0.9501618122977347, "progress_pct": 15.83, "epoch_pct": 15.84, "eta": "58:14:10", "max_grad_norm": 0.8, "loss": 0.7061861753463745, "grad_norm": 1.2159804105758667, "learning_rate": 9.822649388346396e-05}
+{"ts": "2025-12-24T01:20:58", "event": "train_log", "step": 2204, "epoch": 0.951024811218986, "progress_pct": 15.85, "epoch_pct": 15.85, "eta": "58:12:01", "max_grad_norm": 0.8, "loss": 0.824293851852417, "grad_norm": 1.1189815998077393, "learning_rate": 9.822000664773278e-05}
+{"ts": "2025-12-24T01:21:17", "event": "train_log", "step": 2206, "epoch": 0.9518878101402374, "progress_pct": 15.86, "epoch_pct": 15.86, "eta": "58:09:59", "max_grad_norm": 0.8, "loss": 0.7230689525604248, "grad_norm": 0.9968506097793579, "learning_rate": 9.821350778396439e-05}
+{"ts": "2025-12-24T01:21:36", "event": "train_log", "step": 2208, "epoch": 0.9527508090614887, "progress_pct": 15.88, "epoch_pct": 15.88, "eta": "58:07:55", "max_grad_norm": 0.8, "loss": 0.6633790135383606, "grad_norm": 1.101393222808838, "learning_rate": 9.820699729372593e-05}
+{"ts": "2025-12-24T01:21:56", "event": "train_log", "step": 2210, "epoch": 0.95361380798274, "progress_pct": 15.89, "epoch_pct": 15.89, "eta": "58:05:56", "max_grad_norm": 0.8, "loss": 0.6548227071762085, "grad_norm": 1.084307312965393, "learning_rate": 9.820047517858741e-05}
+{"ts": "2025-12-24T01:22:15", "event": "train_log", "step": 2212, "epoch": 0.9544768069039914, "progress_pct": 15.9, "epoch_pct": 15.91, "eta": "58:03:51", "max_grad_norm": 0.8, "loss": 0.6434295177459717, "grad_norm": 1.1291496753692627, "learning_rate": 9.819394144012158e-05}
+{"ts": "2025-12-24T01:22:34", "event": "train_log", "step": 2214, "epoch": 0.9553398058252427, "progress_pct": 15.92, "epoch_pct": 15.92, "eta": "58:01:46", "max_grad_norm": 0.8, "loss": 0.8122013211250305, "grad_norm": 2.303483724594116, "learning_rate": 9.818739607990404e-05}
+{"ts": "2025-12-24T01:22:53", "event": "train_log", "step": 2216, "epoch": 0.9562028047464941, "progress_pct": 15.93, "epoch_pct": 15.94, "eta": "57:59:42", "max_grad_norm": 0.8, "loss": 0.6120955348014832, "grad_norm": 1.244541883468628, "learning_rate": 9.818083909951315e-05}
+{"ts": "2025-12-24T01:23:12", "event": "train_log", "step": 2218, "epoch": 0.9570658036677454, "progress_pct": 15.95, "epoch_pct": 15.95, "eta": "57:57:37", "max_grad_norm": 0.8, "loss": 0.7569628953933716, "grad_norm": 1.1051502227783203, "learning_rate": 9.817427050053012e-05}
+{"ts": "2025-12-24T01:23:31", "event": "train_log", "step": 2220, "epoch": 0.9579288025889967, "progress_pct": 15.96, "epoch_pct": 15.97, "eta": "57:55:34", "max_grad_norm": 0.8, "loss": 0.7588980197906494, "grad_norm": 1.3161911964416504, "learning_rate": 9.816769028453893e-05}
+{"ts": "2025-12-24T01:23:49", "event": "train_log", "step": 2222, "epoch": 0.9587918015102481, "progress_pct": 15.98, "epoch_pct": 15.98, "eta": "57:53:24", "max_grad_norm": 0.8, "loss": 0.743370771408081, "grad_norm": 1.2526928186416626, "learning_rate": 9.816109845312636e-05}
+{"ts": "2025-12-24T01:24:07", "event": "train_log", "step": 2224, "epoch": 0.9596548004314994, "progress_pct": 15.99, "epoch_pct": 15.99, "eta": "57:51:19", "max_grad_norm": 0.8, "loss": 0.80516517162323, "grad_norm": 1.196215033531189, "learning_rate": 9.815449500788202e-05}
+{"ts": "2025-12-24T01:24:25", "event": "train_log", "step": 2226, "epoch": 0.9605177993527508, "progress_pct": 16.01, "epoch_pct": 16.01, "eta": "57:49:10", "max_grad_norm": 0.8, "loss": 0.7331632971763611, "grad_norm": 1.3204832077026367, "learning_rate": 9.81478799503983e-05}
+{"ts": "2025-12-24T01:24:44", "event": "train_log", "step": 2228, "epoch": 0.9613807982740021, "progress_pct": 16.02, "epoch_pct": 16.02, "eta": "57:47:07", "max_grad_norm": 0.8, "loss": 0.6810339093208313, "grad_norm": 1.2442790269851685, "learning_rate": 9.814125328227035e-05}
+{"ts": "2025-12-24T01:25:03", "event": "train_log", "step": 2230, "epoch": 0.9622437971952535, "progress_pct": 16.03, "epoch_pct": 16.04, "eta": "57:45:03", "max_grad_norm": 0.8, "loss": 0.7269608378410339, "grad_norm": 1.1739158630371094, "learning_rate": 9.813461500509623e-05}
+{"ts": "2025-12-24T01:25:22", "event": "train_log", "step": 2232, "epoch": 0.9631067961165048, "progress_pct": 16.05, "epoch_pct": 16.05, "eta": "57:43:00", "max_grad_norm": 0.8, "loss": 0.7159227132797241, "grad_norm": 1.1016241312026978, "learning_rate": 9.812796512047666e-05}
+{"ts": "2025-12-24T01:25:42", "event": "train_log", "step": 2234, "epoch": 0.9639697950377561, "progress_pct": 16.06, "epoch_pct": 16.07, "eta": "57:41:02", "max_grad_norm": 0.8, "loss": 0.7145758867263794, "grad_norm": 1.1654549837112427, "learning_rate": 9.812130363001529e-05}
+{"ts": "2025-12-24T01:26:02", "event": "train_log", "step": 2236, "epoch": 0.9648327939590076, "progress_pct": 16.08, "epoch_pct": 16.08, "eta": "57:39:05", "max_grad_norm": 0.8, "loss": 0.6873835325241089, "grad_norm": 1.0521938800811768, "learning_rate": 9.811463053531846e-05}
+{"ts": "2025-12-24T01:26:20", "event": "train_log", "step": 2238, "epoch": 0.9656957928802589, "progress_pct": 16.09, "epoch_pct": 16.09, "eta": "57:37:00", "max_grad_norm": 0.8, "loss": 0.6683027148246765, "grad_norm": 1.2285776138305664, "learning_rate": 9.810794583799538e-05}
+{"ts": "2025-12-24T01:26:39", "event": "train_log", "step": 2240, "epoch": 0.9665587918015103, "progress_pct": 16.11, "epoch_pct": 16.11, "eta": "57:34:58", "max_grad_norm": 0.8, "loss": 0.663141131401062, "grad_norm": 0.9826368689537048, "learning_rate": 9.810124953965805e-05}
+{"ts": "2025-12-24T01:26:59", "event": "train_log", "step": 2242, "epoch": 0.9674217907227616, "progress_pct": 16.12, "epoch_pct": 16.12, "eta": "57:33:02", "max_grad_norm": 0.8, "loss": 0.6688778400421143, "grad_norm": 1.0102336406707764, "learning_rate": 9.809454164192123e-05}
+{"ts": "2025-12-24T01:27:19", "event": "train_log", "step": 2244, "epoch": 0.968284789644013, "progress_pct": 16.13, "epoch_pct": 16.14, "eta": "57:31:08", "max_grad_norm": 0.8, "loss": 0.671240508556366, "grad_norm": 0.9781575202941895, "learning_rate": 9.808782214640251e-05}
+{"ts": "2025-12-24T01:27:38", "event": "train_log", "step": 2246, "epoch": 0.9691477885652643, "progress_pct": 16.15, "epoch_pct": 16.15, "eta": "57:29:07", "max_grad_norm": 0.8, "loss": 0.7673252820968628, "grad_norm": 1.137385368347168, "learning_rate": 9.808109105472225e-05}
+{"ts": "2025-12-24T01:27:57", "event": "train_log", "step": 2248, "epoch": 0.9700107874865156, "progress_pct": 16.16, "epoch_pct": 16.17, "eta": "57:27:02", "max_grad_norm": 0.8, "loss": 0.8085554838180542, "grad_norm": 1.3359450101852417, "learning_rate": 9.807434836850362e-05}
+{"ts": "2025-12-24T01:28:16", "event": "train_log", "step": 2250, "epoch": 0.970873786407767, "progress_pct": 16.18, "epoch_pct": 16.18, "eta": "57:25:02", "max_grad_norm": 0.8, "loss": 0.6750202178955078, "grad_norm": 1.150694727897644, "learning_rate": 9.806759408937262e-05}
+{"ts": "2025-12-24T01:28:36", "event": "train_log", "step": 2252, "epoch": 0.9717367853290183, "progress_pct": 16.19, "epoch_pct": 16.2, "eta": "57:23:05", "max_grad_norm": 0.8, "loss": 0.5928418040275574, "grad_norm": 1.080316424369812, "learning_rate": 9.8060828218958e-05}
+{"ts": "2025-12-24T01:28:54", "event": "train_log", "step": 2254, "epoch": 0.9725997842502697, "progress_pct": 16.21, "epoch_pct": 16.21, "eta": "57:21:03", "max_grad_norm": 0.8, "loss": 0.6947471499443054, "grad_norm": 1.0489610433578491, "learning_rate": 9.805405075889129e-05}
+{"ts": "2025-12-24T01:29:14", "event": "train_log", "step": 2256, "epoch": 0.973462783171521, "progress_pct": 16.22, "epoch_pct": 16.22, "eta": "57:19:08", "max_grad_norm": 0.8, "loss": 0.7273924350738525, "grad_norm": 1.199119210243225, "learning_rate": 9.80472617108069e-05}
+{"ts": "2025-12-24T01:29:34", "event": "train_log", "step": 2258, "epoch": 0.9743257820927724, "progress_pct": 16.24, "epoch_pct": 16.24, "eta": "57:17:11", "max_grad_norm": 0.8, "loss": 0.668613612651825, "grad_norm": 0.9830896258354187, "learning_rate": 9.804046107634193e-05}
+{"ts": "2025-12-24T01:29:55", "event": "train_log", "step": 2260, "epoch": 0.9751887810140237, "progress_pct": 16.25, "epoch_pct": 16.25, "eta": "57:15:22", "max_grad_norm": 0.8, "loss": 0.7255772352218628, "grad_norm": 1.1655877828598022, "learning_rate": 9.803364885713633e-05}
+{"ts": "2025-12-24T01:30:17", "event": "train_log", "step": 2262, "epoch": 0.976051779935275, "progress_pct": 16.26, "epoch_pct": 16.27, "eta": "57:13:37", "max_grad_norm": 0.8, "loss": 0.6865858435630798, "grad_norm": 1.0289462804794312, "learning_rate": 9.802682505483284e-05}
+{"ts": "2025-12-24T01:30:37", "event": "train_log", "step": 2264, "epoch": 0.9769147788565264, "progress_pct": 16.28, "epoch_pct": 16.28, "eta": "57:11:44", "max_grad_norm": 0.8, "loss": 0.6680663228034973, "grad_norm": 1.1020126342773438, "learning_rate": 9.801998967107702e-05}
+{"ts": "2025-12-24T01:30:57", "event": "train_log", "step": 2266, "epoch": 0.9777777777777777, "progress_pct": 16.29, "epoch_pct": 16.3, "eta": "57:09:48", "max_grad_norm": 0.8, "loss": 0.7476468086242676, "grad_norm": 1.0846027135849, "learning_rate": 9.801314270751716e-05}
+{"ts": "2025-12-24T01:31:16", "event": "train_log", "step": 2268, "epoch": 0.9786407766990292, "progress_pct": 16.31, "epoch_pct": 16.31, "eta": "57:07:48", "max_grad_norm": 0.8, "loss": 0.7173228859901428, "grad_norm": 1.1834278106689453, "learning_rate": 9.800628416580437e-05}
+{"ts": "2025-12-24T01:31:38", "event": "train_log", "step": 2270, "epoch": 0.9795037756202805, "progress_pct": 16.32, "epoch_pct": 16.33, "eta": "57:06:03", "max_grad_norm": 0.8, "loss": 0.6768618226051331, "grad_norm": 1.0082321166992188, "learning_rate": 9.799941404759257e-05}
+{"ts": "2025-12-24T01:31:56", "event": "train_log", "step": 2272, "epoch": 0.9803667745415319, "progress_pct": 16.34, "epoch_pct": 16.34, "eta": "57:04:04", "max_grad_norm": 0.8, "loss": 0.7877311110496521, "grad_norm": 1.1573102474212646, "learning_rate": 9.799253235453846e-05}
+{"ts": "2025-12-24T01:32:16", "event": "train_log", "step": 2274, "epoch": 0.9812297734627832, "progress_pct": 16.35, "epoch_pct": 16.35, "eta": "57:02:07", "max_grad_norm": 0.8, "loss": 0.6843966841697693, "grad_norm": 1.2030092477798462, "learning_rate": 9.798563908830152e-05}
+{"ts": "2025-12-24T01:32:37", "event": "train_log", "step": 2276, "epoch": 0.9820927723840345, "progress_pct": 16.36, "epoch_pct": 16.37, "eta": "57:00:21", "max_grad_norm": 0.8, "loss": 0.6084349155426025, "grad_norm": 1.080560564994812, "learning_rate": 9.797873425054404e-05}
+{"ts": "2025-12-24T01:32:56", "event": "train_log", "step": 2278, "epoch": 0.9829557713052859, "progress_pct": 16.38, "epoch_pct": 16.38, "eta": "56:58:21", "max_grad_norm": 0.8, "loss": 0.6784868836402893, "grad_norm": 1.1488033533096313, "learning_rate": 9.797181784293107e-05}
+{"ts": "2025-12-24T01:33:16", "event": "train_log", "step": 2280, "epoch": 0.9838187702265372, "progress_pct": 16.39, "epoch_pct": 16.4, "eta": "56:56:29", "max_grad_norm": 0.8, "loss": 0.6111500263214111, "grad_norm": 1.144773006439209, "learning_rate": 9.796488986713049e-05}
+{"ts": "2025-12-24T01:33:37", "event": "train_log", "step": 2282, "epoch": 0.9846817691477886, "progress_pct": 16.41, "epoch_pct": 16.41, "eta": "56:54:40", "max_grad_norm": 0.8, "loss": 0.7634603977203369, "grad_norm": 1.1356254816055298, "learning_rate": 9.795795032481297e-05}
+{"ts": "2025-12-24T01:33:57", "event": "train_log", "step": 2284, "epoch": 0.9855447680690399, "progress_pct": 16.42, "epoch_pct": 16.43, "eta": "56:52:46", "max_grad_norm": 0.8, "loss": 0.6470903158187866, "grad_norm": 1.139446496963501, "learning_rate": 9.79509992176519e-05}
+{"ts": "2025-12-24T01:34:15", "event": "train_log", "step": 2286, "epoch": 0.9864077669902913, "progress_pct": 16.44, "epoch_pct": 16.44, "eta": "56:50:46", "max_grad_norm": 0.8, "loss": 0.71474289894104, "grad_norm": 1.1564133167266846, "learning_rate": 9.794403654732354e-05}
+{"ts": "2025-12-24T01:34:35", "event": "train_log", "step": 2288, "epoch": 0.9872707659115426, "progress_pct": 16.45, "epoch_pct": 16.45, "eta": "56:48:50", "max_grad_norm": 0.8, "loss": 0.6413968801498413, "grad_norm": 1.0605655908584595, "learning_rate": 9.793706231550689e-05}
+{"ts": "2025-12-24T01:34:53", "event": "train_log", "step": 2290, "epoch": 0.9881337648327939, "progress_pct": 16.47, "epoch_pct": 16.47, "eta": "56:46:51", "max_grad_norm": 0.8, "loss": 0.6595705151557922, "grad_norm": 1.1660863161087036, "learning_rate": 9.793007652388378e-05}
+{"ts": "2025-12-24T01:35:13", "event": "train_log", "step": 2292, "epoch": 0.9889967637540453, "progress_pct": 16.48, "epoch_pct": 16.48, "eta": "56:44:58", "max_grad_norm": 0.8, "loss": 0.7144066095352173, "grad_norm": 1.1661914587020874, "learning_rate": 9.792307917413877e-05}
+{"ts": "2025-12-24T01:35:33", "event": "train_log", "step": 2294, "epoch": 0.9898597626752966, "progress_pct": 16.49, "epoch_pct": 16.5, "eta": "56:43:02", "max_grad_norm": 0.8, "loss": 0.7956018447875977, "grad_norm": 1.2250727415084839, "learning_rate": 9.791607026795926e-05}
+{"ts": "2025-12-24T01:35:54", "event": "train_log", "step": 2296, "epoch": 0.990722761596548, "progress_pct": 16.51, "epoch_pct": 16.51, "eta": "56:41:20", "max_grad_norm": 0.8, "loss": 0.5786031484603882, "grad_norm": 0.9593650698661804, "learning_rate": 9.790904980703541e-05}
+{"ts": "2025-12-24T01:36:15", "event": "train_log", "step": 2298, "epoch": 0.9915857605177993, "progress_pct": 16.52, "epoch_pct": 16.53, "eta": "56:39:31", "max_grad_norm": 0.8, "loss": 0.6817460656166077, "grad_norm": 1.0697777271270752, "learning_rate": 9.790201779306018e-05}
+{"ts": "2025-12-24T01:36:35", "event": "train_log", "step": 2300, "epoch": 0.9924487594390508, "progress_pct": 16.54, "epoch_pct": 16.54, "eta": "56:37:38", "max_grad_norm": 0.8, "loss": 0.6845148801803589, "grad_norm": 1.2658408880233765, "learning_rate": 9.789497422772931e-05}
+{"ts": "2025-12-24T01:50:21", "event": "train_log", "step": 2300, "epoch": 0.9924487594390508, "progress_pct": 16.54, "epoch_pct": 16.54, "eta": "57:47:06", "max_grad_norm": 0.8, "eval_loss": 0.7360134720802307, "eval_runtime": 825.8559, "eval_samples_per_second": 2.494, "eval_steps_per_second": 2.494}
+{"ts": "2025-12-24T01:50:40", "event": "train_log", "step": 2302, "epoch": 0.993311758360302, "progress_pct": 16.55, "epoch_pct": 16.56, "eta": "57:45:08", "max_grad_norm": 0.8, "loss": 0.6689707040786743, "grad_norm": 1.0869544744491577, "learning_rate": 9.788791911274131e-05}
+{"ts": "2025-12-24T01:51:00", "event": "train_log", "step": 2304, "epoch": 0.9941747572815534, "progress_pct": 16.57, "epoch_pct": 16.57, "eta": "57:43:11", "max_grad_norm": 0.8, "loss": 0.6470839381217957, "grad_norm": 1.0049875974655151, "learning_rate": 9.788085244979748e-05}
+{"ts": "2025-12-24T01:51:18", "event": "train_log", "step": 2306, "epoch": 0.9950377562028048, "progress_pct": 16.58, "epoch_pct": 16.58, "eta": "57:41:07", "max_grad_norm": 0.8, "loss": 0.8018016815185547, "grad_norm": 1.4471046924591064, "learning_rate": 9.787377424060196e-05}
+{"ts": "2025-12-24T01:51:35", "event": "train_log", "step": 2308, "epoch": 0.9959007551240561, "progress_pct": 16.59, "epoch_pct": 16.6, "eta": "57:38:56", "max_grad_norm": 0.8, "loss": 0.8078322410583496, "grad_norm": 1.3750648498535156, "learning_rate": 9.786668448686159e-05}
+{"ts": "2025-12-24T01:51:54", "event": "train_log", "step": 2310, "epoch": 0.9967637540453075, "progress_pct": 16.61, "epoch_pct": 16.61, "eta": "57:36:55", "max_grad_norm": 0.8, "loss": 0.6740908622741699, "grad_norm": 1.098283290863037, "learning_rate": 9.7859583190286e-05}
+{"ts": "2025-12-24T01:52:13", "event": "train_log", "step": 2312, "epoch": 0.9976267529665588, "progress_pct": 16.62, "epoch_pct": 16.63, "eta": "57:34:59", "max_grad_norm": 0.8, "loss": 0.6844386458396912, "grad_norm": 1.0312142372131348, "learning_rate": 9.785247035258773e-05}
+{"ts": "2025-12-24T01:52:32", "event": "train_log", "step": 2314, "epoch": 0.9984897518878102, "progress_pct": 16.64, "epoch_pct": 16.64, "eta": "57:32:59", "max_grad_norm": 0.8, "loss": 0.7547479271888733, "grad_norm": 1.4482319355010986, "learning_rate": 9.784534597548193e-05}
+{"ts": "2025-12-24T01:52:51", "event": "train_log", "step": 2316, "epoch": 0.9993527508090615, "progress_pct": 16.65, "epoch_pct": 16.66, "eta": "57:30:59", "max_grad_norm": 0.8, "loss": 0.7713987231254578, "grad_norm": 1.0897735357284546, "learning_rate": 9.783821006068663e-05}
+{"ts": "2025-12-24T01:53:06", "event": "train_log", "step": 2318, "epoch": 1.0, "progress_pct": 16.67, "epoch_pct": 16.67, "eta": "57:28:37", "max_grad_norm": 0.8, "loss": 0.6868399381637573, "grad_norm": 1.7887805700302124, "learning_rate": 9.783106260992264e-05}
+{"ts": "2025-12-24T01:53:25", "event": "train_log", "step": 2320, "epoch": 1.0008629989212514, "progress_pct": 16.68, "epoch_pct": 16.68, "eta": "57:26:41", "max_grad_norm": 0.8, "loss": 0.696355938911438, "grad_norm": 1.0021663904190063, "learning_rate": 9.782390362491353e-05}
+{"ts": "2025-12-24T01:53:45", "event": "train_log", "step": 2322, "epoch": 1.0017259978425026, "progress_pct": 16.7, "epoch_pct": 16.7, "eta": "57:24:44", "max_grad_norm": 0.8, "loss": 0.7315902709960938, "grad_norm": 1.3650091886520386, "learning_rate": 9.781673310738564e-05}
+{"ts": "2025-12-24T01:54:05", "event": "train_log", "step": 2324, "epoch": 1.002588996763754, "progress_pct": 16.71, "epoch_pct": 16.71, "eta": "57:22:49", "max_grad_norm": 0.8, "loss": 0.7667547464370728, "grad_norm": 1.2247875928878784, "learning_rate": 9.780955105906813e-05}
+{"ts": "2025-12-24T01:54:24", "event": "train_log", "step": 2326, "epoch": 1.0034519956850054, "progress_pct": 16.72, "epoch_pct": 16.72, "eta": "57:20:53", "max_grad_norm": 0.8, "loss": 0.7380582094192505, "grad_norm": 1.1917641162872314, "learning_rate": 9.780235748169291e-05}
+{"ts": "2025-12-24T01:54:45", "event": "train_log", "step": 2328, "epoch": 1.0043149946062568, "progress_pct": 16.74, "epoch_pct": 16.74, "eta": "57:19:03", "max_grad_norm": 0.8, "loss": 0.6780733466148376, "grad_norm": 1.146159052848816, "learning_rate": 9.779515237699466e-05}
+{"ts": "2025-12-24T01:55:04", "event": "train_log", "step": 2330, "epoch": 1.005177993527508, "progress_pct": 16.75, "epoch_pct": 16.75, "eta": "57:17:05", "max_grad_norm": 0.8, "loss": 0.8116409182548523, "grad_norm": 1.4877620935440063, "learning_rate": 9.778793574671089e-05}
+{"ts": "2025-12-24T01:55:24", "event": "train_log", "step": 2332, "epoch": 1.0060409924487594, "progress_pct": 16.77, "epoch_pct": 16.77, "eta": "57:15:14", "max_grad_norm": 0.8, "loss": 0.65245521068573, "grad_norm": 1.3488554954528809, "learning_rate": 9.778070759258185e-05}
+{"ts": "2025-12-24T01:55:44", "event": "train_log", "step": 2334, "epoch": 1.0069039913700109, "progress_pct": 16.78, "epoch_pct": 16.78, "eta": "57:13:20", "max_grad_norm": 0.8, "loss": 0.584334135055542, "grad_norm": 1.096686840057373, "learning_rate": 9.777346791635056e-05}
+{"ts": "2025-12-24T01:56:05", "event": "train_log", "step": 2336, "epoch": 1.007766990291262, "progress_pct": 16.8, "epoch_pct": 16.8, "eta": "57:11:31", "max_grad_norm": 0.8, "loss": 0.6594714522361755, "grad_norm": 1.063981056213379, "learning_rate": 9.776621671976285e-05}
+{"ts": "2025-12-24T01:56:25", "event": "train_log", "step": 2338, "epoch": 1.0086299892125135, "progress_pct": 16.81, "epoch_pct": 16.81, "eta": "57:09:41", "max_grad_norm": 0.8, "loss": 0.6963099837303162, "grad_norm": 1.2811797857284546, "learning_rate": 9.775895400456729e-05}
+{"ts": "2025-12-24T01:56:45", "event": "train_log", "step": 2340, "epoch": 1.0094929881337649, "progress_pct": 16.82, "epoch_pct": 16.82, "eta": "57:07:48", "max_grad_norm": 0.8, "loss": 0.6246375441551208, "grad_norm": 0.9491971731185913, "learning_rate": 9.775167977251529e-05}
+{"ts": "2025-12-24T01:57:05", "event": "train_log", "step": 2342, "epoch": 1.0103559870550163, "progress_pct": 16.84, "epoch_pct": 16.84, "eta": "57:05:54", "max_grad_norm": 0.8, "loss": 0.6352161169052124, "grad_norm": 1.198683500289917, "learning_rate": 9.774439402536098e-05}
+{"ts": "2025-12-24T01:57:26", "event": "train_log", "step": 2344, "epoch": 1.0112189859762675, "progress_pct": 16.85, "epoch_pct": 16.85, "eta": "57:04:06", "max_grad_norm": 0.8, "loss": 0.6884894967079163, "grad_norm": 1.2178813219070435, "learning_rate": 9.773709676486129e-05}
+{"ts": "2025-12-24T01:57:47", "event": "train_log", "step": 2346, "epoch": 1.0120819848975189, "progress_pct": 16.87, "epoch_pct": 16.87, "eta": "57:02:20", "max_grad_norm": 0.8, "loss": 0.5945048928260803, "grad_norm": 1.1905319690704346, "learning_rate": 9.77297879927759e-05}
+{"ts": "2025-12-24T01:58:08", "event": "train_log", "step": 2348, "epoch": 1.0129449838187703, "progress_pct": 16.88, "epoch_pct": 16.88, "eta": "57:00:30", "max_grad_norm": 0.8, "loss": 0.6533923149108887, "grad_norm": 1.210302710533142, "learning_rate": 9.77224677108673e-05}
+{"ts": "2025-12-24T01:58:26", "event": "train_log", "step": 2350, "epoch": 1.0138079827400215, "progress_pct": 16.9, "epoch_pct": 16.9, "eta": "56:58:30", "max_grad_norm": 0.8, "loss": 0.6619405746459961, "grad_norm": 1.15508234500885, "learning_rate": 9.771513592090076e-05}
+{"ts": "2025-12-24T01:58:46", "event": "train_log", "step": 2352, "epoch": 1.014670981661273, "progress_pct": 16.91, "epoch_pct": 16.91, "eta": "56:56:38", "max_grad_norm": 0.8, "loss": 0.6581310033798218, "grad_norm": 1.0833762884140015, "learning_rate": 9.77077926246443e-05}
+{"ts": "2025-12-24T01:59:06", "event": "train_log", "step": 2354, "epoch": 1.0155339805825243, "progress_pct": 16.93, "epoch_pct": 16.93, "eta": "56:54:47", "max_grad_norm": 0.8, "loss": 0.6670945882797241, "grad_norm": 1.0330514907836914, "learning_rate": 9.770043782386869e-05}
+{"ts": "2025-12-24T01:59:26", "event": "train_log", "step": 2356, "epoch": 1.0163969795037757, "progress_pct": 16.94, "epoch_pct": 16.94, "eta": "56:52:56", "max_grad_norm": 0.8, "loss": 0.6801814436912537, "grad_norm": 1.1472195386886597, "learning_rate": 9.769307152034755e-05}
+{"ts": "2025-12-24T01:59:43", "event": "train_log", "step": 2358, "epoch": 1.017259978425027, "progress_pct": 16.95, "epoch_pct": 16.95, "eta": "56:50:51", "max_grad_norm": 0.8, "loss": 0.7206465005874634, "grad_norm": 1.2204132080078125, "learning_rate": 9.76856937158572e-05}
+{"ts": "2025-12-24T02:00:00", "event": "train_log", "step": 2360, "epoch": 1.0181229773462783, "progress_pct": 16.97, "epoch_pct": 16.97, "eta": "56:48:47", "max_grad_norm": 0.8, "loss": 0.6884574294090271, "grad_norm": 1.2335962057113647, "learning_rate": 9.76783044121768e-05}
+{"ts": "2025-12-24T02:00:20", "event": "train_log", "step": 2362, "epoch": 1.0189859762675297, "progress_pct": 16.98, "epoch_pct": 16.98, "eta": "56:46:54", "max_grad_norm": 0.8, "loss": 0.655097484588623, "grad_norm": 1.0858685970306396, "learning_rate": 9.767090361108822e-05}
+{"ts": "2025-12-24T02:00:41", "event": "train_log", "step": 2364, "epoch": 1.019848975188781, "progress_pct": 17.0, "epoch_pct": 17.0, "eta": "56:45:06", "max_grad_norm": 0.8, "loss": 0.6478561758995056, "grad_norm": 1.117907166481018, "learning_rate": 9.766349131437614e-05}
+{"ts": "2025-12-24T02:01:00", "event": "train_log", "step": 2366, "epoch": 1.0207119741100323, "progress_pct": 17.01, "epoch_pct": 17.01, "eta": "56:43:13", "max_grad_norm": 0.8, "loss": 0.6310951113700867, "grad_norm": 1.1995056867599487, "learning_rate": 9.765606752382797e-05}
+{"ts": "2025-12-24T02:01:19", "event": "train_log", "step": 2368, "epoch": 1.0215749730312838, "progress_pct": 17.03, "epoch_pct": 17.03, "eta": "56:41:17", "max_grad_norm": 0.8, "loss": 0.6373490691184998, "grad_norm": 1.354819655418396, "learning_rate": 9.764863224123397e-05}
+{"ts": "2025-12-24T02:01:39", "event": "train_log", "step": 2370, "epoch": 1.0224379719525352, "progress_pct": 17.04, "epoch_pct": 17.04, "eta": "56:39:26", "max_grad_norm": 0.8, "loss": 0.6595761179924011, "grad_norm": 1.2194819450378418, "learning_rate": 9.764118546838712e-05}
+{"ts": "2025-12-24T02:01:58", "event": "train_log", "step": 2372, "epoch": 1.0233009708737864, "progress_pct": 17.05, "epoch_pct": 17.06, "eta": "56:37:34", "max_grad_norm": 0.8, "loss": 0.6299480199813843, "grad_norm": 1.1485158205032349, "learning_rate": 9.763372720708314e-05}
+{"ts": "2025-12-24T02:02:17", "event": "train_log", "step": 2374, "epoch": 1.0241639697950378, "progress_pct": 17.07, "epoch_pct": 17.07, "eta": "56:35:38", "max_grad_norm": 0.8, "loss": 0.6788742542266846, "grad_norm": 1.1153662204742432, "learning_rate": 9.76262574591206e-05}
+{"ts": "2025-12-24T02:02:37", "event": "train_log", "step": 2376, "epoch": 1.0250269687162892, "progress_pct": 17.08, "epoch_pct": 17.08, "eta": "56:33:51", "max_grad_norm": 0.8, "loss": 0.559056282043457, "grad_norm": 1.0690817832946777, "learning_rate": 9.761877622630075e-05}
+{"ts": "2025-12-24T02:02:56", "event": "train_log", "step": 2378, "epoch": 1.0258899676375404, "progress_pct": 17.1, "epoch_pct": 17.1, "eta": "56:31:53", "max_grad_norm": 0.8, "loss": 0.7164555191993713, "grad_norm": 1.1896858215332031, "learning_rate": 9.76112835104277e-05}
+{"ts": "2025-12-24T02:03:16", "event": "train_log", "step": 2380, "epoch": 1.0267529665587918, "progress_pct": 17.11, "epoch_pct": 17.11, "eta": "56:30:03", "max_grad_norm": 0.8, "loss": 0.6791456937789917, "grad_norm": 1.1792134046554565, "learning_rate": 9.760377931330824e-05}
+{"ts": "2025-12-24T02:03:34", "event": "train_log", "step": 2382, "epoch": 1.0276159654800432, "progress_pct": 17.13, "epoch_pct": 17.13, "eta": "56:28:07", "max_grad_norm": 0.8, "loss": 0.7421526312828064, "grad_norm": 1.361763834953308, "learning_rate": 9.759626363675202e-05}
+{"ts": "2025-12-24T02:03:54", "event": "train_log", "step": 2384, "epoch": 1.0284789644012946, "progress_pct": 17.14, "epoch_pct": 17.14, "eta": "56:26:18", "max_grad_norm": 0.8, "loss": 0.7032517194747925, "grad_norm": 1.1960662603378296, "learning_rate": 9.758873648257138e-05}
+{"ts": "2025-12-24T02:04:15", "event": "train_log", "step": 2386, "epoch": 1.0293419633225458, "progress_pct": 17.16, "epoch_pct": 17.16, "eta": "56:24:35", "max_grad_norm": 0.8, "loss": 0.6953344345092773, "grad_norm": 1.323255181312561, "learning_rate": 9.758119785258146e-05}
+{"ts": "2025-12-24T02:04:35", "event": "train_log", "step": 2388, "epoch": 1.0302049622437972, "progress_pct": 17.17, "epoch_pct": 17.17, "eta": "56:22:43", "max_grad_norm": 0.8, "loss": 0.7287829518318176, "grad_norm": 1.2343937158584595, "learning_rate": 9.75736477486002e-05}
+{"ts": "2025-12-24T02:04:55", "event": "train_log", "step": 2390, "epoch": 1.0310679611650486, "progress_pct": 17.18, "epoch_pct": 17.18, "eta": "56:20:53", "max_grad_norm": 0.8, "loss": 0.6812928915023804, "grad_norm": 1.0712062120437622, "learning_rate": 9.756608617244823e-05}
+{"ts": "2025-12-24T02:05:12", "event": "train_log", "step": 2392, "epoch": 1.0319309600862998, "progress_pct": 17.2, "epoch_pct": 17.2, "eta": "56:18:53", "max_grad_norm": 0.8, "loss": 0.7173963785171509, "grad_norm": 1.264726996421814, "learning_rate": 9.755851312594902e-05}
+{"ts": "2025-12-24T02:05:32", "event": "train_log", "step": 2394, "epoch": 1.0327939590075512, "progress_pct": 17.21, "epoch_pct": 17.21, "eta": "56:17:05", "max_grad_norm": 0.8, "loss": 0.6509836316108704, "grad_norm": 1.1563389301300049, "learning_rate": 9.755092861092874e-05}
+{"ts": "2025-12-24T02:05:51", "event": "train_log", "step": 2396, "epoch": 1.0336569579288026, "progress_pct": 17.23, "epoch_pct": 17.23, "eta": "56:15:08", "max_grad_norm": 0.8, "loss": 0.718437135219574, "grad_norm": 1.326188087463379, "learning_rate": 9.75433326292164e-05}
+{"ts": "2025-12-24T02:06:10", "event": "train_log", "step": 2398, "epoch": 1.0345199568500538, "progress_pct": 17.24, "epoch_pct": 17.24, "eta": "56:13:16", "max_grad_norm": 0.8, "loss": 0.664516031742096, "grad_norm": 1.2263392210006714, "learning_rate": 9.753572518264373e-05}
+{"ts": "2025-12-24T02:06:28", "event": "train_log", "step": 2400, "epoch": 1.0353829557713052, "progress_pct": 17.26, "epoch_pct": 17.26, "eta": "56:11:21", "max_grad_norm": 0.8, "loss": 0.6754546165466309, "grad_norm": 1.2874478101730347, "learning_rate": 9.752810627304523e-05}
+{"ts": "2025-12-24T02:20:14", "event": "train_log", "step": 2400, "epoch": 1.0353829557713052, "progress_pct": 17.26, "epoch_pct": 17.26, "eta": "57:17:21", "max_grad_norm": 0.8, "eval_loss": 0.7346783876419067, "eval_runtime": 825.8356, "eval_samples_per_second": 2.494, "eval_steps_per_second": 2.494}
+{"ts": "2025-12-24T02:20:32", "event": "train_log", "step": 2402, "epoch": 1.0362459546925566, "progress_pct": 17.27, "epoch_pct": 17.27, "eta": "57:15:22", "max_grad_norm": 0.8, "loss": 0.708167314529419, "grad_norm": 1.304518461227417, "learning_rate": 9.752047590225814e-05}
+{"ts": "2025-12-24T02:20:52", "event": "train_log", "step": 2404, "epoch": 1.037108953613808, "progress_pct": 17.29, "epoch_pct": 17.29, "eta": "57:13:26", "max_grad_norm": 0.8, "loss": 0.7213018536567688, "grad_norm": 1.21835196018219, "learning_rate": 9.751283407212252e-05}
+{"ts": "2025-12-24T02:21:11", "event": "train_log", "step": 2406, "epoch": 1.0379719525350593, "progress_pct": 17.3, "epoch_pct": 17.3, "eta": "57:11:30", "max_grad_norm": 0.8, "loss": 0.6739231944084167, "grad_norm": 1.2947579622268677, "learning_rate": 9.750518078448116e-05}
+{"ts": "2025-12-24T02:21:31", "event": "train_log", "step": 2408, "epoch": 1.0388349514563107, "progress_pct": 17.31, "epoch_pct": 17.31, "eta": "57:09:41", "max_grad_norm": 0.8, "loss": 0.7671371102333069, "grad_norm": 1.138368010520935, "learning_rate": 9.749751604117959e-05}
+{"ts": "2025-12-24T02:21:50", "event": "train_log", "step": 2410, "epoch": 1.039697950377562, "progress_pct": 17.33, "epoch_pct": 17.33, "eta": "57:07:46", "max_grad_norm": 0.8, "loss": 0.6593353748321533, "grad_norm": 1.1255316734313965, "learning_rate": 9.748983984406616e-05}
+{"ts": "2025-12-24T02:22:10", "event": "train_log", "step": 2412, "epoch": 1.0405609492988135, "progress_pct": 17.34, "epoch_pct": 17.34, "eta": "57:05:54", "max_grad_norm": 0.8, "loss": 0.713056743144989, "grad_norm": 1.2362794876098633, "learning_rate": 9.748215219499193e-05}
+{"ts": "2025-12-24T02:22:29", "event": "train_log", "step": 2414, "epoch": 1.0414239482200647, "progress_pct": 17.36, "epoch_pct": 17.36, "eta": "57:03:56", "max_grad_norm": 0.8, "loss": 0.6989394426345825, "grad_norm": 1.2021304368972778, "learning_rate": 9.747445309581076e-05}
+{"ts": "2025-12-24T02:22:46", "event": "train_log", "step": 2416, "epoch": 1.042286947141316, "progress_pct": 17.37, "epoch_pct": 17.37, "eta": "57:01:55", "max_grad_norm": 0.8, "loss": 0.623035192489624, "grad_norm": 1.127341389656067, "learning_rate": 9.746674254837925e-05}
+{"ts": "2025-12-24T02:23:07", "event": "train_log", "step": 2418, "epoch": 1.0431499460625675, "progress_pct": 17.39, "epoch_pct": 17.39, "eta": "57:00:10", "max_grad_norm": 0.8, "loss": 0.6490310430526733, "grad_norm": 1.1254255771636963, "learning_rate": 9.745902055455672e-05}
+{"ts": "2025-12-24T02:23:28", "event": "train_log", "step": 2420, "epoch": 1.0440129449838187, "progress_pct": 17.4, "epoch_pct": 17.4, "eta": "56:58:20", "max_grad_norm": 0.8, "loss": 0.6210505366325378, "grad_norm": 1.2142512798309326, "learning_rate": 9.745128711620535e-05}
+{"ts": "2025-12-24T02:23:47", "event": "train_log", "step": 2422, "epoch": 1.04487594390507, "progress_pct": 17.41, "epoch_pct": 17.41, "eta": "56:56:27", "max_grad_norm": 0.8, "loss": 0.6243326663970947, "grad_norm": 1.1306709051132202, "learning_rate": 9.744354223519001e-05}
+{"ts": "2025-12-24T02:24:07", "event": "train_log", "step": 2424, "epoch": 1.0457389428263215, "progress_pct": 17.43, "epoch_pct": 17.43, "eta": "56:54:38", "max_grad_norm": 0.8, "loss": 0.6312575340270996, "grad_norm": 1.202621340751648, "learning_rate": 9.743578591337833e-05}
+{"ts": "2025-12-24T02:24:28", "event": "train_log", "step": 2426, "epoch": 1.0466019417475727, "progress_pct": 17.44, "epoch_pct": 17.44, "eta": "56:52:52", "max_grad_norm": 0.8, "loss": 0.6190714836120605, "grad_norm": 1.0556137561798096, "learning_rate": 9.742801815264072e-05}
+{"ts": "2025-12-24T02:24:48", "event": "train_log", "step": 2428, "epoch": 1.0474649406688241, "progress_pct": 17.46, "epoch_pct": 17.46, "eta": "56:51:00", "max_grad_norm": 0.8, "loss": 0.7496671080589294, "grad_norm": 1.5071814060211182, "learning_rate": 9.742023895485033e-05}
+{"ts": "2025-12-24T02:25:08", "event": "train_log", "step": 2430, "epoch": 1.0483279395900755, "progress_pct": 17.47, "epoch_pct": 17.47, "eta": "56:49:13", "max_grad_norm": 0.8, "loss": 0.6288847327232361, "grad_norm": 1.1448185443878174, "learning_rate": 9.74124483218831e-05}
+{"ts": "2025-12-24T02:25:30", "event": "train_log", "step": 2432, "epoch": 1.049190938511327, "progress_pct": 17.49, "epoch_pct": 17.49, "eta": "56:47:34", "max_grad_norm": 0.8, "loss": 0.6938424706459045, "grad_norm": 1.0072964429855347, "learning_rate": 9.74046462556177e-05}
+{"ts": "2025-12-24T02:25:47", "event": "train_log", "step": 2434, "epoch": 1.0500539374325781, "progress_pct": 17.5, "epoch_pct": 17.5, "eta": "56:45:31", "max_grad_norm": 0.8, "loss": 0.7529639601707458, "grad_norm": 1.198836326599121, "learning_rate": 9.739683275793554e-05}
+{"ts": "2025-12-24T02:26:08", "event": "train_log", "step": 2436, "epoch": 1.0509169363538295, "progress_pct": 17.52, "epoch_pct": 17.52, "eta": "56:43:45", "max_grad_norm": 0.8, "loss": 0.7193960547447205, "grad_norm": 1.2170380353927612, "learning_rate": 9.738900783072083e-05}
+{"ts": "2025-12-24T02:26:29", "event": "train_log", "step": 2438, "epoch": 1.051779935275081, "progress_pct": 17.53, "epoch_pct": 17.53, "eta": "56:42:00", "max_grad_norm": 0.8, "loss": 0.6488150954246521, "grad_norm": 1.0384336709976196, "learning_rate": 9.738117147586051e-05}
+{"ts": "2025-12-24T02:26:48", "event": "train_log", "step": 2440, "epoch": 1.0526429341963321, "progress_pct": 17.54, "epoch_pct": 17.54, "eta": "56:40:09", "max_grad_norm": 0.8, "loss": 0.6672247052192688, "grad_norm": 1.1562492847442627, "learning_rate": 9.737332369524429e-05}
+{"ts": "2025-12-24T02:27:10", "event": "train_log", "step": 2442, "epoch": 1.0535059331175836, "progress_pct": 17.56, "epoch_pct": 17.56, "eta": "56:38:27", "max_grad_norm": 0.8, "loss": 0.651012122631073, "grad_norm": 1.1553802490234375, "learning_rate": 9.736546449076462e-05}
+{"ts": "2025-12-24T02:27:31", "event": "train_log", "step": 2444, "epoch": 1.054368932038835, "progress_pct": 17.57, "epoch_pct": 17.57, "eta": "56:36:42", "max_grad_norm": 0.8, "loss": 0.681246817111969, "grad_norm": 1.136838674545288, "learning_rate": 9.735759386431671e-05}
+{"ts": "2025-12-24T02:27:51", "event": "train_log", "step": 2446, "epoch": 1.0552319309600864, "progress_pct": 17.59, "epoch_pct": 17.59, "eta": "56:34:56", "max_grad_norm": 0.8, "loss": 0.7331019639968872, "grad_norm": 1.177869200706482, "learning_rate": 9.734971181779854e-05}
+{"ts": "2025-12-24T02:28:12", "event": "train_log", "step": 2448, "epoch": 1.0560949298813376, "progress_pct": 17.6, "epoch_pct": 17.6, "eta": "56:33:10", "max_grad_norm": 0.8, "loss": 0.7316935658454895, "grad_norm": 1.2983415126800537, "learning_rate": 9.73418183531108e-05}
+{"ts": "2025-12-24T02:28:33", "event": "train_log", "step": 2450, "epoch": 1.056957928802589, "progress_pct": 17.62, "epoch_pct": 17.62, "eta": "56:31:27", "max_grad_norm": 0.8, "loss": 0.6513608694076538, "grad_norm": 1.1189959049224854, "learning_rate": 9.733391347215699e-05}
+{"ts": "2025-12-24T02:28:52", "event": "train_log", "step": 2452, "epoch": 1.0578209277238404, "progress_pct": 17.63, "epoch_pct": 17.63, "eta": "56:29:35", "max_grad_norm": 0.8, "loss": 0.6469123363494873, "grad_norm": 1.3256937265396118, "learning_rate": 9.732599717684334e-05}
+{"ts": "2025-12-24T02:29:12", "event": "train_log", "step": 2454, "epoch": 1.0586839266450916, "progress_pct": 17.64, "epoch_pct": 17.64, "eta": "56:27:47", "max_grad_norm": 0.8, "loss": 0.5860153436660767, "grad_norm": 0.9766954779624939, "learning_rate": 9.731806946907881e-05}
+{"ts": "2025-12-24T02:29:31", "event": "train_log", "step": 2456, "epoch": 1.059546925566343, "progress_pct": 17.66, "epoch_pct": 17.66, "eta": "56:25:54", "max_grad_norm": 0.8, "loss": 0.6783872842788696, "grad_norm": 1.1360441446304321, "learning_rate": 9.731013035077514e-05}
+{"ts": "2025-12-24T02:29:49", "event": "train_log", "step": 2458, "epoch": 1.0604099244875944, "progress_pct": 17.67, "epoch_pct": 17.67, "eta": "56:23:57", "max_grad_norm": 0.8, "loss": 0.7894490361213684, "grad_norm": 1.3483080863952637, "learning_rate": 9.73021798238468e-05}
+{"ts": "2025-12-24T02:30:07", "event": "train_log", "step": 2460, "epoch": 1.0612729234088458, "progress_pct": 17.69, "epoch_pct": 17.69, "eta": "56:22:02", "max_grad_norm": 0.8, "loss": 0.7098361849784851, "grad_norm": 1.1412159204483032, "learning_rate": 9.729421789021105e-05}
+{"ts": "2025-12-24T02:30:25", "event": "train_log", "step": 2462, "epoch": 1.062135922330097, "progress_pct": 17.7, "epoch_pct": 17.7, "eta": "56:20:06", "max_grad_norm": 0.8, "loss": 0.7138097286224365, "grad_norm": 1.2670512199401855, "learning_rate": 9.728624455178785e-05}
+{"ts": "2025-12-24T02:30:43", "event": "train_log", "step": 2464, "epoch": 1.0629989212513484, "progress_pct": 17.72, "epoch_pct": 17.72, "eta": "56:18:09", "max_grad_norm": 0.8, "loss": 0.7136340141296387, "grad_norm": 1.2577205896377563, "learning_rate": 9.727825981049994e-05}
+{"ts": "2025-12-24T02:31:03", "event": "train_log", "step": 2466, "epoch": 1.0638619201725998, "progress_pct": 17.73, "epoch_pct": 17.73, "eta": "56:16:19", "max_grad_norm": 0.8, "loss": 0.6273087859153748, "grad_norm": 1.098997712135315, "learning_rate": 9.727026366827283e-05}
+{"ts": "2025-12-24T02:31:21", "event": "train_log", "step": 2468, "epoch": 1.064724919093851, "progress_pct": 17.75, "epoch_pct": 17.75, "eta": "56:14:26", "max_grad_norm": 0.8, "loss": 0.7076449394226074, "grad_norm": 1.1550003290176392, "learning_rate": 9.72622561270347e-05}
+{"ts": "2025-12-24T02:31:40", "event": "train_log", "step": 2470, "epoch": 1.0655879180151024, "progress_pct": 17.76, "epoch_pct": 17.76, "eta": "56:12:35", "max_grad_norm": 0.8, "loss": 0.6856985092163086, "grad_norm": 1.0861129760742188, "learning_rate": 9.725423718871659e-05}
+{"ts": "2025-12-24T02:31:59", "event": "train_log", "step": 2472, "epoch": 1.0664509169363539, "progress_pct": 17.77, "epoch_pct": 17.77, "eta": "56:10:43", "max_grad_norm": 0.8, "loss": 0.7311446070671082, "grad_norm": 1.2742570638656616, "learning_rate": 9.724620685525219e-05}
+{"ts": "2025-12-24T02:32:17", "event": "train_log", "step": 2474, "epoch": 1.0673139158576053, "progress_pct": 17.79, "epoch_pct": 17.79, "eta": "56:08:48", "max_grad_norm": 0.8, "loss": 0.7728472948074341, "grad_norm": 1.504339337348938, "learning_rate": 9.7238165128578e-05}
+{"ts": "2025-12-24T02:32:37", "event": "train_log", "step": 2476, "epoch": 1.0681769147788565, "progress_pct": 17.8, "epoch_pct": 17.8, "eta": "56:06:58", "max_grad_norm": 0.8, "loss": 0.6259000301361084, "grad_norm": 1.2232030630111694, "learning_rate": 9.723011201063323e-05}
+{"ts": "2025-12-24T02:32:55", "event": "train_log", "step": 2478, "epoch": 1.0690399137001079, "progress_pct": 17.82, "epoch_pct": 17.82, "eta": "56:05:05", "max_grad_norm": 0.8, "loss": 0.7132798433303833, "grad_norm": 1.3238953351974487, "learning_rate": 9.722204750335989e-05}
+{"ts": "2025-12-24T02:33:15", "event": "train_log", "step": 2480, "epoch": 1.0699029126213593, "progress_pct": 17.83, "epoch_pct": 17.83, "eta": "56:03:19", "max_grad_norm": 0.8, "loss": 0.6067986488342285, "grad_norm": 1.2286484241485596, "learning_rate": 9.721397160870265e-05}
+{"ts": "2025-12-24T02:33:33", "event": "train_log", "step": 2482, "epoch": 1.0707659115426105, "progress_pct": 17.85, "epoch_pct": 17.85, "eta": "56:01:24", "max_grad_norm": 0.8, "loss": 0.6884648203849792, "grad_norm": 1.3101580142974854, "learning_rate": 9.7205884328609e-05}
+{"ts": "2025-12-24T02:33:51", "event": "train_log", "step": 2484, "epoch": 1.0716289104638619, "progress_pct": 17.86, "epoch_pct": 17.86, "eta": "55:59:28", "max_grad_norm": 0.8, "loss": 0.7164831757545471, "grad_norm": 1.437833309173584, "learning_rate": 9.719778566502915e-05}
+{"ts": "2025-12-24T02:34:10", "event": "train_log", "step": 2486, "epoch": 1.0724919093851133, "progress_pct": 17.87, "epoch_pct": 17.87, "eta": "55:57:39", "max_grad_norm": 0.8, "loss": 0.6326627731323242, "grad_norm": 1.0980390310287476, "learning_rate": 9.718967561991606e-05}
+{"ts": "2025-12-24T02:34:30", "event": "train_log", "step": 2488, "epoch": 1.0733549083063647, "progress_pct": 17.89, "epoch_pct": 17.89, "eta": "55:55:55", "max_grad_norm": 0.8, "loss": 0.5712647438049316, "grad_norm": 0.9919570088386536, "learning_rate": 9.718155419522543e-05}
+{"ts": "2025-12-24T02:34:49", "event": "train_log", "step": 2490, "epoch": 1.074217907227616, "progress_pct": 17.9, "epoch_pct": 17.9, "eta": "55:54:05", "max_grad_norm": 0.8, "loss": 0.6923331022262573, "grad_norm": 1.0773793458938599, "learning_rate": 9.717342139291569e-05}
+{"ts": "2025-12-24T02:35:08", "event": "train_log", "step": 2492, "epoch": 1.0750809061488673, "progress_pct": 17.92, "epoch_pct": 17.92, "eta": "55:52:16", "max_grad_norm": 0.8, "loss": 0.6737485527992249, "grad_norm": 1.1553318500518799, "learning_rate": 9.716527721494805e-05}
+{"ts": "2025-12-24T02:35:27", "event": "train_log", "step": 2494, "epoch": 1.0759439050701187, "progress_pct": 17.93, "epoch_pct": 17.93, "eta": "55:50:26", "max_grad_norm": 0.8, "loss": 0.6422063708305359, "grad_norm": 1.112865924835205, "learning_rate": 9.715712166328644e-05}
+{"ts": "2025-12-24T02:35:47", "event": "train_log", "step": 2496, "epoch": 1.07680690399137, "progress_pct": 17.95, "epoch_pct": 17.95, "eta": "55:48:42", "max_grad_norm": 0.8, "loss": 0.6399799585342407, "grad_norm": 1.0651512145996094, "learning_rate": 9.71489547398975e-05}
+{"ts": "2025-12-24T02:36:08", "event": "train_log", "step": 2498, "epoch": 1.0776699029126213, "progress_pct": 17.96, "epoch_pct": 17.96, "eta": "55:46:59", "max_grad_norm": 0.8, "loss": 0.6218312382698059, "grad_norm": 1.1394895315170288, "learning_rate": 9.71407764467507e-05}
+{"ts": "2025-12-24T02:36:26", "event": "train_log", "step": 2500, "epoch": 1.0785329018338727, "progress_pct": 17.98, "epoch_pct": 17.98, "eta": "55:45:07", "max_grad_norm": 0.8, "loss": 0.65947425365448, "grad_norm": 1.3323278427124023, "learning_rate": 9.713258678581815e-05}
+{"ts": "2025-12-24T02:50:23", "event": "train_log", "step": 2500, "epoch": 1.0785329018338727, "progress_pct": 17.98, "epoch_pct": 17.98, "eta": "56:48:44", "max_grad_norm": 0.8, "eval_loss": 0.7294915914535522, "eval_runtime": 836.4643, "eval_samples_per_second": 2.463, "eval_steps_per_second": 2.463}
+{"ts": "2025-12-24T02:50:43", "event": "train_log", "step": 2502, "epoch": 1.0793959007551241, "progress_pct": 17.99, "epoch_pct": 17.99, "eta": "56:46:57", "max_grad_norm": 0.8, "loss": 0.66681307554245, "grad_norm": 1.3122533559799194, "learning_rate": 9.712438575907479e-05}
+{"ts": "2025-12-24T02:51:02", "event": "train_log", "step": 2504, "epoch": 1.0802588996763753, "progress_pct": 18.0, "epoch_pct": 18.0, "eta": "56:45:04", "max_grad_norm": 0.8, "loss": 0.6926553249359131, "grad_norm": 1.3631974458694458, "learning_rate": 9.711617336849824e-05}
+{"ts": "2025-12-24T02:51:21", "event": "train_log", "step": 2506, "epoch": 1.0811218985976268, "progress_pct": 18.02, "epoch_pct": 18.02, "eta": "56:43:14", "max_grad_norm": 0.8, "loss": 0.6706596612930298, "grad_norm": 1.2383692264556885, "learning_rate": 9.710794961606886e-05}
+{"ts": "2025-12-24T02:51:40", "event": "train_log", "step": 2508, "epoch": 1.0819848975188782, "progress_pct": 18.03, "epoch_pct": 18.03, "eta": "56:41:18", "max_grad_norm": 0.8, "loss": 0.6011359095573425, "grad_norm": 1.1660504341125488, "learning_rate": 9.709971450376982e-05}
+{"ts": "2025-12-24T02:51:59", "event": "train_log", "step": 2510, "epoch": 1.0828478964401294, "progress_pct": 18.05, "epoch_pct": 18.05, "eta": "56:39:30", "max_grad_norm": 0.8, "loss": 0.689009428024292, "grad_norm": 1.3021246194839478, "learning_rate": 9.709146803358693e-05}
+{"ts": "2025-12-24T02:52:18", "event": "train_log", "step": 2512, "epoch": 1.0837108953613808, "progress_pct": 18.06, "epoch_pct": 18.06, "eta": "56:37:37", "max_grad_norm": 0.8, "loss": 0.6216740012168884, "grad_norm": 1.0553252696990967, "learning_rate": 9.708321020750883e-05}
+{"ts": "2025-12-24T02:52:37", "event": "train_log", "step": 2514, "epoch": 1.0845738942826322, "progress_pct": 18.08, "epoch_pct": 18.08, "eta": "56:35:44", "max_grad_norm": 0.8, "loss": 0.6439616680145264, "grad_norm": 1.1862177848815918, "learning_rate": 9.707494102752683e-05}
+{"ts": "2025-12-24T02:52:56", "event": "train_log", "step": 2516, "epoch": 1.0854368932038836, "progress_pct": 18.09, "epoch_pct": 18.09, "eta": "56:33:54", "max_grad_norm": 0.8, "loss": 0.6825361251831055, "grad_norm": 1.1166558265686035, "learning_rate": 9.706666049563504e-05}
+{"ts": "2025-12-24T02:53:15", "event": "train_log", "step": 2518, "epoch": 1.0862998921251348, "progress_pct": 18.1, "epoch_pct": 18.1, "eta": "56:32:02", "max_grad_norm": 0.8, "loss": 0.6471308469772339, "grad_norm": 1.128812551498413, "learning_rate": 9.705836861383022e-05}
+{"ts": "2025-12-24T02:53:33", "event": "train_log", "step": 2520, "epoch": 1.0871628910463862, "progress_pct": 18.12, "epoch_pct": 18.12, "eta": "56:30:07", "max_grad_norm": 0.8, "loss": 0.6535511016845703, "grad_norm": 1.324802279472351, "learning_rate": 9.705006538411197e-05}
+{"ts": "2025-12-24T02:53:52", "event": "train_log", "step": 2522, "epoch": 1.0880258899676376, "progress_pct": 18.13, "epoch_pct": 18.13, "eta": "56:28:13", "max_grad_norm": 0.8, "loss": 0.678022563457489, "grad_norm": 1.3586591482162476, "learning_rate": 9.704175080848255e-05}
+{"ts": "2025-12-24T02:54:11", "event": "train_log", "step": 2524, "epoch": 1.0888888888888888, "progress_pct": 18.15, "epoch_pct": 18.15, "eta": "56:26:22", "max_grad_norm": 0.8, "loss": 0.5904073119163513, "grad_norm": 1.1334232091903687, "learning_rate": 9.703342488894697e-05}
+{"ts": "2025-12-24T02:54:31", "event": "train_log", "step": 2526, "epoch": 1.0897518878101402, "progress_pct": 18.16, "epoch_pct": 18.16, "eta": "56:24:36", "max_grad_norm": 0.8, "loss": 0.6072601079940796, "grad_norm": 1.182215929031372, "learning_rate": 9.702508762751303e-05}
+{"ts": "2025-12-24T02:54:50", "event": "train_log", "step": 2528, "epoch": 1.0906148867313916, "progress_pct": 18.18, "epoch_pct": 18.18, "eta": "56:22:45", "max_grad_norm": 0.8, "loss": 0.6752429604530334, "grad_norm": 1.2525769472122192, "learning_rate": 9.701673902619119e-05}
+{"ts": "2025-12-24T02:55:08", "event": "train_log", "step": 2530, "epoch": 1.091477885652643, "progress_pct": 18.19, "epoch_pct": 18.19, "eta": "56:20:50", "max_grad_norm": 0.8, "loss": 0.7192611694335938, "grad_norm": 1.2811026573181152, "learning_rate": 9.70083790869947e-05}
+{"ts": "2025-12-24T02:55:26", "event": "train_log", "step": 2532, "epoch": 1.0923408845738942, "progress_pct": 18.21, "epoch_pct": 18.21, "eta": "56:18:57", "max_grad_norm": 0.8, "loss": 0.6991751194000244, "grad_norm": 1.1434061527252197, "learning_rate": 9.700000781193949e-05}
+{"ts": "2025-12-24T02:55:45", "event": "train_log", "step": 2534, "epoch": 1.0932038834951456, "progress_pct": 18.22, "epoch_pct": 18.22, "eta": "56:17:04", "max_grad_norm": 0.8, "loss": 0.7010498046875, "grad_norm": 1.26658296585083, "learning_rate": 9.699162520304428e-05}
+{"ts": "2025-12-24T02:56:04", "event": "train_log", "step": 2536, "epoch": 1.094066882416397, "progress_pct": 18.23, "epoch_pct": 18.23, "eta": "56:15:13", "max_grad_norm": 0.8, "loss": 0.6945615410804749, "grad_norm": 1.2083529233932495, "learning_rate": 9.698323126233048e-05}
+{"ts": "2025-12-24T02:56:23", "event": "train_log", "step": 2538, "epoch": 1.0949298813376482, "progress_pct": 18.25, "epoch_pct": 18.25, "eta": "56:13:25", "max_grad_norm": 0.8, "loss": 0.689496636390686, "grad_norm": 1.2064114809036255, "learning_rate": 9.697482599182227e-05}
+{"ts": "2025-12-24T02:56:42", "event": "train_log", "step": 2540, "epoch": 1.0957928802588996, "progress_pct": 18.26, "epoch_pct": 18.26, "eta": "56:11:34", "max_grad_norm": 0.8, "loss": 0.6966884136199951, "grad_norm": 1.17113196849823, "learning_rate": 9.696640939354655e-05}
+{"ts": "2025-12-24T02:57:01", "event": "train_log", "step": 2542, "epoch": 1.096655879180151, "progress_pct": 18.28, "epoch_pct": 18.28, "eta": "56:09:45", "max_grad_norm": 0.8, "loss": 0.6524666547775269, "grad_norm": 1.1551356315612793, "learning_rate": 9.695798146953291e-05}
+{"ts": "2025-12-24T02:57:22", "event": "train_log", "step": 2544, "epoch": 1.0975188781014025, "progress_pct": 18.29, "epoch_pct": 18.29, "eta": "56:08:04", "max_grad_norm": 0.8, "loss": 0.5769726037979126, "grad_norm": 1.0966521501541138, "learning_rate": 9.694954222181374e-05}
+{"ts": "2025-12-24T02:57:39", "event": "train_log", "step": 2546, "epoch": 1.0983818770226537, "progress_pct": 18.31, "epoch_pct": 18.31, "eta": "56:06:08", "max_grad_norm": 0.8, "loss": 0.6986333727836609, "grad_norm": 1.320669174194336, "learning_rate": 9.694109165242411e-05}
+{"ts": "2025-12-24T02:57:58", "event": "train_log", "step": 2548, "epoch": 1.099244875943905, "progress_pct": 18.32, "epoch_pct": 18.32, "eta": "56:04:15", "max_grad_norm": 0.8, "loss": 0.7014455199241638, "grad_norm": 1.1678471565246582, "learning_rate": 9.693262976340183e-05}
+{"ts": "2025-12-24T02:58:18", "event": "train_log", "step": 2550, "epoch": 1.1001078748651565, "progress_pct": 18.33, "epoch_pct": 18.34, "eta": "56:02:31", "max_grad_norm": 0.8, "loss": 0.5786730647087097, "grad_norm": 1.1527155637741089, "learning_rate": 9.692415655678745e-05}
+{"ts": "2025-12-24T02:58:37", "event": "train_log", "step": 2552, "epoch": 1.1009708737864077, "progress_pct": 18.35, "epoch_pct": 18.35, "eta": "56:00:43", "max_grad_norm": 0.8, "loss": 0.755945086479187, "grad_norm": 1.2511217594146729, "learning_rate": 9.691567203462428e-05}
+{"ts": "2025-12-24T02:58:56", "event": "train_log", "step": 2554, "epoch": 1.101833872707659, "progress_pct": 18.36, "epoch_pct": 18.36, "eta": "55:58:56", "max_grad_norm": 0.8, "loss": 0.6487610936164856, "grad_norm": 1.0910221338272095, "learning_rate": 9.690717619895828e-05}
+{"ts": "2025-12-24T02:59:16", "event": "train_log", "step": 2556, "epoch": 1.1026968716289105, "progress_pct": 18.38, "epoch_pct": 18.38, "eta": "55:57:09", "max_grad_norm": 0.8, "loss": 0.6346665024757385, "grad_norm": 1.3682197332382202, "learning_rate": 9.68986690518382e-05}
+{"ts": "2025-12-24T02:59:35", "event": "train_log", "step": 2558, "epoch": 1.103559870550162, "progress_pct": 18.39, "epoch_pct": 18.39, "eta": "55:55:21", "max_grad_norm": 0.8, "loss": 0.647144615650177, "grad_norm": 1.2161564826965332, "learning_rate": 9.68901505953155e-05}
+{"ts": "2025-12-24T02:59:54", "event": "train_log", "step": 2560, "epoch": 1.104422869471413, "progress_pct": 18.41, "epoch_pct": 18.41, "eta": "55:53:31", "max_grad_norm": 0.8, "loss": 0.6461755633354187, "grad_norm": 1.2559088468551636, "learning_rate": 9.688162083144436e-05}
+{"ts": "2025-12-24T03:00:12", "event": "train_log", "step": 2562, "epoch": 1.1052858683926645, "progress_pct": 18.42, "epoch_pct": 18.42, "eta": "55:51:40", "max_grad_norm": 0.8, "loss": 0.6888763904571533, "grad_norm": 1.1260225772857666, "learning_rate": 9.68730797622817e-05}
+{"ts": "2025-12-24T03:00:31", "event": "train_log", "step": 2564, "epoch": 1.106148867313916, "progress_pct": 18.44, "epoch_pct": 18.44, "eta": "55:49:50", "max_grad_norm": 0.8, "loss": 0.6318327784538269, "grad_norm": 1.1499868631362915, "learning_rate": 9.686452738988716e-05}
+{"ts": "2025-12-24T03:00:49", "event": "train_log", "step": 2566, "epoch": 1.1070118662351671, "progress_pct": 18.45, "epoch_pct": 18.45, "eta": "55:47:59", "max_grad_norm": 0.8, "loss": 0.6538340449333191, "grad_norm": 1.1775025129318237, "learning_rate": 9.68559637163231e-05}
+{"ts": "2025-12-24T03:01:08", "event": "train_log", "step": 2568, "epoch": 1.1078748651564185, "progress_pct": 18.46, "epoch_pct": 18.46, "eta": "55:46:11", "max_grad_norm": 0.8, "loss": 0.6706726551055908, "grad_norm": 1.264656901359558, "learning_rate": 9.684738874365465e-05}
+{"ts": "2025-12-24T03:01:27", "event": "train_log", "step": 2570, "epoch": 1.10873786407767, "progress_pct": 18.48, "epoch_pct": 18.48, "eta": "55:44:23", "max_grad_norm": 0.8, "loss": 0.7573289275169373, "grad_norm": 1.1130270957946777, "learning_rate": 9.683880247394956e-05}
+{"ts": "2025-12-24T03:01:45", "event": "train_log", "step": 2572, "epoch": 1.1096008629989211, "progress_pct": 18.49, "epoch_pct": 18.49, "eta": "55:42:33", "max_grad_norm": 0.8, "loss": 0.6603306531906128, "grad_norm": 1.2781217098236084, "learning_rate": 9.683020490927841e-05}
+{"ts": "2025-12-24T03:02:05", "event": "train_log", "step": 2574, "epoch": 1.1104638619201725, "progress_pct": 18.51, "epoch_pct": 18.51, "eta": "55:40:46", "max_grad_norm": 0.8, "loss": 0.6523779630661011, "grad_norm": 1.251062035560608, "learning_rate": 9.682159605171446e-05}
+{"ts": "2025-12-24T03:02:23", "event": "train_log", "step": 2576, "epoch": 1.111326860841424, "progress_pct": 18.52, "epoch_pct": 18.52, "eta": "55:38:56", "max_grad_norm": 0.8, "loss": 0.677550733089447, "grad_norm": 1.2631903886795044, "learning_rate": 9.68129759033337e-05}
+{"ts": "2025-12-24T03:02:42", "event": "train_log", "step": 2578, "epoch": 1.1121898597626754, "progress_pct": 18.54, "epoch_pct": 18.54, "eta": "55:37:09", "max_grad_norm": 0.8, "loss": 0.666540801525116, "grad_norm": 1.0212602615356445, "learning_rate": 9.680434446621483e-05}
+{"ts": "2025-12-24T03:03:00", "event": "train_log", "step": 2580, "epoch": 1.1130528586839266, "progress_pct": 18.55, "epoch_pct": 18.55, "eta": "55:35:19", "max_grad_norm": 0.8, "loss": 0.7467086315155029, "grad_norm": 1.221718430519104, "learning_rate": 9.679570174243929e-05}
+{"ts": "2025-12-24T03:03:19", "event": "train_log", "step": 2582, "epoch": 1.113915857605178, "progress_pct": 18.56, "epoch_pct": 18.57, "eta": "55:33:33", "max_grad_norm": 0.8, "loss": 0.673492968082428, "grad_norm": 1.1175146102905273, "learning_rate": 9.678704773409124e-05}
+{"ts": "2025-12-24T03:03:39", "event": "train_log", "step": 2584, "epoch": 1.1147788565264294, "progress_pct": 18.58, "epoch_pct": 18.58, "eta": "55:31:48", "max_grad_norm": 0.8, "loss": 0.5724847912788391, "grad_norm": 1.1274988651275635, "learning_rate": 9.677838244325754e-05}
+{"ts": "2025-12-24T03:03:58", "event": "train_log", "step": 2586, "epoch": 1.1156418554476808, "progress_pct": 18.59, "epoch_pct": 18.59, "eta": "55:30:01", "max_grad_norm": 0.8, "loss": 0.634985625743866, "grad_norm": 1.297654390335083, "learning_rate": 9.676970587202779e-05}
+{"ts": "2025-12-24T03:04:18", "event": "train_log", "step": 2588, "epoch": 1.116504854368932, "progress_pct": 18.61, "epoch_pct": 18.61, "eta": "55:28:18", "max_grad_norm": 0.8, "loss": 0.6813890933990479, "grad_norm": 1.1908061504364014, "learning_rate": 9.67610180224943e-05}
+{"ts": "2025-12-24T03:04:36", "event": "train_log", "step": 2590, "epoch": 1.1173678532901834, "progress_pct": 18.62, "epoch_pct": 18.62, "eta": "55:26:29", "max_grad_norm": 0.8, "loss": 0.6738101243972778, "grad_norm": 1.1787118911743164, "learning_rate": 9.675231889675212e-05}
+{"ts": "2025-12-24T03:04:56", "event": "train_log", "step": 2592, "epoch": 1.1182308522114348, "progress_pct": 18.64, "epoch_pct": 18.64, "eta": "55:24:46", "max_grad_norm": 0.8, "loss": 0.6662938594818115, "grad_norm": 1.1327152252197266, "learning_rate": 9.674360849689899e-05}
+{"ts": "2025-12-24T03:05:16", "event": "train_log", "step": 2594, "epoch": 1.119093851132686, "progress_pct": 18.65, "epoch_pct": 18.65, "eta": "55:23:04", "max_grad_norm": 0.8, "loss": 0.677423357963562, "grad_norm": 1.1246299743652344, "learning_rate": 9.67348868250354e-05}
+{"ts": "2025-12-24T03:05:36", "event": "train_log", "step": 2596, "epoch": 1.1199568500539374, "progress_pct": 18.67, "epoch_pct": 18.67, "eta": "55:21:24", "max_grad_norm": 0.8, "loss": 0.6257020235061646, "grad_norm": 1.1411702632904053, "learning_rate": 9.672615388326454e-05}
+{"ts": "2025-12-24T03:05:55", "event": "train_log", "step": 2598, "epoch": 1.1208198489751888, "progress_pct": 18.68, "epoch_pct": 18.68, "eta": "55:19:36", "max_grad_norm": 0.8, "loss": 0.7029470205307007, "grad_norm": 1.1607346534729004, "learning_rate": 9.671740967369231e-05}
+{"ts": "2025-12-24T03:06:16", "event": "train_log", "step": 2600, "epoch": 1.12168284789644, "progress_pct": 18.69, "epoch_pct": 18.69, "eta": "55:18:00", "max_grad_norm": 0.8, "loss": 0.6040990352630615, "grad_norm": 1.047934651374817, "learning_rate": 9.670865419842735e-05}
+{"ts": "2025-12-24T03:20:08", "event": "train_log", "step": 2600, "epoch": 1.12168284789644, "progress_pct": 18.69, "epoch_pct": 18.69, "eta": "56:18:18", "max_grad_norm": 0.8, "eval_loss": 0.7287447452545166, "eval_runtime": 831.8138, "eval_samples_per_second": 2.477, "eval_steps_per_second": 2.477}
+{"ts": "2025-12-24T03:20:27", "event": "train_log", "step": 2602, "epoch": 1.1225458468176914, "progress_pct": 18.71, "epoch_pct": 18.71, "eta": "56:16:30", "max_grad_norm": 0.8, "loss": 0.6872879266738892, "grad_norm": 1.0571783781051636, "learning_rate": 9.669988745958098e-05}
+{"ts": "2025-12-24T03:20:47", "event": "train_log", "step": 2604, "epoch": 1.1234088457389428, "progress_pct": 18.72, "epoch_pct": 18.72, "eta": "56:14:45", "max_grad_norm": 0.8, "loss": 0.6173275113105774, "grad_norm": 1.0940403938293457, "learning_rate": 9.669110945926727e-05}
+{"ts": "2025-12-24T03:21:06", "event": "train_log", "step": 2606, "epoch": 1.1242718446601943, "progress_pct": 18.74, "epoch_pct": 18.74, "eta": "56:12:56", "max_grad_norm": 0.8, "loss": 0.6142664551734924, "grad_norm": 1.1180222034454346, "learning_rate": 9.6682320199603e-05}
+{"ts": "2025-12-24T03:21:24", "event": "train_log", "step": 2608, "epoch": 1.1251348435814454, "progress_pct": 18.75, "epoch_pct": 18.75, "eta": "56:11:05", "max_grad_norm": 0.8, "loss": 0.7225713133811951, "grad_norm": 1.251098871231079, "learning_rate": 9.667351968270768e-05}
+{"ts": "2025-12-24T03:21:43", "event": "train_log", "step": 2610, "epoch": 1.1259978425026969, "progress_pct": 18.77, "epoch_pct": 18.77, "eta": "56:09:15", "max_grad_norm": 0.8, "loss": 0.6830481290817261, "grad_norm": 1.2333059310913086, "learning_rate": 9.666470791070347e-05}
+{"ts": "2025-12-24T03:22:02", "event": "train_log", "step": 2612, "epoch": 1.1268608414239483, "progress_pct": 18.78, "epoch_pct": 18.78, "eta": "56:07:25", "max_grad_norm": 0.8, "loss": 0.6835238337516785, "grad_norm": 1.2904123067855835, "learning_rate": 9.665588488571531e-05}
+{"ts": "2025-12-24T03:22:21", "event": "train_log", "step": 2614, "epoch": 1.1277238403451997, "progress_pct": 18.79, "epoch_pct": 18.8, "eta": "56:05:36", "max_grad_norm": 0.8, "loss": 0.6993544697761536, "grad_norm": 1.4779443740844727, "learning_rate": 9.664705060987085e-05}
+{"ts": "2025-12-24T03:22:39", "event": "train_log", "step": 2616, "epoch": 1.1285868392664509, "progress_pct": 18.81, "epoch_pct": 18.81, "eta": "56:03:45", "max_grad_norm": 0.8, "loss": 0.6819394826889038, "grad_norm": 1.267021656036377, "learning_rate": 9.663820508530042e-05}
+{"ts": "2025-12-24T03:22:58", "event": "train_log", "step": 2618, "epoch": 1.1294498381877023, "progress_pct": 18.82, "epoch_pct": 18.82, "eta": "56:01:59", "max_grad_norm": 0.8, "loss": 0.6200308799743652, "grad_norm": 1.0755046606063843, "learning_rate": 9.662934831413708e-05}
+{"ts": "2025-12-24T03:23:17", "event": "train_log", "step": 2620, "epoch": 1.1303128371089537, "progress_pct": 18.84, "epoch_pct": 18.84, "eta": "56:00:08", "max_grad_norm": 0.8, "loss": 0.7138658165931702, "grad_norm": 1.2043033838272095, "learning_rate": 9.66204802985166e-05}
+{"ts": "2025-12-24T03:23:36", "event": "train_log", "step": 2622, "epoch": 1.1311758360302049, "progress_pct": 18.85, "epoch_pct": 18.85, "eta": "55:58:21", "max_grad_norm": 0.8, "loss": 0.7238332033157349, "grad_norm": 1.2641128301620483, "learning_rate": 9.661160104057744e-05}
+{"ts": "2025-12-24T03:23:55", "event": "train_log", "step": 2624, "epoch": 1.1320388349514563, "progress_pct": 18.87, "epoch_pct": 18.87, "eta": "55:56:34", "max_grad_norm": 0.8, "loss": 0.6401007175445557, "grad_norm": 1.0422990322113037, "learning_rate": 9.660271054246085e-05}
+{"ts": "2025-12-24T03:24:14", "event": "train_log", "step": 2626, "epoch": 1.1329018338727077, "progress_pct": 18.88, "epoch_pct": 18.88, "eta": "55:54:46", "max_grad_norm": 0.8, "loss": 0.6305879950523376, "grad_norm": 1.2298130989074707, "learning_rate": 9.659380880631068e-05}
+{"ts": "2025-12-24T03:24:33", "event": "train_log", "step": 2628, "epoch": 1.133764832793959, "progress_pct": 18.9, "epoch_pct": 18.9, "eta": "55:53:01", "max_grad_norm": 0.8, "loss": 0.5751872062683105, "grad_norm": 1.254573106765747, "learning_rate": 9.658489583427357e-05}
+{"ts": "2025-12-24T03:24:53", "event": "train_log", "step": 2630, "epoch": 1.1346278317152103, "progress_pct": 18.91, "epoch_pct": 18.91, "eta": "55:51:16", "max_grad_norm": 0.8, "loss": 0.6629595756530762, "grad_norm": 1.15679132938385, "learning_rate": 9.657597162849883e-05}
+{"ts": "2025-12-24T03:25:12", "event": "train_log", "step": 2632, "epoch": 1.1354908306364617, "progress_pct": 18.92, "epoch_pct": 18.92, "eta": "55:49:30", "max_grad_norm": 0.8, "loss": 0.6832258105278015, "grad_norm": 1.193629264831543, "learning_rate": 9.656703619113849e-05}
+{"ts": "2025-12-24T03:25:29", "event": "train_log", "step": 2634, "epoch": 1.1363538295577131, "progress_pct": 18.94, "epoch_pct": 18.94, "eta": "55:47:36", "max_grad_norm": 0.8, "loss": 0.6823931932449341, "grad_norm": 1.2330398559570312, "learning_rate": 9.655808952434732e-05}
+{"ts": "2025-12-24T03:25:48", "event": "train_log", "step": 2636, "epoch": 1.1372168284789643, "progress_pct": 18.95, "epoch_pct": 18.95, "eta": "55:45:46", "max_grad_norm": 0.8, "loss": 0.6671395897865295, "grad_norm": 1.3926069736480713, "learning_rate": 9.654913163028272e-05}
+{"ts": "2025-12-24T03:26:06", "event": "train_log", "step": 2638, "epoch": 1.1380798274002157, "progress_pct": 18.97, "epoch_pct": 18.97, "eta": "55:43:59", "max_grad_norm": 0.8, "loss": 0.662219226360321, "grad_norm": 1.2995383739471436, "learning_rate": 9.65401625111049e-05}
+{"ts": "2025-12-24T03:26:25", "event": "train_log", "step": 2640, "epoch": 1.1389428263214671, "progress_pct": 18.98, "epoch_pct": 18.98, "eta": "55:42:09", "max_grad_norm": 0.8, "loss": 0.6419461965560913, "grad_norm": 1.1584415435791016, "learning_rate": 9.653118216897667e-05}
+{"ts": "2025-12-24T03:26:43", "event": "train_log", "step": 2642, "epoch": 1.1398058252427186, "progress_pct": 19.0, "epoch_pct": 19.0, "eta": "55:40:20", "max_grad_norm": 0.8, "loss": 0.6844428181648254, "grad_norm": 1.3584444522857666, "learning_rate": 9.652219060606364e-05}
+{"ts": "2025-12-24T03:27:02", "event": "train_log", "step": 2644, "epoch": 1.1406688241639698, "progress_pct": 19.01, "epoch_pct": 19.01, "eta": "55:38:34", "max_grad_norm": 0.8, "loss": 0.6761999130249023, "grad_norm": 1.2619858980178833, "learning_rate": 9.651318782453407e-05}
+{"ts": "2025-12-24T03:27:21", "event": "train_log", "step": 2646, "epoch": 1.1415318230852212, "progress_pct": 19.03, "epoch_pct": 19.03, "eta": "55:36:48", "max_grad_norm": 0.8, "loss": 0.7258123159408569, "grad_norm": 1.2202177047729492, "learning_rate": 9.650417382655892e-05}
+{"ts": "2025-12-24T03:27:43", "event": "train_log", "step": 2648, "epoch": 1.1423948220064726, "progress_pct": 19.04, "epoch_pct": 19.04, "eta": "55:35:16", "max_grad_norm": 0.8, "loss": 0.5709917545318604, "grad_norm": 1.0225578546524048, "learning_rate": 9.64951486143119e-05}
+{"ts": "2025-12-24T03:28:04", "event": "train_log", "step": 2650, "epoch": 1.1432578209277238, "progress_pct": 19.05, "epoch_pct": 19.05, "eta": "55:33:35", "max_grad_norm": 0.8, "loss": 0.7353709936141968, "grad_norm": 1.0961005687713623, "learning_rate": 9.64861121899694e-05}
+{"ts": "2025-12-24T03:28:24", "event": "train_log", "step": 2652, "epoch": 1.1441208198489752, "progress_pct": 19.07, "epoch_pct": 19.07, "eta": "55:31:55", "max_grad_norm": 0.8, "loss": 0.7083833813667297, "grad_norm": 0.9393200874328613, "learning_rate": 9.64770645557105e-05}
+{"ts": "2025-12-24T03:28:44", "event": "train_log", "step": 2654, "epoch": 1.1449838187702266, "progress_pct": 19.08, "epoch_pct": 19.08, "eta": "55:30:13", "max_grad_norm": 0.8, "loss": 0.6405303478240967, "grad_norm": 1.089733362197876, "learning_rate": 9.646800571371701e-05}
+{"ts": "2025-12-24T03:29:03", "event": "train_log", "step": 2656, "epoch": 1.1458468176914778, "progress_pct": 19.1, "epoch_pct": 19.1, "eta": "55:28:30", "max_grad_norm": 0.8, "loss": 0.617733359336853, "grad_norm": 1.1581398248672485, "learning_rate": 9.645893566617341e-05}
+{"ts": "2025-12-24T03:29:23", "event": "train_log", "step": 2658, "epoch": 1.1467098166127292, "progress_pct": 19.11, "epoch_pct": 19.11, "eta": "55:26:47", "max_grad_norm": 0.8, "loss": 0.6948232650756836, "grad_norm": 1.1336220502853394, "learning_rate": 9.644985441526693e-05}
+{"ts": "2025-12-24T03:29:44", "event": "train_log", "step": 2660, "epoch": 1.1475728155339806, "progress_pct": 19.13, "epoch_pct": 19.13, "eta": "55:25:10", "max_grad_norm": 0.8, "loss": 0.6392919421195984, "grad_norm": 1.028591513633728, "learning_rate": 9.644076196318743e-05}
+{"ts": "2025-12-24T03:30:06", "event": "train_log", "step": 2662, "epoch": 1.148435814455232, "progress_pct": 19.14, "epoch_pct": 19.14, "eta": "55:23:37", "max_grad_norm": 0.8, "loss": 0.6384931206703186, "grad_norm": 1.2088173627853394, "learning_rate": 9.643165831212756e-05}
+{"ts": "2025-12-24T03:30:26", "event": "train_log", "step": 2664, "epoch": 1.1492988133764832, "progress_pct": 19.15, "epoch_pct": 19.15, "eta": "55:21:57", "max_grad_norm": 0.8, "loss": 0.5909387469291687, "grad_norm": 0.9905838966369629, "learning_rate": 9.642254346428259e-05}
+{"ts": "2025-12-24T03:30:43", "event": "train_log", "step": 2666, "epoch": 1.1501618122977346, "progress_pct": 19.17, "epoch_pct": 19.17, "eta": "55:20:06", "max_grad_norm": 0.8, "loss": 0.700904130935669, "grad_norm": 1.1670424938201904, "learning_rate": 9.641341742185054e-05}
+{"ts": "2025-12-24T03:31:04", "event": "train_log", "step": 2668, "epoch": 1.151024811218986, "progress_pct": 19.18, "epoch_pct": 19.18, "eta": "55:18:27", "max_grad_norm": 0.8, "loss": 0.6036953330039978, "grad_norm": 1.164746880531311, "learning_rate": 9.640428018703211e-05}
+{"ts": "2025-12-24T03:31:24", "event": "train_log", "step": 2670, "epoch": 1.1518878101402374, "progress_pct": 19.2, "epoch_pct": 19.2, "eta": "55:16:50", "max_grad_norm": 0.8, "loss": 0.6158549785614014, "grad_norm": 1.00895357131958, "learning_rate": 9.639513176203072e-05}
+{"ts": "2025-12-24T03:31:43", "event": "train_log", "step": 2672, "epoch": 1.1527508090614886, "progress_pct": 19.21, "epoch_pct": 19.21, "eta": "55:15:04", "max_grad_norm": 0.8, "loss": 0.7280431389808655, "grad_norm": 1.2236207723617554, "learning_rate": 9.638597214905244e-05}
+{"ts": "2025-12-24T03:32:03", "event": "train_log", "step": 2674, "epoch": 1.15361380798274, "progress_pct": 19.23, "epoch_pct": 19.23, "eta": "55:13:23", "max_grad_norm": 0.8, "loss": 0.6434410810470581, "grad_norm": 1.0930732488632202, "learning_rate": 9.637680135030609e-05}
+{"ts": "2025-12-24T03:32:23", "event": "train_log", "step": 2676, "epoch": 1.1544768069039915, "progress_pct": 19.24, "epoch_pct": 19.24, "eta": "55:11:41", "max_grad_norm": 0.8, "loss": 0.6771191954612732, "grad_norm": 1.366325855255127, "learning_rate": 9.636761936800315e-05}
+{"ts": "2025-12-24T03:32:41", "event": "train_log", "step": 2678, "epoch": 1.1553398058252426, "progress_pct": 19.26, "epoch_pct": 19.26, "eta": "55:09:56", "max_grad_norm": 0.8, "loss": 0.6382432579994202, "grad_norm": 2.249835968017578, "learning_rate": 9.635842620435785e-05}
+{"ts": "2025-12-24T03:33:01", "event": "train_log", "step": 2680, "epoch": 1.156202804746494, "progress_pct": 19.27, "epoch_pct": 19.27, "eta": "55:08:13", "max_grad_norm": 0.8, "loss": 0.6733657121658325, "grad_norm": 1.274483561515808, "learning_rate": 9.634922186158704e-05}
+{"ts": "2025-12-24T03:33:19", "event": "train_log", "step": 2682, "epoch": 1.1570658036677455, "progress_pct": 19.28, "epoch_pct": 19.28, "eta": "55:06:29", "max_grad_norm": 0.8, "loss": 0.7601200938224792, "grad_norm": 1.3715782165527344, "learning_rate": 9.634000634191033e-05}
+{"ts": "2025-12-24T03:33:40", "event": "train_log", "step": 2684, "epoch": 1.1579288025889967, "progress_pct": 19.3, "epoch_pct": 19.3, "eta": "55:04:51", "max_grad_norm": 0.8, "loss": 0.5959098935127258, "grad_norm": 1.1312460899353027, "learning_rate": 9.633077964754998e-05}
+{"ts": "2025-12-24T03:33:59", "event": "train_log", "step": 2686, "epoch": 1.158791801510248, "progress_pct": 19.31, "epoch_pct": 19.31, "eta": "55:03:10", "max_grad_norm": 0.8, "loss": 0.615096926689148, "grad_norm": 1.063233494758606, "learning_rate": 9.632154178073098e-05}
+{"ts": "2025-12-24T03:34:19", "event": "train_log", "step": 2688, "epoch": 1.1596548004314995, "progress_pct": 19.33, "epoch_pct": 19.33, "eta": "55:01:29", "max_grad_norm": 0.8, "loss": 0.6553784608840942, "grad_norm": 1.074887752532959, "learning_rate": 9.6312292743681e-05}
+{"ts": "2025-12-24T03:34:39", "event": "train_log", "step": 2690, "epoch": 1.160517799352751, "progress_pct": 19.34, "epoch_pct": 19.34, "eta": "54:59:51", "max_grad_norm": 0.8, "loss": 0.6963269114494324, "grad_norm": 1.17954683303833, "learning_rate": 9.630303253863042e-05}
+{"ts": "2025-12-24T03:34:58", "event": "train_log", "step": 2692, "epoch": 1.161380798274002, "progress_pct": 19.36, "epoch_pct": 19.36, "eta": "54:58:05", "max_grad_norm": 0.8, "loss": 0.6633999347686768, "grad_norm": 1.2549158334732056, "learning_rate": 9.629376116781226e-05}
+{"ts": "2025-12-24T03:35:17", "event": "train_log", "step": 2694, "epoch": 1.1622437971952535, "progress_pct": 19.37, "epoch_pct": 19.37, "eta": "54:56:22", "max_grad_norm": 0.8, "loss": 0.6727597117424011, "grad_norm": 1.237196683883667, "learning_rate": 9.628447863346229e-05}
+{"ts": "2025-12-24T03:35:35", "event": "train_log", "step": 2696, "epoch": 1.163106796116505, "progress_pct": 19.38, "epoch_pct": 19.39, "eta": "54:54:36", "max_grad_norm": 0.8, "loss": 0.7018169164657593, "grad_norm": 1.4047741889953613, "learning_rate": 9.627518493781895e-05}
+{"ts": "2025-12-24T03:35:55", "event": "train_log", "step": 2698, "epoch": 1.1639697950377563, "progress_pct": 19.4, "epoch_pct": 19.4, "eta": "54:52:59", "max_grad_norm": 0.8, "loss": 0.6566510796546936, "grad_norm": 1.15412175655365, "learning_rate": 9.626588008312339e-05}
+{"ts": "2025-12-24T03:36:15", "event": "train_log", "step": 2700, "epoch": 1.1648327939590075, "progress_pct": 19.41, "epoch_pct": 19.41, "eta": "54:51:18", "max_grad_norm": 0.8, "loss": 0.7017931342124939, "grad_norm": 1.194110631942749, "learning_rate": 9.625656407161938e-05}
+{"ts": "2025-12-24T03:49:59", "event": "train_log", "step": 2700, "epoch": 1.1648327939590075, "progress_pct": 19.41, "epoch_pct": 19.41, "eta": "55:48:19", "max_grad_norm": 0.8, "eval_loss": 0.72404944896698, "eval_runtime": 824.2877, "eval_samples_per_second": 2.499, "eval_steps_per_second": 2.499}
+{"ts": "2025-12-24T03:50:18", "event": "train_log", "step": 2702, "epoch": 1.165695792880259, "progress_pct": 19.43, "epoch_pct": 19.43, "eta": "55:46:34", "max_grad_norm": 0.8, "loss": 0.7058532238006592, "grad_norm": 1.1512954235076904, "learning_rate": 9.62472369055535e-05}
+{"ts": "2025-12-24T03:50:35", "event": "train_log", "step": 2704, "epoch": 1.1665587918015103, "progress_pct": 19.44, "epoch_pct": 19.44, "eta": "55:44:40", "max_grad_norm": 0.8, "loss": 0.7458376288414001, "grad_norm": 1.4743460416793823, "learning_rate": 9.623789858717492e-05}
+{"ts": "2025-12-24T03:50:55", "event": "train_log", "step": 2706, "epoch": 1.1674217907227615, "progress_pct": 19.46, "epoch_pct": 19.46, "eta": "55:43:00", "max_grad_norm": 0.8, "loss": 0.6526424288749695, "grad_norm": 1.0526450872421265, "learning_rate": 9.622854911873551e-05}
+{"ts": "2025-12-24T03:51:18", "event": "train_log", "step": 2708, "epoch": 1.168284789644013, "progress_pct": 19.47, "epoch_pct": 19.47, "eta": "55:41:28", "max_grad_norm": 0.8, "loss": 0.6055451035499573, "grad_norm": 1.0408949851989746, "learning_rate": 9.62191885024899e-05}
+{"ts": "2025-12-24T03:51:36", "event": "train_log", "step": 2710, "epoch": 1.1691477885652644, "progress_pct": 19.49, "epoch_pct": 19.49, "eta": "55:39:41", "max_grad_norm": 0.8, "loss": 0.666462779045105, "grad_norm": 1.214052677154541, "learning_rate": 9.620981674069534e-05}
+{"ts": "2025-12-24T03:51:55", "event": "train_log", "step": 2712, "epoch": 1.1700107874865155, "progress_pct": 19.5, "epoch_pct": 19.5, "eta": "55:37:57", "max_grad_norm": 0.8, "loss": 0.7208677530288696, "grad_norm": 1.2526963949203491, "learning_rate": 9.62004338356118e-05}
+{"ts": "2025-12-24T03:52:15", "event": "train_log", "step": 2714, "epoch": 1.170873786407767, "progress_pct": 19.51, "epoch_pct": 19.51, "eta": "55:36:15", "max_grad_norm": 0.8, "loss": 0.6269269585609436, "grad_norm": 1.2097395658493042, "learning_rate": 9.61910397895019e-05}
+{"ts": "2025-12-24T03:52:35", "event": "train_log", "step": 2716, "epoch": 1.1717367853290184, "progress_pct": 19.53, "epoch_pct": 19.53, "eta": "55:34:36", "max_grad_norm": 0.8, "loss": 0.6063299179077148, "grad_norm": 1.035733699798584, "learning_rate": 9.618163460463096e-05}
+{"ts": "2025-12-24T03:52:56", "event": "train_log", "step": 2718, "epoch": 1.1725997842502698, "progress_pct": 19.54, "epoch_pct": 19.54, "eta": "55:32:58", "max_grad_norm": 0.8, "loss": 0.7084641456604004, "grad_norm": 1.2881336212158203, "learning_rate": 9.617221828326706e-05}
+{"ts": "2025-12-24T03:53:17", "event": "train_log", "step": 2720, "epoch": 1.173462783171521, "progress_pct": 19.56, "epoch_pct": 19.56, "eta": "55:31:20", "max_grad_norm": 0.8, "loss": 0.6515718102455139, "grad_norm": 1.2808257341384888, "learning_rate": 9.616279082768085e-05}
+{"ts": "2025-12-24T03:53:38", "event": "train_log", "step": 2722, "epoch": 1.1743257820927724, "progress_pct": 19.57, "epoch_pct": 19.57, "eta": "55:29:43", "max_grad_norm": 0.8, "loss": 0.6745902299880981, "grad_norm": 1.1174249649047852, "learning_rate": 9.615335224014574e-05}
+{"ts": "2025-12-24T03:53:58", "event": "train_log", "step": 2724, "epoch": 1.1751887810140238, "progress_pct": 19.59, "epoch_pct": 19.59, "eta": "55:28:05", "max_grad_norm": 0.8, "loss": 0.6824861168861389, "grad_norm": 1.1222290992736816, "learning_rate": 9.61439025229378e-05}
+{"ts": "2025-12-24T03:54:17", "event": "train_log", "step": 2726, "epoch": 1.176051779935275, "progress_pct": 19.6, "epoch_pct": 19.6, "eta": "55:26:19", "max_grad_norm": 0.8, "loss": 0.6020817160606384, "grad_norm": 1.156187891960144, "learning_rate": 9.613444167833579e-05}
+{"ts": "2025-12-24T03:54:37", "event": "train_log", "step": 2728, "epoch": 1.1769147788565264, "progress_pct": 19.61, "epoch_pct": 19.62, "eta": "55:24:40", "max_grad_norm": 0.8, "loss": 0.6430572271347046, "grad_norm": 1.2502888441085815, "learning_rate": 9.612496970862113e-05}
+{"ts": "2025-12-24T03:54:56", "event": "train_log", "step": 2730, "epoch": 1.1777777777777778, "progress_pct": 19.63, "epoch_pct": 19.63, "eta": "55:22:55", "max_grad_norm": 0.8, "loss": 0.7195103168487549, "grad_norm": 1.6157052516937256, "learning_rate": 9.611548661607797e-05}
+{"ts": "2025-12-24T03:55:14", "event": "train_log", "step": 2732, "epoch": 1.1786407766990292, "progress_pct": 19.64, "epoch_pct": 19.64, "eta": "55:21:07", "max_grad_norm": 0.8, "loss": 0.710418164730072, "grad_norm": 1.3196234703063965, "learning_rate": 9.610599240299308e-05}
+{"ts": "2025-12-24T03:55:33", "event": "train_log", "step": 2734, "epoch": 1.1795037756202804, "progress_pct": 19.66, "epoch_pct": 19.66, "eta": "55:19:25", "max_grad_norm": 0.8, "loss": 0.650095522403717, "grad_norm": 1.2117719650268555, "learning_rate": 9.6096487071656e-05}
+{"ts": "2025-12-24T03:55:52", "event": "train_log", "step": 2736, "epoch": 1.1803667745415318, "progress_pct": 19.67, "epoch_pct": 19.67, "eta": "55:17:41", "max_grad_norm": 0.8, "loss": 0.7150285243988037, "grad_norm": 2.2723631858825684, "learning_rate": 9.608697062435884e-05}
+{"ts": "2025-12-24T03:56:12", "event": "train_log", "step": 2738, "epoch": 1.1812297734627832, "progress_pct": 19.69, "epoch_pct": 19.69, "eta": "55:16:01", "max_grad_norm": 0.8, "loss": 0.6271997690200806, "grad_norm": 1.0523152351379395, "learning_rate": 9.607744306339647e-05}
+{"ts": "2025-12-24T03:56:32", "event": "train_log", "step": 2740, "epoch": 1.1820927723840344, "progress_pct": 19.7, "epoch_pct": 19.7, "eta": "55:14:21", "max_grad_norm": 0.8, "loss": 0.659799337387085, "grad_norm": 1.2567129135131836, "learning_rate": 9.606790439106646e-05}
+{"ts": "2025-12-24T03:56:51", "event": "train_log", "step": 2742, "epoch": 1.1829557713052858, "progress_pct": 19.72, "epoch_pct": 19.72, "eta": "55:12:40", "max_grad_norm": 0.8, "loss": 0.7097326517105103, "grad_norm": 1.1448849439620972, "learning_rate": 9.605835460966896e-05}
+{"ts": "2025-12-24T03:57:12", "event": "train_log", "step": 2744, "epoch": 1.1838187702265373, "progress_pct": 19.73, "epoch_pct": 19.73, "eta": "55:11:04", "max_grad_norm": 0.8, "loss": 0.6692788004875183, "grad_norm": 1.3698351383209229, "learning_rate": 9.604879372150688e-05}
+{"ts": "2025-12-24T03:57:32", "event": "train_log", "step": 2746, "epoch": 1.1846817691477884, "progress_pct": 19.74, "epoch_pct": 19.74, "eta": "55:09:25", "max_grad_norm": 0.8, "loss": 0.6453474760055542, "grad_norm": 1.1424204111099243, "learning_rate": 9.603922172888577e-05}
+{"ts": "2025-12-24T03:57:51", "event": "train_log", "step": 2748, "epoch": 1.1855447680690399, "progress_pct": 19.76, "epoch_pct": 19.76, "eta": "55:07:43", "max_grad_norm": 0.8, "loss": 0.6197408437728882, "grad_norm": 1.2701301574707031, "learning_rate": 9.60296386341139e-05}
+{"ts": "2025-12-24T03:58:12", "event": "train_log", "step": 2750, "epoch": 1.1864077669902913, "progress_pct": 19.77, "epoch_pct": 19.77, "eta": "55:06:05", "max_grad_norm": 0.8, "loss": 0.6525035500526428, "grad_norm": 1.1171884536743164, "learning_rate": 9.602004443950217e-05}
+{"ts": "2025-12-24T03:58:30", "event": "train_log", "step": 2752, "epoch": 1.1872707659115427, "progress_pct": 19.79, "epoch_pct": 19.79, "eta": "55:04:20", "max_grad_norm": 0.8, "loss": 0.6459797620773315, "grad_norm": 1.132633090019226, "learning_rate": 9.60104391473642e-05}
+{"ts": "2025-12-24T03:58:50", "event": "train_log", "step": 2754, "epoch": 1.1881337648327939, "progress_pct": 19.8, "epoch_pct": 19.8, "eta": "55:02:41", "max_grad_norm": 0.8, "loss": 0.6175895929336548, "grad_norm": 1.1362011432647705, "learning_rate": 9.600082276001623e-05}
+{"ts": "2025-12-24T03:59:09", "event": "train_log", "step": 2756, "epoch": 1.1889967637540453, "progress_pct": 19.82, "epoch_pct": 19.82, "eta": "55:01:01", "max_grad_norm": 0.8, "loss": 0.6782001852989197, "grad_norm": 1.1216920614242554, "learning_rate": 9.599119527977722e-05}
+{"ts": "2025-12-24T03:59:30", "event": "train_log", "step": 2758, "epoch": 1.1898597626752967, "progress_pct": 19.83, "epoch_pct": 19.83, "eta": "54:59:24", "max_grad_norm": 0.8, "loss": 0.6387497782707214, "grad_norm": 1.024833083152771, "learning_rate": 9.59815567089688e-05}
+{"ts": "2025-12-24T03:59:49", "event": "train_log", "step": 2760, "epoch": 1.190722761596548, "progress_pct": 19.84, "epoch_pct": 19.85, "eta": "54:57:42", "max_grad_norm": 0.8, "loss": 0.6974280476570129, "grad_norm": 1.3539615869522095, "learning_rate": 9.597190704991527e-05}
+{"ts": "2025-12-24T04:00:09", "event": "train_log", "step": 2762, "epoch": 1.1915857605177993, "progress_pct": 19.86, "epoch_pct": 19.86, "eta": "54:56:06", "max_grad_norm": 0.8, "loss": 0.6039770245552063, "grad_norm": 1.002167820930481, "learning_rate": 9.59622463049436e-05}
+{"ts": "2025-12-24T04:00:26", "event": "train_log", "step": 2764, "epoch": 1.1924487594390507, "progress_pct": 19.87, "epoch_pct": 19.87, "eta": "54:54:17", "max_grad_norm": 0.8, "loss": 0.6543651819229126, "grad_norm": 1.150636076927185, "learning_rate": 9.595257447638343e-05}
+{"ts": "2025-12-24T04:00:46", "event": "train_log", "step": 2766, "epoch": 1.1933117583603021, "progress_pct": 19.89, "epoch_pct": 19.89, "eta": "54:52:36", "max_grad_norm": 0.8, "loss": 0.7388567328453064, "grad_norm": 1.413001537322998, "learning_rate": 9.594289156656709e-05}
+{"ts": "2025-12-24T04:01:05", "event": "train_log", "step": 2768, "epoch": 1.1941747572815533, "progress_pct": 19.9, "epoch_pct": 19.9, "eta": "54:50:56", "max_grad_norm": 0.8, "loss": 0.7128161191940308, "grad_norm": 1.2041958570480347, "learning_rate": 9.593319757782955e-05}
+{"ts": "2025-12-24T04:01:25", "event": "train_log", "step": 2770, "epoch": 1.1950377562028047, "progress_pct": 19.92, "epoch_pct": 19.92, "eta": "54:49:18", "max_grad_norm": 0.8, "loss": 0.6593868136405945, "grad_norm": 1.2151374816894531, "learning_rate": 9.592349251250847e-05}
+{"ts": "2025-12-24T04:01:44", "event": "train_log", "step": 2772, "epoch": 1.1959007551240561, "progress_pct": 19.93, "epoch_pct": 19.93, "eta": "54:47:37", "max_grad_norm": 0.8, "loss": 0.6128022074699402, "grad_norm": 1.6225353479385376, "learning_rate": 9.59137763729442e-05}
+{"ts": "2025-12-24T04:02:04", "event": "train_log", "step": 2774, "epoch": 1.1967637540453073, "progress_pct": 19.95, "epoch_pct": 19.95, "eta": "54:45:59", "max_grad_norm": 0.8, "loss": 0.6537567377090454, "grad_norm": 1.1904171705245972, "learning_rate": 9.590404916147976e-05}
+{"ts": "2025-12-24T04:02:24", "event": "train_log", "step": 2776, "epoch": 1.1976267529665587, "progress_pct": 19.96, "epoch_pct": 19.96, "eta": "54:44:22", "max_grad_norm": 0.8, "loss": 0.6325924396514893, "grad_norm": 1.159506916999817, "learning_rate": 9.589431088046079e-05}
+{"ts": "2025-12-24T04:02:44", "event": "train_log", "step": 2778, "epoch": 1.1984897518878102, "progress_pct": 19.97, "epoch_pct": 19.97, "eta": "54:42:44", "max_grad_norm": 0.8, "loss": 0.6438963413238525, "grad_norm": 1.1314560174942017, "learning_rate": 9.588456153223565e-05}
+{"ts": "2025-12-24T04:03:04", "event": "train_log", "step": 2780, "epoch": 1.1993527508090616, "progress_pct": 19.99, "epoch_pct": 19.99, "eta": "54:41:07", "max_grad_norm": 0.8, "loss": 0.6703475117683411, "grad_norm": 1.0182478427886963, "learning_rate": 9.587480111915532e-05}
+{"ts": "2025-12-24T04:03:20", "event": "train_log", "step": 2782, "epoch": 1.2002157497303128, "progress_pct": 20.0, "epoch_pct": 20.0, "eta": "54:39:16", "max_grad_norm": 0.8, "loss": 0.7694190144538879, "grad_norm": 1.5506517887115479, "learning_rate": 9.586502964357352e-05}
+{"ts": "2025-12-24T04:03:40", "event": "train_log", "step": 2784, "epoch": 1.2010787486515642, "progress_pct": 20.02, "epoch_pct": 20.02, "eta": "54:37:37", "max_grad_norm": 0.8, "loss": 0.6117724180221558, "grad_norm": 1.0769826173782349, "learning_rate": 9.585524710784659e-05}
+{"ts": "2025-12-24T04:03:58", "event": "train_log", "step": 2786, "epoch": 1.2019417475728156, "progress_pct": 20.03, "epoch_pct": 20.03, "eta": "54:35:53", "max_grad_norm": 0.8, "loss": 0.6375927329063416, "grad_norm": 1.207175850868225, "learning_rate": 9.584545351433352e-05}
+{"ts": "2025-12-24T04:04:17", "event": "train_log", "step": 2788, "epoch": 1.202804746494067, "progress_pct": 20.05, "epoch_pct": 20.05, "eta": "54:34:14", "max_grad_norm": 0.8, "loss": 0.6526537537574768, "grad_norm": 1.1291100978851318, "learning_rate": 9.583564886539602e-05}
+{"ts": "2025-12-24T04:04:36", "event": "train_log", "step": 2790, "epoch": 1.2036677454153182, "progress_pct": 20.06, "epoch_pct": 20.06, "eta": "54:32:33", "max_grad_norm": 0.8, "loss": 0.7080113887786865, "grad_norm": 1.3076926469802856, "learning_rate": 9.582583316339843e-05}
+{"ts": "2025-12-24T04:04:55", "event": "train_log", "step": 2792, "epoch": 1.2045307443365696, "progress_pct": 20.07, "epoch_pct": 20.08, "eta": "54:30:53", "max_grad_norm": 0.8, "loss": 0.6415606141090393, "grad_norm": 1.166823148727417, "learning_rate": 9.581600641070773e-05}
+{"ts": "2025-12-24T04:05:14", "event": "train_log", "step": 2794, "epoch": 1.205393743257821, "progress_pct": 20.09, "epoch_pct": 20.09, "eta": "54:29:14", "max_grad_norm": 0.8, "loss": 0.6538968086242676, "grad_norm": 1.1248292922973633, "learning_rate": 9.580616860969364e-05}
+{"ts": "2025-12-24T04:05:34", "event": "train_log", "step": 2796, "epoch": 1.2062567421790722, "progress_pct": 20.1, "epoch_pct": 20.1, "eta": "54:27:36", "max_grad_norm": 0.8, "loss": 0.694507896900177, "grad_norm": 1.406731128692627, "learning_rate": 9.579631976272849e-05}
+{"ts": "2025-12-24T04:05:52", "event": "train_log", "step": 2798, "epoch": 1.2071197411003236, "progress_pct": 20.12, "epoch_pct": 20.12, "eta": "54:25:53", "max_grad_norm": 0.8, "loss": 0.7064871788024902, "grad_norm": 1.185895562171936, "learning_rate": 9.578645987218728e-05}
+{"ts": "2025-12-24T04:06:12", "event": "train_log", "step": 2800, "epoch": 1.207982740021575, "progress_pct": 20.13, "epoch_pct": 20.13, "eta": "54:24:15", "max_grad_norm": 0.8, "loss": 0.5790154933929443, "grad_norm": 1.0860708951950073, "learning_rate": 9.577658894044768e-05}
+{"ts": "2025-12-24T04:20:03", "event": "train_log", "step": 2800, "epoch": 1.207982740021575, "progress_pct": 20.13, "epoch_pct": 20.13, "eta": "55:19:14", "max_grad_norm": 0.8, "eval_loss": 0.7220398187637329, "eval_runtime": 831.5812, "eval_samples_per_second": 2.477, "eval_steps_per_second": 2.477}
+{"ts": "2025-12-24T04:20:23", "event": "train_log", "step": 2802, "epoch": 1.2088457389428262, "progress_pct": 20.15, "epoch_pct": 20.15, "eta": "55:17:34", "max_grad_norm": 0.8, "loss": 0.6645335555076599, "grad_norm": 1.2761744260787964, "learning_rate": 9.576670696989001e-05}
+{"ts": "2025-12-24T04:20:44", "event": "train_log", "step": 2804, "epoch": 1.2097087378640776, "progress_pct": 20.16, "epoch_pct": 20.16, "eta": "55:15:58", "max_grad_norm": 0.8, "loss": 0.6320346593856812, "grad_norm": 1.3502917289733887, "learning_rate": 9.575681396289728e-05}
+{"ts": "2025-12-24T04:21:02", "event": "train_log", "step": 2806, "epoch": 1.210571736785329, "progress_pct": 20.18, "epoch_pct": 20.18, "eta": "55:14:14", "max_grad_norm": 0.8, "loss": 0.6297362446784973, "grad_norm": 1.2638593912124634, "learning_rate": 9.574690992185514e-05}
+{"ts": "2025-12-24T04:21:22", "event": "train_log", "step": 2808, "epoch": 1.2114347357065804, "progress_pct": 20.19, "epoch_pct": 20.19, "eta": "55:12:37", "max_grad_norm": 0.8, "loss": 0.7569923996925354, "grad_norm": 1.0787774324417114, "learning_rate": 9.573699484915189e-05}
+{"ts": "2025-12-24T04:21:42", "event": "train_log", "step": 2810, "epoch": 1.2122977346278316, "progress_pct": 20.2, "epoch_pct": 20.2, "eta": "55:10:58", "max_grad_norm": 0.8, "loss": 0.6712964773178101, "grad_norm": 1.097254991531372, "learning_rate": 9.572706874717851e-05}
+{"ts": "2025-12-24T04:22:04", "event": "train_log", "step": 2812, "epoch": 1.213160733549083, "progress_pct": 20.22, "epoch_pct": 20.22, "eta": "55:09:24", "max_grad_norm": 0.8, "loss": 0.6775450706481934, "grad_norm": 1.0995726585388184, "learning_rate": 9.571713161832865e-05}
+{"ts": "2025-12-24T04:22:23", "event": "train_log", "step": 2814, "epoch": 1.2140237324703345, "progress_pct": 20.23, "epoch_pct": 20.23, "eta": "55:07:43", "max_grad_norm": 0.8, "loss": 0.6691418290138245, "grad_norm": 1.7386797666549683, "learning_rate": 9.570718346499859e-05}
+{"ts": "2025-12-24T04:22:42", "event": "train_log", "step": 2816, "epoch": 1.2148867313915859, "progress_pct": 20.25, "epoch_pct": 20.25, "eta": "55:06:04", "max_grad_norm": 0.8, "loss": 0.6406403183937073, "grad_norm": 1.2281049489974976, "learning_rate": 9.569722428958728e-05}
+{"ts": "2025-12-24T04:23:02", "event": "train_log", "step": 2818, "epoch": 1.215749730312837, "progress_pct": 20.26, "epoch_pct": 20.26, "eta": "55:04:24", "max_grad_norm": 0.8, "loss": 0.7008584141731262, "grad_norm": 1.1326305866241455, "learning_rate": 9.568725409449635e-05}
+{"ts": "2025-12-24T04:23:21", "event": "train_log", "step": 2820, "epoch": 1.2166127292340885, "progress_pct": 20.28, "epoch_pct": 20.28, "eta": "55:02:42", "max_grad_norm": 0.8, "loss": 0.7438415288925171, "grad_norm": 1.3039381504058838, "learning_rate": 9.567727288213005e-05}
+{"ts": "2025-12-24T04:23:40", "event": "train_log", "step": 2822, "epoch": 1.2174757281553399, "progress_pct": 20.29, "epoch_pct": 20.29, "eta": "55:01:03", "max_grad_norm": 0.8, "loss": 0.6937777996063232, "grad_norm": 1.219136357307434, "learning_rate": 9.566728065489531e-05}
+{"ts": "2025-12-24T04:23:58", "event": "train_log", "step": 2824, "epoch": 1.218338727076591, "progress_pct": 20.3, "epoch_pct": 20.31, "eta": "54:59:15", "max_grad_norm": 0.8, "loss": 0.7586468458175659, "grad_norm": 1.321285605430603, "learning_rate": 9.56572774152017e-05}
+{"ts": "2025-12-24T04:24:18", "event": "train_log", "step": 2826, "epoch": 1.2192017259978425, "progress_pct": 20.32, "epoch_pct": 20.32, "eta": "54:57:38", "max_grad_norm": 0.8, "loss": 0.6356340050697327, "grad_norm": 1.3538484573364258, "learning_rate": 9.564726316546147e-05}
+{"ts": "2025-12-24T04:24:37", "event": "train_log", "step": 2828, "epoch": 1.220064724919094, "progress_pct": 20.33, "epoch_pct": 20.33, "eta": "54:55:59", "max_grad_norm": 0.8, "loss": 0.6304616928100586, "grad_norm": 1.1015816926956177, "learning_rate": 9.563723790808952e-05}
+{"ts": "2025-12-24T04:25:00", "event": "train_log", "step": 2830, "epoch": 1.220927723840345, "progress_pct": 20.35, "epoch_pct": 20.35, "eta": "54:54:30", "max_grad_norm": 0.8, "loss": 0.6459048986434937, "grad_norm": 0.9658390283584595, "learning_rate": 9.562720164550337e-05}
+{"ts": "2025-12-24T04:25:18", "event": "train_log", "step": 2832, "epoch": 1.2217907227615965, "progress_pct": 20.36, "epoch_pct": 20.36, "eta": "54:52:46", "max_grad_norm": 0.8, "loss": 0.7034182548522949, "grad_norm": 1.3981443643569946, "learning_rate": 9.561715438012323e-05}
+{"ts": "2025-12-24T04:25:36", "event": "train_log", "step": 2834, "epoch": 1.222653721682848, "progress_pct": 20.38, "epoch_pct": 20.38, "eta": "54:51:04", "max_grad_norm": 0.8, "loss": 0.6629860997200012, "grad_norm": 1.1737093925476074, "learning_rate": 9.560709611437195e-05}
+{"ts": "2025-12-24T04:25:56", "event": "train_log", "step": 2836, "epoch": 1.2235167206040993, "progress_pct": 20.39, "epoch_pct": 20.39, "eta": "54:49:25", "max_grad_norm": 0.8, "loss": 0.7166537642478943, "grad_norm": 1.2071303129196167, "learning_rate": 9.559702685067506e-05}
+{"ts": "2025-12-24T04:26:16", "event": "train_log", "step": 2838, "epoch": 1.2243797195253505, "progress_pct": 20.41, "epoch_pct": 20.41, "eta": "54:47:48", "max_grad_norm": 0.8, "loss": 0.6711867451667786, "grad_norm": 1.2520930767059326, "learning_rate": 9.558694659146068e-05}
+{"ts": "2025-12-24T04:26:34", "event": "train_log", "step": 2840, "epoch": 1.225242718446602, "progress_pct": 20.42, "epoch_pct": 20.42, "eta": "54:46:06", "max_grad_norm": 0.8, "loss": 0.7135099768638611, "grad_norm": 1.2304543256759644, "learning_rate": 9.557685533915964e-05}
+{"ts": "2025-12-24T04:26:54", "event": "train_log", "step": 2842, "epoch": 1.2261057173678533, "progress_pct": 20.43, "epoch_pct": 20.44, "eta": "54:44:26", "max_grad_norm": 0.8, "loss": 0.6279205083847046, "grad_norm": 1.1509613990783691, "learning_rate": 9.556675309620539e-05}
+{"ts": "2025-12-24T04:27:13", "event": "train_log", "step": 2844, "epoch": 1.2269687162891048, "progress_pct": 20.45, "epoch_pct": 20.45, "eta": "54:42:48", "max_grad_norm": 0.8, "loss": 0.6848570704460144, "grad_norm": 1.2024176120758057, "learning_rate": 9.555663986503406e-05}
+{"ts": "2025-12-24T04:27:31", "event": "train_log", "step": 2846, "epoch": 1.227831715210356, "progress_pct": 20.46, "epoch_pct": 20.46, "eta": "54:41:04", "max_grad_norm": 0.8, "loss": 0.7109206914901733, "grad_norm": 1.360257863998413, "learning_rate": 9.55465156480844e-05}
+{"ts": "2025-12-24T04:27:50", "event": "train_log", "step": 2848, "epoch": 1.2286947141316074, "progress_pct": 20.48, "epoch_pct": 20.48, "eta": "54:39:25", "max_grad_norm": 0.8, "loss": 0.5983237028121948, "grad_norm": 1.1556848287582397, "learning_rate": 9.55363804477978e-05}
+{"ts": "2025-12-24T04:28:09", "event": "train_log", "step": 2850, "epoch": 1.2295577130528588, "progress_pct": 20.49, "epoch_pct": 20.49, "eta": "54:37:46", "max_grad_norm": 0.8, "loss": 0.6387401223182678, "grad_norm": 1.131439447402954, "learning_rate": 9.552623426661836e-05}
+{"ts": "2025-12-24T04:28:29", "event": "train_log", "step": 2852, "epoch": 1.23042071197411, "progress_pct": 20.51, "epoch_pct": 20.51, "eta": "54:36:08", "max_grad_norm": 0.8, "loss": 0.6584593057632446, "grad_norm": 1.1854766607284546, "learning_rate": 9.551607710699273e-05}
+{"ts": "2025-12-24T04:28:49", "event": "train_log", "step": 2854, "epoch": 1.2312837108953614, "progress_pct": 20.52, "epoch_pct": 20.52, "eta": "54:34:32", "max_grad_norm": 0.8, "loss": 0.6284405589103699, "grad_norm": 1.1811782121658325, "learning_rate": 9.55059089713703e-05}
+{"ts": "2025-12-24T04:29:08", "event": "train_log", "step": 2856, "epoch": 1.2321467098166128, "progress_pct": 20.53, "epoch_pct": 20.54, "eta": "54:32:53", "max_grad_norm": 0.8, "loss": 0.6183417439460754, "grad_norm": 1.1038708686828613, "learning_rate": 9.549572986220307e-05}
+{"ts": "2025-12-24T04:29:27", "event": "train_log", "step": 2858, "epoch": 1.233009708737864, "progress_pct": 20.55, "epoch_pct": 20.55, "eta": "54:31:12", "max_grad_norm": 0.8, "loss": 0.7266764640808105, "grad_norm": 1.3645228147506714, "learning_rate": 9.548553978194567e-05}
+{"ts": "2025-12-24T04:29:46", "event": "train_log", "step": 2860, "epoch": 1.2338727076591154, "progress_pct": 20.56, "epoch_pct": 20.56, "eta": "54:29:33", "max_grad_norm": 0.8, "loss": 0.6580738425254822, "grad_norm": 1.1896798610687256, "learning_rate": 9.547533873305542e-05}
+{"ts": "2025-12-24T04:30:06", "event": "train_log", "step": 2862, "epoch": 1.2347357065803668, "progress_pct": 20.58, "epoch_pct": 20.58, "eta": "54:27:57", "max_grad_norm": 0.8, "loss": 0.669810950756073, "grad_norm": 1.1351168155670166, "learning_rate": 9.546512671799224e-05}
+{"ts": "2025-12-24T04:30:25", "event": "train_log", "step": 2864, "epoch": 1.2355987055016182, "progress_pct": 20.59, "epoch_pct": 20.59, "eta": "54:26:18", "max_grad_norm": 0.8, "loss": 0.730329155921936, "grad_norm": 1.1937936544418335, "learning_rate": 9.545490373921869e-05}
+{"ts": "2025-12-24T04:30:44", "event": "train_log", "step": 2866, "epoch": 1.2364617044228694, "progress_pct": 20.61, "epoch_pct": 20.61, "eta": "54:24:40", "max_grad_norm": 0.8, "loss": 0.6867446303367615, "grad_norm": 1.094347357749939, "learning_rate": 9.544466979920002e-05}
+{"ts": "2025-12-24T04:31:03", "event": "train_log", "step": 2868, "epoch": 1.2373247033441208, "progress_pct": 20.62, "epoch_pct": 20.62, "eta": "54:23:00", "max_grad_norm": 0.8, "loss": 0.735468327999115, "grad_norm": 1.4509600400924683, "learning_rate": 9.54344249004041e-05}
+{"ts": "2025-12-24T04:31:23", "event": "train_log", "step": 2870, "epoch": 1.2381877022653722, "progress_pct": 20.64, "epoch_pct": 20.64, "eta": "54:21:25", "max_grad_norm": 0.8, "loss": 0.6316167116165161, "grad_norm": 1.1195857524871826, "learning_rate": 9.542416904530142e-05}
+{"ts": "2025-12-24T04:31:42", "event": "train_log", "step": 2872, "epoch": 1.2390507011866236, "progress_pct": 20.65, "epoch_pct": 20.65, "eta": "54:19:49", "max_grad_norm": 0.8, "loss": 0.6637873649597168, "grad_norm": 1.0227746963500977, "learning_rate": 9.541390223636517e-05}
+{"ts": "2025-12-24T04:32:01", "event": "train_log", "step": 2874, "epoch": 1.2399137001078748, "progress_pct": 20.66, "epoch_pct": 20.67, "eta": "54:18:09", "max_grad_norm": 0.8, "loss": 0.6579877734184265, "grad_norm": 1.1871414184570312, "learning_rate": 9.540362447607111e-05}
+{"ts": "2025-12-24T04:32:19", "event": "train_log", "step": 2876, "epoch": 1.2407766990291262, "progress_pct": 20.68, "epoch_pct": 20.68, "eta": "54:16:26", "max_grad_norm": 0.8, "loss": 0.7563158273696899, "grad_norm": 1.3265694379806519, "learning_rate": 9.539333576689768e-05}
+{"ts": "2025-12-24T04:32:40", "event": "train_log", "step": 2878, "epoch": 1.2416396979503777, "progress_pct": 20.69, "epoch_pct": 20.69, "eta": "54:14:54", "max_grad_norm": 0.8, "loss": 0.6373291015625, "grad_norm": 1.0074095726013184, "learning_rate": 9.538303611132598e-05}
+{"ts": "2025-12-24T04:32:58", "event": "train_log", "step": 2880, "epoch": 1.2425026968716288, "progress_pct": 20.71, "epoch_pct": 20.71, "eta": "54:13:13", "max_grad_norm": 0.8, "loss": 0.6989303827285767, "grad_norm": 1.421204924583435, "learning_rate": 9.537272551183971e-05}
+{"ts": "2025-12-24T04:33:17", "event": "train_log", "step": 2882, "epoch": 1.2433656957928803, "progress_pct": 20.72, "epoch_pct": 20.72, "eta": "54:11:36", "max_grad_norm": 0.8, "loss": 0.7590261697769165, "grad_norm": 1.256486415863037, "learning_rate": 9.536240397092522e-05}
+{"ts": "2025-12-24T04:33:37", "event": "train_log", "step": 2884, "epoch": 1.2442286947141317, "progress_pct": 20.74, "epoch_pct": 20.74, "eta": "54:10:00", "max_grad_norm": 0.8, "loss": 0.6652005314826965, "grad_norm": 1.1513274908065796, "learning_rate": 9.53520714910715e-05}
+{"ts": "2025-12-24T04:33:57", "event": "train_log", "step": 2886, "epoch": 1.2450916936353829, "progress_pct": 20.75, "epoch_pct": 20.75, "eta": "54:08:26", "max_grad_norm": 0.8, "loss": 0.6510938405990601, "grad_norm": 1.2129496335983276, "learning_rate": 9.53417280747702e-05}
+{"ts": "2025-12-24T04:34:16", "event": "train_log", "step": 2888, "epoch": 1.2459546925566343, "progress_pct": 20.77, "epoch_pct": 20.77, "eta": "54:06:51", "max_grad_norm": 0.8, "loss": 0.6405928730964661, "grad_norm": 1.1393296718597412, "learning_rate": 9.533137372451557e-05}
+{"ts": "2025-12-24T04:34:35", "event": "train_log", "step": 2890, "epoch": 1.2468176914778857, "progress_pct": 20.78, "epoch_pct": 20.78, "eta": "54:05:12", "max_grad_norm": 0.8, "loss": 0.6819307208061218, "grad_norm": 1.1467162370681763, "learning_rate": 9.532100844280451e-05}
+{"ts": "2025-12-24T04:34:54", "event": "train_log", "step": 2892, "epoch": 1.247680690399137, "progress_pct": 20.79, "epoch_pct": 20.79, "eta": "54:03:35", "max_grad_norm": 0.8, "loss": 0.6333560347557068, "grad_norm": 1.1516613960266113, "learning_rate": 9.531063223213659e-05}
+{"ts": "2025-12-24T04:35:12", "event": "train_log", "step": 2894, "epoch": 1.2485436893203883, "progress_pct": 20.81, "epoch_pct": 20.81, "eta": "54:01:52", "max_grad_norm": 0.8, "loss": 0.7259839177131653, "grad_norm": 1.463189959526062, "learning_rate": 9.530024509501396e-05}
+{"ts": "2025-12-24T04:35:31", "event": "train_log", "step": 2896, "epoch": 1.2494066882416397, "progress_pct": 20.82, "epoch_pct": 20.82, "eta": "54:00:16", "max_grad_norm": 0.8, "loss": 0.661693811416626, "grad_norm": 1.0450547933578491, "learning_rate": 9.528984703394144e-05}
+{"ts": "2025-12-24T04:35:52", "event": "train_log", "step": 2898, "epoch": 1.250269687162891, "progress_pct": 20.84, "epoch_pct": 20.84, "eta": "53:58:45", "max_grad_norm": 0.8, "loss": 0.5889303684234619, "grad_norm": 1.0525774955749512, "learning_rate": 9.527943805142648e-05}
+{"ts": "2025-12-24T04:36:11", "event": "train_log", "step": 2900, "epoch": 1.2511326860841425, "progress_pct": 20.85, "epoch_pct": 20.85, "eta": "53:57:09", "max_grad_norm": 0.8, "loss": 0.6380610466003418, "grad_norm": 1.130048394203186, "learning_rate": 9.526901814997914e-05}
+{"ts": "2025-12-24T04:50:12", "event": "train_log", "step": 2900, "epoch": 1.2511326860841425, "progress_pct": 20.85, "epoch_pct": 20.85, "eta": "54:50:21", "max_grad_norm": 0.8, "eval_loss": 0.7182289958000183, "eval_runtime": 840.8156, "eval_samples_per_second": 2.45, "eval_steps_per_second": 2.45}
+{"ts": "2025-12-24T04:50:32", "event": "train_log", "step": 2902, "epoch": 1.2519956850053937, "progress_pct": 20.87, "epoch_pct": 20.87, "eta": "54:48:46", "max_grad_norm": 0.8, "loss": 0.6397356986999512, "grad_norm": 1.0194777250289917, "learning_rate": 9.525858733211217e-05}
+{"ts": "2025-12-24T04:50:51", "event": "train_log", "step": 2904, "epoch": 1.2528586839266451, "progress_pct": 20.88, "epoch_pct": 20.88, "eta": "54:47:04", "max_grad_norm": 0.8, "loss": 0.7931577563285828, "grad_norm": 1.3902885913848877, "learning_rate": 9.524814560034087e-05}
+{"ts": "2025-12-24T04:51:10", "event": "train_log", "step": 2906, "epoch": 1.2537216828478965, "progress_pct": 20.89, "epoch_pct": 20.9, "eta": "54:45:24", "max_grad_norm": 0.8, "loss": 0.6997087597846985, "grad_norm": 1.258169412612915, "learning_rate": 9.523769295718325e-05}
+{"ts": "2025-12-24T04:51:28", "event": "train_log", "step": 2908, "epoch": 1.2545846817691477, "progress_pct": 20.91, "epoch_pct": 20.91, "eta": "54:43:41", "max_grad_norm": 0.8, "loss": 0.6946825385093689, "grad_norm": 1.3403711318969727, "learning_rate": 9.522722940515991e-05}
+{"ts": "2025-12-24T04:51:47", "event": "train_log", "step": 2910, "epoch": 1.2554476806903991, "progress_pct": 20.92, "epoch_pct": 20.92, "eta": "54:42:02", "max_grad_norm": 0.8, "loss": 0.6309347748756409, "grad_norm": 1.1724950075149536, "learning_rate": 9.521675494679409e-05}
+{"ts": "2025-12-24T04:52:05", "event": "train_log", "step": 2912, "epoch": 1.2563106796116505, "progress_pct": 20.94, "epoch_pct": 20.94, "eta": "54:40:21", "max_grad_norm": 0.8, "loss": 0.6453542113304138, "grad_norm": 1.210274338722229, "learning_rate": 9.520626958461163e-05}
+{"ts": "2025-12-24T04:52:24", "event": "train_log", "step": 2914, "epoch": 1.2571736785329017, "progress_pct": 20.95, "epoch_pct": 20.95, "eta": "54:38:42", "max_grad_norm": 0.8, "loss": 0.6747867465019226, "grad_norm": 1.233112096786499, "learning_rate": 9.519577332114106e-05}
+{"ts": "2025-12-24T04:52:44", "event": "train_log", "step": 2916, "epoch": 1.2580366774541532, "progress_pct": 20.97, "epoch_pct": 20.97, "eta": "54:37:05", "max_grad_norm": 0.8, "loss": 0.6729235053062439, "grad_norm": 1.1785951852798462, "learning_rate": 9.51852661589135e-05}
+{"ts": "2025-12-24T04:53:02", "event": "train_log", "step": 2918, "epoch": 1.2588996763754046, "progress_pct": 20.98, "epoch_pct": 20.98, "eta": "54:35:24", "max_grad_norm": 0.8, "loss": 0.6581047177314758, "grad_norm": 1.0781456232070923, "learning_rate": 9.51747481004627e-05}
+{"ts": "2025-12-24T04:53:22", "event": "train_log", "step": 2920, "epoch": 1.2597626752966558, "progress_pct": 21.0, "epoch_pct": 21.0, "eta": "54:33:48", "max_grad_norm": 0.8, "loss": 0.6576398015022278, "grad_norm": 1.1122184991836548, "learning_rate": 9.516421914832506e-05}
+{"ts": "2025-12-24T04:53:41", "event": "train_log", "step": 2922, "epoch": 1.2606256742179072, "progress_pct": 21.01, "epoch_pct": 21.01, "eta": "54:32:08", "max_grad_norm": 0.8, "loss": 0.7026275396347046, "grad_norm": 1.3378703594207764, "learning_rate": 9.515367930503956e-05}
+{"ts": "2025-12-24T04:54:00", "event": "train_log", "step": 2924, "epoch": 1.2614886731391586, "progress_pct": 21.02, "epoch_pct": 21.02, "eta": "54:30:30", "max_grad_norm": 0.8, "loss": 0.6111703515052795, "grad_norm": 1.127657175064087, "learning_rate": 9.514312857314785e-05}
+{"ts": "2025-12-24T04:54:18", "event": "train_log", "step": 2926, "epoch": 1.26235167206041, "progress_pct": 21.04, "epoch_pct": 21.04, "eta": "54:28:50", "max_grad_norm": 0.8, "loss": 0.6698760986328125, "grad_norm": 1.2058607339859009, "learning_rate": 9.513256695519419e-05}
+{"ts": "2025-12-24T04:54:38", "event": "train_log", "step": 2928, "epoch": 1.2632146709816614, "progress_pct": 21.05, "epoch_pct": 21.05, "eta": "54:27:14", "max_grad_norm": 0.8, "loss": 0.6793603897094727, "grad_norm": 1.2206004858016968, "learning_rate": 9.512199445372546e-05}
+{"ts": "2025-12-24T04:54:57", "event": "train_log", "step": 2930, "epoch": 1.2640776699029126, "progress_pct": 21.07, "epoch_pct": 21.07, "eta": "54:25:36", "max_grad_norm": 0.8, "loss": 0.709406316280365, "grad_norm": 1.167311668395996, "learning_rate": 9.511141107129119e-05}
+{"ts": "2025-12-24T04:55:16", "event": "train_log", "step": 2932, "epoch": 1.264940668824164, "progress_pct": 21.08, "epoch_pct": 21.08, "eta": "54:23:57", "max_grad_norm": 0.8, "loss": 0.685570240020752, "grad_norm": 1.3347517251968384, "learning_rate": 9.51008168104435e-05}
+{"ts": "2025-12-24T04:55:35", "event": "train_log", "step": 2934, "epoch": 1.2658036677454154, "progress_pct": 21.1, "epoch_pct": 21.1, "eta": "54:22:19", "max_grad_norm": 0.8, "loss": 0.6775280833244324, "grad_norm": 1.186468482017517, "learning_rate": 9.509021167373714e-05}
+{"ts": "2025-12-24T04:55:54", "event": "train_log", "step": 2936, "epoch": 1.2666666666666666, "progress_pct": 21.11, "epoch_pct": 21.11, "eta": "54:20:40", "max_grad_norm": 0.8, "loss": 0.6587507128715515, "grad_norm": 1.3934271335601807, "learning_rate": 9.507959566372953e-05}
+{"ts": "2025-12-24T04:56:12", "event": "train_log", "step": 2938, "epoch": 1.267529665587918, "progress_pct": 21.12, "epoch_pct": 21.13, "eta": "54:18:58", "max_grad_norm": 0.8, "loss": 0.692707896232605, "grad_norm": 1.3766281604766846, "learning_rate": 9.506896878298061e-05}
+{"ts": "2025-12-24T04:56:31", "event": "train_log", "step": 2940, "epoch": 1.2683926645091694, "progress_pct": 21.14, "epoch_pct": 21.14, "eta": "54:17:21", "max_grad_norm": 0.8, "loss": 0.6481655836105347, "grad_norm": 1.034883975982666, "learning_rate": 9.505833103405307e-05}
+{"ts": "2025-12-24T04:56:50", "event": "train_log", "step": 2942, "epoch": 1.2692556634304206, "progress_pct": 21.15, "epoch_pct": 21.15, "eta": "54:15:45", "max_grad_norm": 0.8, "loss": 0.5867606997489929, "grad_norm": 1.113044261932373, "learning_rate": 9.504768241951212e-05}
+{"ts": "2025-12-24T04:57:10", "event": "train_log", "step": 2944, "epoch": 1.270118662351672, "progress_pct": 21.17, "epoch_pct": 21.17, "eta": "54:14:09", "max_grad_norm": 0.8, "loss": 0.6147398948669434, "grad_norm": 1.0769455432891846, "learning_rate": 9.503702294192563e-05}
+{"ts": "2025-12-24T04:57:29", "event": "train_log", "step": 2946, "epoch": 1.2709816612729234, "progress_pct": 21.18, "epoch_pct": 21.18, "eta": "54:12:33", "max_grad_norm": 0.8, "loss": 0.650299608707428, "grad_norm": 1.100616216659546, "learning_rate": 9.502635260386408e-05}
+{"ts": "2025-12-24T04:57:49", "event": "train_log", "step": 2948, "epoch": 1.2718446601941746, "progress_pct": 21.2, "epoch_pct": 21.2, "eta": "54:10:58", "max_grad_norm": 0.8, "loss": 0.6896426677703857, "grad_norm": 1.1982520818710327, "learning_rate": 9.501567140790057e-05}
+{"ts": "2025-12-24T04:58:07", "event": "train_log", "step": 2950, "epoch": 1.272707659115426, "progress_pct": 21.21, "epoch_pct": 21.21, "eta": "54:09:20", "max_grad_norm": 0.8, "loss": 0.6780133247375488, "grad_norm": 1.276911735534668, "learning_rate": 9.500497935661085e-05}
+{"ts": "2025-12-24T04:58:26", "event": "train_log", "step": 2952, "epoch": 1.2735706580366775, "progress_pct": 21.23, "epoch_pct": 21.23, "eta": "54:07:42", "max_grad_norm": 0.8, "loss": 0.6906163692474365, "grad_norm": 1.131345272064209, "learning_rate": 9.499427645257323e-05}
+{"ts": "2025-12-24T04:58:46", "event": "train_log", "step": 2954, "epoch": 1.2744336569579289, "progress_pct": 21.24, "epoch_pct": 21.24, "eta": "54:06:08", "max_grad_norm": 0.8, "loss": 0.6418634057044983, "grad_norm": 1.125251293182373, "learning_rate": 9.498356269836868e-05}
+{"ts": "2025-12-24T04:59:05", "event": "train_log", "step": 2956, "epoch": 1.2752966558791803, "progress_pct": 21.25, "epoch_pct": 21.25, "eta": "54:04:32", "max_grad_norm": 0.8, "loss": 0.690779983997345, "grad_norm": 1.2508443593978882, "learning_rate": 9.497283809658078e-05}
+{"ts": "2025-12-24T04:59:25", "event": "train_log", "step": 2958, "epoch": 1.2761596548004315, "progress_pct": 21.27, "epoch_pct": 21.27, "eta": "54:02:58", "max_grad_norm": 0.8, "loss": 0.645685613155365, "grad_norm": 0.9558717608451843, "learning_rate": 9.49621026497957e-05}
+{"ts": "2025-12-24T04:59:44", "event": "train_log", "step": 2960, "epoch": 1.2770226537216829, "progress_pct": 21.28, "epoch_pct": 21.28, "eta": "54:01:20", "max_grad_norm": 0.8, "loss": 0.6313972473144531, "grad_norm": 1.1781591176986694, "learning_rate": 9.495135636060226e-05}
+{"ts": "2025-12-24T05:00:03", "event": "train_log", "step": 2962, "epoch": 1.2778856526429343, "progress_pct": 21.3, "epoch_pct": 21.3, "eta": "53:59:44", "max_grad_norm": 0.8, "loss": 0.5853794813156128, "grad_norm": 0.9713736772537231, "learning_rate": 9.494059923159189e-05}
+{"ts": "2025-12-24T05:00:21", "event": "train_log", "step": 2964, "epoch": 1.2787486515641855, "progress_pct": 21.31, "epoch_pct": 21.31, "eta": "53:58:05", "max_grad_norm": 0.8, "loss": 0.6966362595558167, "grad_norm": 1.3793153762817383, "learning_rate": 9.492983126535859e-05}
+{"ts": "2025-12-24T05:00:39", "event": "train_log", "step": 2966, "epoch": 1.279611650485437, "progress_pct": 21.33, "epoch_pct": 21.33, "eta": "53:56:25", "max_grad_norm": 0.8, "loss": 0.7281587719917297, "grad_norm": 1.250563383102417, "learning_rate": 9.491905246449903e-05}
+{"ts": "2025-12-24T05:00:58", "event": "train_log", "step": 2968, "epoch": 1.2804746494066883, "progress_pct": 21.34, "epoch_pct": 21.34, "eta": "53:54:46", "max_grad_norm": 0.8, "loss": 0.7322974801063538, "grad_norm": 1.3268930912017822, "learning_rate": 9.490826283161247e-05}
+{"ts": "2025-12-24T05:01:17", "event": "train_log", "step": 2970, "epoch": 1.2813376483279395, "progress_pct": 21.35, "epoch_pct": 21.36, "eta": "53:53:10", "max_grad_norm": 0.8, "loss": 0.7875081896781921, "grad_norm": 1.3876935243606567, "learning_rate": 9.489746236930077e-05}
+{"ts": "2025-12-24T05:01:37", "event": "train_log", "step": 2972, "epoch": 1.282200647249191, "progress_pct": 21.37, "epoch_pct": 21.37, "eta": "53:51:38", "max_grad_norm": 0.8, "loss": 0.601652979850769, "grad_norm": 1.02049720287323, "learning_rate": 9.488665108016842e-05}
+{"ts": "2025-12-24T05:01:57", "event": "train_log", "step": 2974, "epoch": 1.2830636461704423, "progress_pct": 21.38, "epoch_pct": 21.38, "eta": "53:50:05", "max_grad_norm": 0.8, "loss": 0.6255780458450317, "grad_norm": 1.0868756771087646, "learning_rate": 9.487582896682252e-05}
+{"ts": "2025-12-24T05:02:16", "event": "train_log", "step": 2976, "epoch": 1.2839266450916935, "progress_pct": 21.4, "epoch_pct": 21.4, "eta": "53:48:32", "max_grad_norm": 0.8, "loss": 0.5967418551445007, "grad_norm": 1.00597083568573, "learning_rate": 9.486499603187275e-05}
+{"ts": "2025-12-24T05:02:36", "event": "train_log", "step": 2978, "epoch": 1.284789644012945, "progress_pct": 21.41, "epoch_pct": 21.41, "eta": "53:46:59", "max_grad_norm": 0.8, "loss": 0.719573438167572, "grad_norm": 1.3068969249725342, "learning_rate": 9.485415227793145e-05}
+{"ts": "2025-12-24T05:02:54", "event": "train_log", "step": 2980, "epoch": 1.2856526429341963, "progress_pct": 21.43, "epoch_pct": 21.43, "eta": "53:45:21", "max_grad_norm": 0.8, "loss": 0.665522038936615, "grad_norm": 1.2478386163711548, "learning_rate": 9.484329770761353e-05}
+{"ts": "2025-12-24T05:03:15", "event": "train_log", "step": 2982, "epoch": 1.2865156418554478, "progress_pct": 21.44, "epoch_pct": 21.44, "eta": "53:43:52", "max_grad_norm": 0.8, "loss": 0.6798994541168213, "grad_norm": 1.1220918893814087, "learning_rate": 9.483243232353652e-05}
+{"ts": "2025-12-24T05:03:35", "event": "train_log", "step": 2984, "epoch": 1.287378640776699, "progress_pct": 21.46, "epoch_pct": 21.46, "eta": "53:42:20", "max_grad_norm": 0.8, "loss": 0.672002911567688, "grad_norm": 1.1369609832763672, "learning_rate": 9.482155612832057e-05}
+{"ts": "2025-12-24T05:03:56", "event": "train_log", "step": 2986, "epoch": 1.2882416396979504, "progress_pct": 21.47, "epoch_pct": 21.47, "eta": "53:40:50", "max_grad_norm": 0.8, "loss": 0.6755662560462952, "grad_norm": 1.2077715396881104, "learning_rate": 9.48106691245884e-05}
+{"ts": "2025-12-24T05:04:14", "event": "train_log", "step": 2988, "epoch": 1.2891046386192018, "progress_pct": 21.48, "epoch_pct": 21.49, "eta": "53:39:12", "max_grad_norm": 0.8, "loss": 0.7504231333732605, "grad_norm": 1.3299074172973633, "learning_rate": 9.479977131496539e-05}
+{"ts": "2025-12-24T05:04:33", "event": "train_log", "step": 2990, "epoch": 1.2899676375404532, "progress_pct": 21.5, "epoch_pct": 21.5, "eta": "53:37:37", "max_grad_norm": 0.8, "loss": 0.6837405562400818, "grad_norm": 1.08982253074646, "learning_rate": 9.478886270207948e-05}
+{"ts": "2025-12-24T05:04:53", "event": "train_log", "step": 2992, "epoch": 1.2908306364617044, "progress_pct": 21.51, "epoch_pct": 21.51, "eta": "53:36:05", "max_grad_norm": 0.8, "loss": 0.6552648544311523, "grad_norm": 1.1015464067459106, "learning_rate": 9.477794328856124e-05}
+{"ts": "2025-12-24T05:05:12", "event": "train_log", "step": 2994, "epoch": 1.2916936353829558, "progress_pct": 21.53, "epoch_pct": 21.53, "eta": "53:34:32", "max_grad_norm": 0.8, "loss": 0.6160483360290527, "grad_norm": 1.1793229579925537, "learning_rate": 9.476701307704384e-05}
+{"ts": "2025-12-24T05:05:32", "event": "train_log", "step": 2996, "epoch": 1.2925566343042072, "progress_pct": 21.54, "epoch_pct": 21.54, "eta": "53:33:00", "max_grad_norm": 0.8, "loss": 0.7406170964241028, "grad_norm": 1.4781638383865356, "learning_rate": 9.475607207016306e-05}
+{"ts": "2025-12-24T05:05:52", "event": "train_log", "step": 2998, "epoch": 1.2934196332254584, "progress_pct": 21.56, "epoch_pct": 21.56, "eta": "53:31:27", "max_grad_norm": 0.8, "loss": 0.5950883626937866, "grad_norm": 1.2391449213027954, "learning_rate": 9.474512027055724e-05}
+{"ts": "2025-12-24T05:06:11", "event": "train_log", "step": 3000, "epoch": 1.2942826321467098, "progress_pct": 21.57, "epoch_pct": 21.57, "eta": "53:29:55", "max_grad_norm": 0.8, "loss": 0.6587089896202087, "grad_norm": 1.166535496711731, "learning_rate": 9.473415768086738e-05}
+{"ts": "2025-12-24T05:20:02", "event": "train_log", "step": 3000, "epoch": 1.2942826321467098, "progress_pct": 21.57, "epoch_pct": 21.57, "eta": "54:20:16", "max_grad_norm": 0.8, "eval_loss": 0.7160727977752686, "eval_runtime": 830.8186, "eval_samples_per_second": 2.479, "eval_steps_per_second": 2.479}
+{"ts": "2025-12-24T05:20:22", "event": "train_log", "step": 3002, "epoch": 1.2951456310679612, "progress_pct": 21.58, "epoch_pct": 21.59, "eta": "54:18:41", "max_grad_norm": 0.8, "loss": 0.5802602767944336, "grad_norm": 0.9833650588989258, "learning_rate": 9.472318430373706e-05}
+{"ts": "2025-12-24T05:20:41", "event": "train_log", "step": 3004, "epoch": 1.2960086299892124, "progress_pct": 21.6, "epoch_pct": 21.6, "eta": "54:17:04", "max_grad_norm": 0.8, "loss": 0.6255064606666565, "grad_norm": 1.047341227531433, "learning_rate": 9.471220014181246e-05}
+{"ts": "2025-12-24T05:21:00", "event": "train_log", "step": 3006, "epoch": 1.2968716289104638, "progress_pct": 21.61, "epoch_pct": 21.61, "eta": "54:15:26", "max_grad_norm": 0.8, "loss": 0.7038446068763733, "grad_norm": 1.2761468887329102, "learning_rate": 9.470120519774237e-05}
+{"ts": "2025-12-24T05:21:19", "event": "train_log", "step": 3008, "epoch": 1.2977346278317152, "progress_pct": 21.63, "epoch_pct": 21.63, "eta": "54:13:51", "max_grad_norm": 0.8, "loss": 0.7063520550727844, "grad_norm": 1.2568607330322266, "learning_rate": 9.469019947417815e-05}
+{"ts": "2025-12-24T05:21:38", "event": "train_log", "step": 3010, "epoch": 1.2985976267529666, "progress_pct": 21.64, "epoch_pct": 21.64, "eta": "54:12:14", "max_grad_norm": 0.8, "loss": 0.6931142807006836, "grad_norm": 1.2337472438812256, "learning_rate": 9.467918297377378e-05}
+{"ts": "2025-12-24T05:21:57", "event": "train_log", "step": 3012, "epoch": 1.2994606256742178, "progress_pct": 21.66, "epoch_pct": 21.66, "eta": "54:10:37", "max_grad_norm": 0.8, "loss": 0.6660345196723938, "grad_norm": 1.1557739973068237, "learning_rate": 9.466815569918588e-05}
+{"ts": "2025-12-24T05:22:16", "event": "train_log", "step": 3014, "epoch": 1.3003236245954692, "progress_pct": 21.67, "epoch_pct": 21.67, "eta": "54:09:02", "max_grad_norm": 0.8, "loss": 0.6522690653800964, "grad_norm": 1.2341985702514648, "learning_rate": 9.465711765307357e-05}
+{"ts": "2025-12-24T05:22:36", "event": "train_log", "step": 3016, "epoch": 1.3011866235167207, "progress_pct": 21.69, "epoch_pct": 21.69, "eta": "54:07:27", "max_grad_norm": 0.8, "loss": 0.6304616332054138, "grad_norm": 1.0495452880859375, "learning_rate": 9.464606883809867e-05}
+{"ts": "2025-12-24T05:22:54", "event": "train_log", "step": 3018, "epoch": 1.302049622437972, "progress_pct": 21.7, "epoch_pct": 21.7, "eta": "54:05:49", "max_grad_norm": 0.8, "loss": 0.7271291613578796, "grad_norm": 1.1920377016067505, "learning_rate": 9.463500925692554e-05}
+{"ts": "2025-12-24T05:23:15", "event": "train_log", "step": 3020, "epoch": 1.3029126213592233, "progress_pct": 21.71, "epoch_pct": 21.72, "eta": "54:04:19", "max_grad_norm": 0.8, "loss": 0.5725325345993042, "grad_norm": 1.0697529315948486, "learning_rate": 9.462393891222112e-05}
+{"ts": "2025-12-24T05:23:33", "event": "train_log", "step": 3022, "epoch": 1.3037756202804747, "progress_pct": 21.73, "epoch_pct": 21.73, "eta": "54:02:40", "max_grad_norm": 0.8, "loss": 0.5725870728492737, "grad_norm": 1.0559576749801636, "learning_rate": 9.461285780665502e-05}
+{"ts": "2025-12-24T05:23:52", "event": "train_log", "step": 3024, "epoch": 1.304638619201726, "progress_pct": 21.74, "epoch_pct": 21.74, "eta": "54:01:04", "max_grad_norm": 0.8, "loss": 0.6498687863349915, "grad_norm": 1.1314871311187744, "learning_rate": 9.460176594289937e-05}
+{"ts": "2025-12-24T05:24:11", "event": "train_log", "step": 3026, "epoch": 1.3055016181229773, "progress_pct": 21.76, "epoch_pct": 21.76, "eta": "53:59:29", "max_grad_norm": 0.8, "loss": 0.5473160147666931, "grad_norm": 1.0620028972625732, "learning_rate": 9.459066332362892e-05}
+{"ts": "2025-12-24T05:24:29", "event": "train_log", "step": 3028, "epoch": 1.3063646170442287, "progress_pct": 21.77, "epoch_pct": 21.77, "eta": "53:57:49", "max_grad_norm": 0.8, "loss": 0.6976684927940369, "grad_norm": 1.245997667312622, "learning_rate": 9.457954995152103e-05}
+{"ts": "2025-12-24T05:24:48", "event": "train_log", "step": 3030, "epoch": 1.30722761596548, "progress_pct": 21.79, "epoch_pct": 21.79, "eta": "53:56:12", "max_grad_norm": 0.8, "loss": 0.6469701528549194, "grad_norm": 1.0973089933395386, "learning_rate": 9.456842582925565e-05}
+{"ts": "2025-12-24T05:25:07", "event": "train_log", "step": 3032, "epoch": 1.3080906148867313, "progress_pct": 21.8, "epoch_pct": 21.8, "eta": "53:54:38", "max_grad_norm": 0.8, "loss": 0.6712765097618103, "grad_norm": 1.1558068990707397, "learning_rate": 9.455729095951527e-05}
+{"ts": "2025-12-24T05:25:27", "event": "train_log", "step": 3034, "epoch": 1.3089536138079827, "progress_pct": 21.81, "epoch_pct": 21.82, "eta": "53:53:05", "max_grad_norm": 0.8, "loss": 0.6504483222961426, "grad_norm": 1.0367828607559204, "learning_rate": 9.454614534498506e-05}
+{"ts": "2025-12-24T05:25:46", "event": "train_log", "step": 3036, "epoch": 1.309816612729234, "progress_pct": 21.83, "epoch_pct": 21.83, "eta": "53:51:29", "max_grad_norm": 0.8, "loss": 0.7000672221183777, "grad_norm": 1.3485349416732788, "learning_rate": 9.453498898835272e-05}
+{"ts": "2025-12-24T05:26:05", "event": "train_log", "step": 3038, "epoch": 1.3106796116504853, "progress_pct": 21.84, "epoch_pct": 21.84, "eta": "53:49:53", "max_grad_norm": 0.8, "loss": 0.6946983337402344, "grad_norm": 1.2435184717178345, "learning_rate": 9.452382189230853e-05}
+{"ts": "2025-12-24T05:26:26", "event": "train_log", "step": 3040, "epoch": 1.3115426105717367, "progress_pct": 21.86, "epoch_pct": 21.86, "eta": "53:48:27", "max_grad_norm": 0.8, "loss": 0.5928027033805847, "grad_norm": 1.0728851556777954, "learning_rate": 9.451264405954541e-05}
+{"ts": "2025-12-24T05:26:46", "event": "train_log", "step": 3042, "epoch": 1.3124056094929881, "progress_pct": 21.87, "epoch_pct": 21.87, "eta": "53:46:55", "max_grad_norm": 0.8, "loss": 0.6589536070823669, "grad_norm": 1.2320033311843872, "learning_rate": 9.450145549275884e-05}
+{"ts": "2025-12-24T05:27:06", "event": "train_log", "step": 3044, "epoch": 1.3132686084142395, "progress_pct": 21.89, "epoch_pct": 21.89, "eta": "53:45:22", "max_grad_norm": 0.8, "loss": 0.6516439914703369, "grad_norm": 1.143538236618042, "learning_rate": 9.449025619464689e-05}
+{"ts": "2025-12-24T05:27:23", "event": "train_log", "step": 3046, "epoch": 1.314131607335491, "progress_pct": 21.9, "epoch_pct": 21.9, "eta": "53:43:42", "max_grad_norm": 0.8, "loss": 0.6889938116073608, "grad_norm": 1.3419140577316284, "learning_rate": 9.447904616791023e-05}
+{"ts": "2025-12-24T05:27:41", "event": "train_log", "step": 3048, "epoch": 1.3149946062567421, "progress_pct": 21.92, "epoch_pct": 21.92, "eta": "53:42:02", "max_grad_norm": 0.8, "loss": 0.7203409075737, "grad_norm": 1.4128848314285278, "learning_rate": 9.44678254152521e-05}
+{"ts": "2025-12-24T05:28:01", "event": "train_log", "step": 3050, "epoch": 1.3158576051779936, "progress_pct": 21.93, "epoch_pct": 21.93, "eta": "53:40:31", "max_grad_norm": 0.8, "loss": 0.656524658203125, "grad_norm": 1.1409581899642944, "learning_rate": 9.445659393937833e-05}
+{"ts": "2025-12-24T05:28:19", "event": "train_log", "step": 3052, "epoch": 1.316720604099245, "progress_pct": 21.94, "epoch_pct": 21.95, "eta": "53:38:55", "max_grad_norm": 0.8, "loss": 0.6888052821159363, "grad_norm": 1.0904145240783691, "learning_rate": 9.444535174299734e-05}
+{"ts": "2025-12-24T05:28:39", "event": "train_log", "step": 3054, "epoch": 1.3175836030204962, "progress_pct": 21.96, "epoch_pct": 21.96, "eta": "53:37:22", "max_grad_norm": 0.8, "loss": 0.6149572730064392, "grad_norm": 1.1533669233322144, "learning_rate": 9.443409882882014e-05}
+{"ts": "2025-12-24T05:29:00", "event": "train_log", "step": 3056, "epoch": 1.3184466019417476, "progress_pct": 21.97, "epoch_pct": 21.97, "eta": "53:35:54", "max_grad_norm": 0.8, "loss": 0.6445736885070801, "grad_norm": 1.0775257349014282, "learning_rate": 9.442283519956031e-05}
+{"ts": "2025-12-24T05:29:19", "event": "train_log", "step": 3058, "epoch": 1.319309600862999, "progress_pct": 21.99, "epoch_pct": 21.99, "eta": "53:34:22", "max_grad_norm": 0.8, "loss": 0.6723806858062744, "grad_norm": 1.1126797199249268, "learning_rate": 9.441156085793403e-05}
+{"ts": "2025-12-24T05:29:39", "event": "train_log", "step": 3060, "epoch": 1.3201725997842502, "progress_pct": 22.0, "epoch_pct": 22.0, "eta": "53:32:50", "max_grad_norm": 0.8, "loss": 0.5901219248771667, "grad_norm": 1.1584032773971558, "learning_rate": 9.440027580666008e-05}
+{"ts": "2025-12-24T05:29:58", "event": "train_log", "step": 3062, "epoch": 1.3210355987055016, "progress_pct": 22.02, "epoch_pct": 22.02, "eta": "53:31:16", "max_grad_norm": 0.8, "loss": 0.6816843748092651, "grad_norm": 1.0951672792434692, "learning_rate": 9.438898004845975e-05}
+{"ts": "2025-12-24T05:30:16", "event": "train_log", "step": 3064, "epoch": 1.321898597626753, "progress_pct": 22.03, "epoch_pct": 22.03, "eta": "53:29:39", "max_grad_norm": 0.8, "loss": 0.6883957386016846, "grad_norm": 1.2165061235427856, "learning_rate": 9.4377673586057e-05}
+{"ts": "2025-12-24T05:30:35", "event": "train_log", "step": 3066, "epoch": 1.3227615965480042, "progress_pct": 22.04, "epoch_pct": 22.05, "eta": "53:28:06", "max_grad_norm": 0.8, "loss": 0.6828808188438416, "grad_norm": 1.1537514925003052, "learning_rate": 9.436635642217832e-05}
+{"ts": "2025-12-24T05:30:56", "event": "train_log", "step": 3068, "epoch": 1.3236245954692556, "progress_pct": 22.06, "epoch_pct": 22.06, "eta": "53:26:36", "max_grad_norm": 0.8, "loss": 0.6086717844009399, "grad_norm": 1.2186168432235718, "learning_rate": 9.435502855955281e-05}
+{"ts": "2025-12-24T05:31:16", "event": "train_log", "step": 3070, "epoch": 1.324487594390507, "progress_pct": 22.07, "epoch_pct": 22.07, "eta": "53:25:08", "max_grad_norm": 0.8, "loss": 0.6000884771347046, "grad_norm": 1.0345975160598755, "learning_rate": 9.43436900009121e-05}
+{"ts": "2025-12-24T05:31:36", "event": "train_log", "step": 3072, "epoch": 1.3253505933117584, "progress_pct": 22.09, "epoch_pct": 22.09, "eta": "53:23:38", "max_grad_norm": 0.8, "loss": 0.6812195181846619, "grad_norm": 1.1357595920562744, "learning_rate": 9.433234074899044e-05}
+{"ts": "2025-12-24T05:31:55", "event": "train_log", "step": 3074, "epoch": 1.3262135922330098, "progress_pct": 22.1, "epoch_pct": 22.1, "eta": "53:22:04", "max_grad_norm": 0.8, "loss": 0.6666595935821533, "grad_norm": 1.1914640665054321, "learning_rate": 9.432098080652467e-05}
+{"ts": "2025-12-24T05:32:18", "event": "train_log", "step": 3076, "epoch": 1.327076591154261, "progress_pct": 22.12, "epoch_pct": 22.12, "eta": "53:20:44", "max_grad_norm": 0.8, "loss": 0.6455721259117126, "grad_norm": 1.0881390571594238, "learning_rate": 9.430961017625418e-05}
+{"ts": "2025-12-24T05:32:37", "event": "train_log", "step": 3078, "epoch": 1.3279395900755124, "progress_pct": 22.13, "epoch_pct": 22.13, "eta": "53:19:12", "max_grad_norm": 0.8, "loss": 0.5861079692840576, "grad_norm": 1.1859697103500366, "learning_rate": 9.429822886092093e-05}
+{"ts": "2025-12-24T05:32:57", "event": "train_log", "step": 3080, "epoch": 1.3288025889967638, "progress_pct": 22.15, "epoch_pct": 22.15, "eta": "53:17:42", "max_grad_norm": 0.8, "loss": 0.6173920035362244, "grad_norm": 1.083703637123108, "learning_rate": 9.42868368632695e-05}
+{"ts": "2025-12-24T05:33:16", "event": "train_log", "step": 3082, "epoch": 1.329665587918015, "progress_pct": 22.16, "epoch_pct": 22.16, "eta": "53:16:09", "max_grad_norm": 0.8, "loss": 0.817629873752594, "grad_norm": 1.2820687294006348, "learning_rate": 9.427543418604701e-05}
+{"ts": "2025-12-24T05:33:35", "event": "train_log", "step": 3084, "epoch": 1.3305285868392664, "progress_pct": 22.17, "epoch_pct": 22.18, "eta": "53:14:34", "max_grad_norm": 0.8, "loss": 0.7454650402069092, "grad_norm": 1.1267426013946533, "learning_rate": 9.426402083200315e-05}
+{"ts": "2025-12-24T05:33:56", "event": "train_log", "step": 3086, "epoch": 1.3313915857605179, "progress_pct": 22.19, "epoch_pct": 22.19, "eta": "53:13:09", "max_grad_norm": 0.8, "loss": 0.625328779220581, "grad_norm": 1.0095305442810059, "learning_rate": 9.425259680389022e-05}
+{"ts": "2025-12-24T05:34:16", "event": "train_log", "step": 3088, "epoch": 1.332254584681769, "progress_pct": 22.2, "epoch_pct": 22.2, "eta": "53:11:41", "max_grad_norm": 0.8, "loss": 0.5977230072021484, "grad_norm": 1.183151364326477, "learning_rate": 9.424116210446305e-05}
+{"ts": "2025-12-24T05:34:36", "event": "train_log", "step": 3090, "epoch": 1.3331175836030205, "progress_pct": 22.22, "epoch_pct": 22.22, "eta": "53:10:09", "max_grad_norm": 0.8, "loss": 0.6251013278961182, "grad_norm": 1.1312501430511475, "learning_rate": 9.42297167364791e-05}
+{"ts": "2025-12-24T05:34:53", "event": "train_log", "step": 3092, "epoch": 1.3339805825242719, "progress_pct": 22.23, "epoch_pct": 22.23, "eta": "53:08:31", "max_grad_norm": 0.8, "loss": 0.7469011545181274, "grad_norm": 1.3226121664047241, "learning_rate": 9.421826070269832e-05}
+{"ts": "2025-12-24T05:35:12", "event": "train_log", "step": 3094, "epoch": 1.334843581445523, "progress_pct": 22.25, "epoch_pct": 22.25, "eta": "53:06:59", "max_grad_norm": 0.8, "loss": 0.7012571096420288, "grad_norm": 1.1389979124069214, "learning_rate": 9.420679400588334e-05}
+{"ts": "2025-12-24T05:35:31", "event": "train_log", "step": 3096, "epoch": 1.3357065803667745, "progress_pct": 22.26, "epoch_pct": 22.26, "eta": "53:05:27", "max_grad_norm": 0.8, "loss": 0.7112689018249512, "grad_norm": 1.2065027952194214, "learning_rate": 9.419531664879925e-05}
+{"ts": "2025-12-24T05:35:52", "event": "train_log", "step": 3098, "epoch": 1.3365695792880259, "progress_pct": 22.27, "epoch_pct": 22.28, "eta": "53:04:01", "max_grad_norm": 0.8, "loss": 0.6219019293785095, "grad_norm": 1.2532200813293457, "learning_rate": 9.418382863421379e-05}
+{"ts": "2025-12-24T05:36:11", "event": "train_log", "step": 3100, "epoch": 1.3374325782092773, "progress_pct": 22.29, "epoch_pct": 22.29, "eta": "53:02:29", "max_grad_norm": 0.8, "loss": 0.6905092597007751, "grad_norm": 1.1859256029129028, "learning_rate": 9.417232996489724e-05}
+{"ts": "2025-12-24T05:49:53", "event": "train_log", "step": 3100, "epoch": 1.3374325782092773, "progress_pct": 22.29, "epoch_pct": 22.29, "eta": "53:50:12", "max_grad_norm": 0.8, "eval_loss": 0.7127581834793091, "eval_runtime": 821.1346, "eval_samples_per_second": 2.509, "eval_steps_per_second": 2.509}
+{"ts": "2025-12-24T05:50:12", "event": "train_log", "step": 3102, "epoch": 1.3382955771305287, "progress_pct": 22.3, "epoch_pct": 22.3, "eta": "53:48:41", "max_grad_norm": 0.8, "loss": 0.6473057866096497, "grad_norm": 1.172967791557312, "learning_rate": 9.416082064362244e-05}
+{"ts": "2025-12-24T05:50:31", "event": "train_log", "step": 3104, "epoch": 1.33915857605178, "progress_pct": 22.32, "epoch_pct": 22.32, "eta": "53:47:04", "max_grad_norm": 0.8, "loss": 0.7377036213874817, "grad_norm": 1.2966243028640747, "learning_rate": 9.414930067316482e-05}
+{"ts": "2025-12-24T05:50:50", "event": "train_log", "step": 3106, "epoch": 1.3400215749730313, "progress_pct": 22.33, "epoch_pct": 22.33, "eta": "53:45:30", "max_grad_norm": 0.8, "loss": 0.6842042207717896, "grad_norm": 1.0844253301620483, "learning_rate": 9.413777005630237e-05}
+{"ts": "2025-12-24T05:51:10", "event": "train_log", "step": 3108, "epoch": 1.3408845738942827, "progress_pct": 22.35, "epoch_pct": 22.35, "eta": "53:43:58", "max_grad_norm": 0.8, "loss": 0.662327229976654, "grad_norm": 0.930098831653595, "learning_rate": 9.412622879581565e-05}
+{"ts": "2025-12-24T05:51:29", "event": "train_log", "step": 3110, "epoch": 1.341747572815534, "progress_pct": 22.36, "epoch_pct": 22.36, "eta": "53:42:26", "max_grad_norm": 0.8, "loss": 0.6979284882545471, "grad_norm": 1.1382031440734863, "learning_rate": 9.411467689448777e-05}
+{"ts": "2025-12-24T05:51:48", "event": "train_log", "step": 3112, "epoch": 1.3426105717367853, "progress_pct": 22.38, "epoch_pct": 22.38, "eta": "53:40:50", "max_grad_norm": 0.8, "loss": 0.7032427191734314, "grad_norm": 1.1555129289627075, "learning_rate": 9.410311435510444e-05}
+{"ts": "2025-12-24T05:52:07", "event": "train_log", "step": 3114, "epoch": 1.3434735706580367, "progress_pct": 22.39, "epoch_pct": 22.39, "eta": "53:39:17", "max_grad_norm": 0.8, "loss": 0.6282536387443542, "grad_norm": 1.147678017616272, "learning_rate": 9.409154118045389e-05}
+{"ts": "2025-12-24T05:52:26", "event": "train_log", "step": 3116, "epoch": 1.344336569579288, "progress_pct": 22.4, "epoch_pct": 22.41, "eta": "53:37:44", "max_grad_norm": 0.8, "loss": 0.6572934985160828, "grad_norm": 1.0925829410552979, "learning_rate": 9.407995737332696e-05}
+{"ts": "2025-12-24T05:52:44", "event": "train_log", "step": 3118, "epoch": 1.3451995685005393, "progress_pct": 22.42, "epoch_pct": 22.42, "eta": "53:36:07", "max_grad_norm": 0.8, "loss": 0.7233583927154541, "grad_norm": 1.3632638454437256, "learning_rate": 9.4068362936517e-05}
+{"ts": "2025-12-24T05:53:03", "event": "train_log", "step": 3120, "epoch": 1.3460625674217908, "progress_pct": 22.43, "epoch_pct": 22.43, "eta": "53:34:33", "max_grad_norm": 0.8, "loss": 0.6465377807617188, "grad_norm": 1.3994121551513672, "learning_rate": 9.405675787282e-05}
+{"ts": "2025-12-24T05:53:23", "event": "train_log", "step": 3122, "epoch": 1.346925566343042, "progress_pct": 22.45, "epoch_pct": 22.45, "eta": "53:33:01", "max_grad_norm": 0.8, "loss": 0.6348391771316528, "grad_norm": 1.1862388849258423, "learning_rate": 9.404514218503444e-05}
+{"ts": "2025-12-24T05:53:43", "event": "train_log", "step": 3124, "epoch": 1.3477885652642934, "progress_pct": 22.46, "epoch_pct": 22.46, "eta": "53:31:31", "max_grad_norm": 0.8, "loss": 0.6559093594551086, "grad_norm": 1.1724227666854858, "learning_rate": 9.403351587596141e-05}
+{"ts": "2025-12-24T05:54:01", "event": "train_log", "step": 3126, "epoch": 1.3486515641855448, "progress_pct": 22.48, "epoch_pct": 22.48, "eta": "53:29:56", "max_grad_norm": 0.8, "loss": 0.6693678498268127, "grad_norm": 1.175567865371704, "learning_rate": 9.402187894840452e-05}
+{"ts": "2025-12-24T05:54:20", "event": "train_log", "step": 3128, "epoch": 1.3495145631067962, "progress_pct": 22.49, "epoch_pct": 22.49, "eta": "53:28:22", "max_grad_norm": 0.8, "loss": 0.6766137480735779, "grad_norm": 1.1273375749588013, "learning_rate": 9.401023140516999e-05}
+{"ts": "2025-12-24T05:54:41", "event": "train_log", "step": 3130, "epoch": 1.3503775620280476, "progress_pct": 22.51, "epoch_pct": 22.51, "eta": "53:26:56", "max_grad_norm": 0.8, "loss": 0.7133797407150269, "grad_norm": 1.2515393495559692, "learning_rate": 9.399857324906652e-05}
+{"ts": "2025-12-24T05:55:00", "event": "train_log", "step": 3132, "epoch": 1.3512405609492988, "progress_pct": 22.52, "epoch_pct": 22.52, "eta": "53:25:20", "max_grad_norm": 0.8, "loss": 0.6572166681289673, "grad_norm": 1.1575038433074951, "learning_rate": 9.398690448290547e-05}
+{"ts": "2025-12-24T05:55:21", "event": "train_log", "step": 3134, "epoch": 1.3521035598705502, "progress_pct": 22.53, "epoch_pct": 22.54, "eta": "53:23:54", "max_grad_norm": 0.8, "loss": 0.6389561891555786, "grad_norm": 1.2350742816925049, "learning_rate": 9.397522510950072e-05}
+{"ts": "2025-12-24T05:55:41", "event": "train_log", "step": 3136, "epoch": 1.3529665587918016, "progress_pct": 22.55, "epoch_pct": 22.55, "eta": "53:22:26", "max_grad_norm": 0.8, "loss": 0.5810164213180542, "grad_norm": 0.9648056626319885, "learning_rate": 9.396353513166867e-05}
+{"ts": "2025-12-24T05:55:59", "event": "train_log", "step": 3138, "epoch": 1.3538295577130528, "progress_pct": 22.56, "epoch_pct": 22.56, "eta": "53:20:51", "max_grad_norm": 0.8, "loss": 0.7157397866249084, "grad_norm": 1.147218108177185, "learning_rate": 9.39518345522283e-05}
+{"ts": "2025-12-24T05:56:18", "event": "train_log", "step": 3140, "epoch": 1.3546925566343042, "progress_pct": 22.58, "epoch_pct": 22.58, "eta": "53:19:16", "max_grad_norm": 0.8, "loss": 0.6749072670936584, "grad_norm": 1.1948896646499634, "learning_rate": 9.394012337400117e-05}
+{"ts": "2025-12-24T05:56:38", "event": "train_log", "step": 3142, "epoch": 1.3555555555555556, "progress_pct": 22.59, "epoch_pct": 22.59, "eta": "53:17:47", "max_grad_norm": 0.8, "loss": 0.63704514503479, "grad_norm": 1.0945216417312622, "learning_rate": 9.392840159981136e-05}
+{"ts": "2025-12-24T05:56:59", "event": "train_log", "step": 3144, "epoch": 1.3564185544768068, "progress_pct": 22.61, "epoch_pct": 22.61, "eta": "53:16:23", "max_grad_norm": 0.8, "loss": 0.6804014444351196, "grad_norm": 1.0654480457305908, "learning_rate": 9.391666923248555e-05}
+{"ts": "2025-12-24T05:57:17", "event": "train_log", "step": 3146, "epoch": 1.3572815533980582, "progress_pct": 22.62, "epoch_pct": 22.62, "eta": "53:14:47", "max_grad_norm": 0.8, "loss": 0.7055573463439941, "grad_norm": 1.3977962732315063, "learning_rate": 9.390492627485292e-05}
+{"ts": "2025-12-24T05:57:38", "event": "train_log", "step": 3148, "epoch": 1.3581445523193096, "progress_pct": 22.63, "epoch_pct": 22.64, "eta": "53:13:20", "max_grad_norm": 0.8, "loss": 0.7383730411529541, "grad_norm": 1.1607128381729126, "learning_rate": 9.389317272974526e-05}
+{"ts": "2025-12-24T05:57:55", "event": "train_log", "step": 3150, "epoch": 1.3590075512405608, "progress_pct": 22.65, "epoch_pct": 22.65, "eta": "53:11:42", "max_grad_norm": 0.8, "loss": 0.7339736223220825, "grad_norm": 1.192213773727417, "learning_rate": 9.388140859999686e-05}
+{"ts": "2025-12-24T05:58:15", "event": "train_log", "step": 3152, "epoch": 1.3598705501618122, "progress_pct": 22.66, "epoch_pct": 22.66, "eta": "53:10:12", "max_grad_norm": 0.8, "loss": 0.6564369797706604, "grad_norm": 1.1597269773483276, "learning_rate": 9.386963388844459e-05}
+{"ts": "2025-12-24T05:58:34", "event": "train_log", "step": 3154, "epoch": 1.3607335490830637, "progress_pct": 22.68, "epoch_pct": 22.68, "eta": "53:08:39", "max_grad_norm": 0.8, "loss": 0.7393131256103516, "grad_norm": 1.2374637126922607, "learning_rate": 9.385784859792788e-05}
+{"ts": "2025-12-24T05:58:54", "event": "train_log", "step": 3156, "epoch": 1.361596548004315, "progress_pct": 22.69, "epoch_pct": 22.69, "eta": "53:07:11", "max_grad_norm": 0.8, "loss": 0.6794763207435608, "grad_norm": 1.1975921392440796, "learning_rate": 9.384605273128868e-05}
+{"ts": "2025-12-24T05:59:13", "event": "train_log", "step": 3158, "epoch": 1.3624595469255665, "progress_pct": 22.71, "epoch_pct": 22.71, "eta": "53:05:40", "max_grad_norm": 0.8, "loss": 0.5906237363815308, "grad_norm": 1.158282995223999, "learning_rate": 9.383424629137154e-05}
+{"ts": "2025-12-24T05:59:34", "event": "train_log", "step": 3160, "epoch": 1.3633225458468177, "progress_pct": 22.72, "epoch_pct": 22.72, "eta": "53:04:14", "max_grad_norm": 0.8, "loss": 0.6813581585884094, "grad_norm": 1.0605288743972778, "learning_rate": 9.38224292810235e-05}
+{"ts": "2025-12-24T05:59:55", "event": "train_log", "step": 3162, "epoch": 1.364185544768069, "progress_pct": 22.74, "epoch_pct": 22.74, "eta": "53:02:49", "max_grad_norm": 0.8, "loss": 0.639259934425354, "grad_norm": 0.9797501564025879, "learning_rate": 9.38106017030942e-05}
+{"ts": "2025-12-24T06:00:15", "event": "train_log", "step": 3164, "epoch": 1.3650485436893205, "progress_pct": 22.75, "epoch_pct": 22.75, "eta": "53:01:22", "max_grad_norm": 0.8, "loss": 0.667378306388855, "grad_norm": 1.1023533344268799, "learning_rate": 9.379876356043578e-05}
+{"ts": "2025-12-24T06:00:34", "event": "train_log", "step": 3166, "epoch": 1.3659115426105717, "progress_pct": 22.76, "epoch_pct": 22.77, "eta": "52:59:50", "max_grad_norm": 0.8, "loss": 0.671601414680481, "grad_norm": 1.1105194091796875, "learning_rate": 9.378691485590299e-05}
+{"ts": "2025-12-24T06:00:55", "event": "train_log", "step": 3168, "epoch": 1.366774541531823, "progress_pct": 22.78, "epoch_pct": 22.78, "eta": "52:58:26", "max_grad_norm": 0.8, "loss": 0.623394787311554, "grad_norm": 0.9441736936569214, "learning_rate": 9.377505559235308e-05}
+{"ts": "2025-12-24T06:01:16", "event": "train_log", "step": 3170, "epoch": 1.3676375404530745, "progress_pct": 22.79, "epoch_pct": 22.79, "eta": "52:57:00", "max_grad_norm": 0.8, "loss": 0.6558915972709656, "grad_norm": 1.308560848236084, "learning_rate": 9.376318577264585e-05}
+{"ts": "2025-12-24T06:01:35", "event": "train_log", "step": 3172, "epoch": 1.3685005393743257, "progress_pct": 22.81, "epoch_pct": 22.81, "eta": "52:55:29", "max_grad_norm": 0.8, "loss": 0.6431660652160645, "grad_norm": 1.065693974494934, "learning_rate": 9.375130539964363e-05}
+{"ts": "2025-12-24T06:01:55", "event": "train_log", "step": 3174, "epoch": 1.369363538295577, "progress_pct": 22.82, "epoch_pct": 22.82, "eta": "52:54:00", "max_grad_norm": 0.8, "loss": 0.6541599631309509, "grad_norm": 1.027817726135254, "learning_rate": 9.373941447621137e-05}
+{"ts": "2025-12-24T06:02:14", "event": "train_log", "step": 3176, "epoch": 1.3702265372168285, "progress_pct": 22.84, "epoch_pct": 22.84, "eta": "52:52:28", "max_grad_norm": 0.8, "loss": 0.6710350513458252, "grad_norm": 1.1651864051818848, "learning_rate": 9.372751300521649e-05}
+{"ts": "2025-12-24T06:02:32", "event": "train_log", "step": 3178, "epoch": 1.3710895361380797, "progress_pct": 22.85, "epoch_pct": 22.85, "eta": "52:50:56", "max_grad_norm": 0.8, "loss": 0.6525241732597351, "grad_norm": 1.1066170930862427, "learning_rate": 9.371560098952896e-05}
+{"ts": "2025-12-24T06:02:51", "event": "train_log", "step": 3180, "epoch": 1.3719525350593311, "progress_pct": 22.86, "epoch_pct": 22.87, "eta": "52:49:25", "max_grad_norm": 0.8, "loss": 0.6362038850784302, "grad_norm": 1.0942546129226685, "learning_rate": 9.370367843202133e-05}
+{"ts": "2025-12-24T06:03:11", "event": "train_log", "step": 3182, "epoch": 1.3728155339805825, "progress_pct": 22.88, "epoch_pct": 22.88, "eta": "52:47:58", "max_grad_norm": 0.8, "loss": 0.6017009615898132, "grad_norm": 1.175432562828064, "learning_rate": 9.369174533556868e-05}
+{"ts": "2025-12-24T06:03:30", "event": "train_log", "step": 3184, "epoch": 1.373678532901834, "progress_pct": 22.89, "epoch_pct": 22.89, "eta": "52:46:27", "max_grad_norm": 0.8, "loss": 0.6853877305984497, "grad_norm": 1.3605095148086548, "learning_rate": 9.367980170304857e-05}
+{"ts": "2025-12-24T06:03:50", "event": "train_log", "step": 3186, "epoch": 1.3745415318230851, "progress_pct": 22.91, "epoch_pct": 22.91, "eta": "52:44:58", "max_grad_norm": 0.8, "loss": 0.6434065699577332, "grad_norm": 1.224565029144287, "learning_rate": 9.366784753734121e-05}
+{"ts": "2025-12-24T06:04:08", "event": "train_log", "step": 3188, "epoch": 1.3754045307443366, "progress_pct": 22.92, "epoch_pct": 22.92, "eta": "52:43:26", "max_grad_norm": 0.8, "loss": 0.596630334854126, "grad_norm": 1.1151230335235596, "learning_rate": 9.365588284132927e-05}
+{"ts": "2025-12-24T06:04:27", "event": "train_log", "step": 3190, "epoch": 1.376267529665588, "progress_pct": 22.94, "epoch_pct": 22.94, "eta": "52:41:54", "max_grad_norm": 0.8, "loss": 0.6716568470001221, "grad_norm": 1.2674700021743774, "learning_rate": 9.364390761789799e-05}
+{"ts": "2025-12-24T06:04:47", "event": "train_log", "step": 3192, "epoch": 1.3771305285868394, "progress_pct": 22.95, "epoch_pct": 22.95, "eta": "52:40:26", "max_grad_norm": 0.8, "loss": 0.6230685710906982, "grad_norm": 1.3221330642700195, "learning_rate": 9.363192186993512e-05}
+{"ts": "2025-12-24T06:05:07", "event": "train_log", "step": 3194, "epoch": 1.3779935275080906, "progress_pct": 22.97, "epoch_pct": 22.97, "eta": "52:38:59", "max_grad_norm": 0.8, "loss": 0.6495190858840942, "grad_norm": 1.2978439331054688, "learning_rate": 9.3619925600331e-05}
+{"ts": "2025-12-24T06:05:26", "event": "train_log", "step": 3196, "epoch": 1.378856526429342, "progress_pct": 22.98, "epoch_pct": 22.98, "eta": "52:37:29", "max_grad_norm": 0.8, "loss": 0.6534912586212158, "grad_norm": 1.1542012691497803, "learning_rate": 9.360791881197846e-05}
+{"ts": "2025-12-24T06:05:46", "event": "train_log", "step": 3198, "epoch": 1.3797195253505934, "progress_pct": 22.99, "epoch_pct": 23.0, "eta": "52:36:03", "max_grad_norm": 0.8, "loss": 0.5706549286842346, "grad_norm": 1.1294280290603638, "learning_rate": 9.359590150777288e-05}
+{"ts": "2025-12-24T06:06:06", "event": "train_log", "step": 3200, "epoch": 1.3805825242718446, "progress_pct": 23.01, "epoch_pct": 23.01, "eta": "52:34:34", "max_grad_norm": 0.8, "loss": 0.6193811893463135, "grad_norm": 1.0622307062149048, "learning_rate": 9.358387369061218e-05}
+{"ts": "2025-12-24T06:19:57", "event": "train_log", "step": 3200, "epoch": 1.3805825242718446, "progress_pct": 23.01, "epoch_pct": 23.01, "eta": "53:20:56", "max_grad_norm": 0.8, "eval_loss": 0.7103806734085083, "eval_runtime": 831.4047, "eval_samples_per_second": 2.478, "eval_steps_per_second": 2.478}
+{"ts": "2025-12-24T06:20:16", "event": "train_log", "step": 3202, "epoch": 1.381445523193096, "progress_pct": 23.02, "epoch_pct": 23.02, "eta": "53:19:26", "max_grad_norm": 0.8, "loss": 0.6021195650100708, "grad_norm": 1.1722023487091064, "learning_rate": 9.357183536339683e-05}
+{"ts": "2025-12-24T06:20:35", "event": "train_log", "step": 3204, "epoch": 1.3823085221143474, "progress_pct": 23.04, "epoch_pct": 23.04, "eta": "53:17:53", "max_grad_norm": 0.8, "loss": 0.7195428013801575, "grad_norm": 1.2733373641967773, "learning_rate": 9.355978652902978e-05}
+{"ts": "2025-12-24T06:20:56", "event": "train_log", "step": 3206, "epoch": 1.3831715210355986, "progress_pct": 23.05, "epoch_pct": 23.05, "eta": "53:16:27", "max_grad_norm": 0.8, "loss": 0.6314851641654968, "grad_norm": 1.019609808921814, "learning_rate": 9.354772719041658e-05}
+{"ts": "2025-12-24T06:21:13", "event": "train_log", "step": 3208, "epoch": 1.38403451995685, "progress_pct": 23.07, "epoch_pct": 23.07, "eta": "53:14:48", "max_grad_norm": 0.8, "loss": 0.722943902015686, "grad_norm": 1.36198890209198, "learning_rate": 9.35356573504653e-05}
+{"ts": "2025-12-24T06:21:33", "event": "train_log", "step": 3210, "epoch": 1.3848975188781014, "progress_pct": 23.08, "epoch_pct": 23.08, "eta": "53:13:20", "max_grad_norm": 0.8, "loss": 0.7058439254760742, "grad_norm": 1.1291067600250244, "learning_rate": 9.35235770120865e-05}
+{"ts": "2025-12-24T06:21:54", "event": "train_log", "step": 3212, "epoch": 1.3857605177993528, "progress_pct": 23.09, "epoch_pct": 23.1, "eta": "53:11:53", "max_grad_norm": 0.8, "loss": 0.6186655759811401, "grad_norm": 1.2160366773605347, "learning_rate": 9.351148617819332e-05}
+{"ts": "2025-12-24T06:22:13", "event": "train_log", "step": 3214, "epoch": 1.386623516720604, "progress_pct": 23.11, "epoch_pct": 23.11, "eta": "53:10:24", "max_grad_norm": 0.8, "loss": 0.5912084579467773, "grad_norm": 1.1663079261779785, "learning_rate": 9.349938485170139e-05}
+{"ts": "2025-12-24T06:22:32", "event": "train_log", "step": 3216, "epoch": 1.3874865156418554, "progress_pct": 23.12, "epoch_pct": 23.12, "eta": "53:08:50", "max_grad_norm": 0.8, "loss": 0.6494051218032837, "grad_norm": 1.2145130634307861, "learning_rate": 9.34872730355289e-05}
+{"ts": "2025-12-24T06:22:53", "event": "train_log", "step": 3218, "epoch": 1.3883495145631068, "progress_pct": 23.14, "epoch_pct": 23.14, "eta": "53:07:26", "max_grad_norm": 0.8, "loss": 0.6258159279823303, "grad_norm": 1.018477439880371, "learning_rate": 9.347515073259657e-05}
+{"ts": "2025-12-24T06:23:12", "event": "train_log", "step": 3220, "epoch": 1.3892125134843583, "progress_pct": 23.15, "epoch_pct": 23.15, "eta": "53:05:53", "max_grad_norm": 0.8, "loss": 0.7076614499092102, "grad_norm": 1.2019606828689575, "learning_rate": 9.346301794582761e-05}
+{"ts": "2025-12-24T06:23:32", "event": "train_log", "step": 3222, "epoch": 1.3900755124056094, "progress_pct": 23.17, "epoch_pct": 23.17, "eta": "53:04:28", "max_grad_norm": 0.8, "loss": 0.5958849191665649, "grad_norm": 1.072648048400879, "learning_rate": 9.345087467814782e-05}
+{"ts": "2025-12-24T06:23:53", "event": "train_log", "step": 3224, "epoch": 1.3909385113268609, "progress_pct": 23.18, "epoch_pct": 23.18, "eta": "53:03:03", "max_grad_norm": 0.8, "loss": 0.5659268498420715, "grad_norm": 1.087908387184143, "learning_rate": 9.343872093248546e-05}
+{"ts": "2025-12-24T06:24:13", "event": "train_log", "step": 3226, "epoch": 1.3918015102481123, "progress_pct": 23.2, "epoch_pct": 23.2, "eta": "53:01:34", "max_grad_norm": 0.8, "loss": 0.6369889974594116, "grad_norm": 1.2628878355026245, "learning_rate": 9.342655671177138e-05}
+{"ts": "2025-12-24T06:24:33", "event": "train_log", "step": 3228, "epoch": 1.3926645091693635, "progress_pct": 23.21, "epoch_pct": 23.21, "eta": "53:00:07", "max_grad_norm": 0.8, "loss": 0.64048171043396, "grad_norm": 1.1580097675323486, "learning_rate": 9.341438201893891e-05}
+{"ts": "2025-12-24T06:24:52", "event": "train_log", "step": 3230, "epoch": 1.3935275080906149, "progress_pct": 23.22, "epoch_pct": 23.23, "eta": "52:58:36", "max_grad_norm": 0.8, "loss": 0.7039018273353577, "grad_norm": 1.2230379581451416, "learning_rate": 9.340219685692395e-05}
+{"ts": "2025-12-24T06:25:12", "event": "train_log", "step": 3232, "epoch": 1.3943905070118663, "progress_pct": 23.24, "epoch_pct": 23.24, "eta": "52:57:08", "max_grad_norm": 0.8, "loss": 0.6461225152015686, "grad_norm": 1.1932404041290283, "learning_rate": 9.339000122866486e-05}
+{"ts": "2025-12-24T06:25:31", "event": "train_log", "step": 3234, "epoch": 1.3952535059331175, "progress_pct": 23.25, "epoch_pct": 23.25, "eta": "52:55:35", "max_grad_norm": 0.8, "loss": 0.6538844108581543, "grad_norm": 1.4337340593338013, "learning_rate": 9.337779513710257e-05}
+{"ts": "2025-12-24T06:25:50", "event": "train_log", "step": 3236, "epoch": 1.396116504854369, "progress_pct": 23.27, "epoch_pct": 23.27, "eta": "52:54:05", "max_grad_norm": 0.8, "loss": 0.6788939237594604, "grad_norm": 1.2300320863723755, "learning_rate": 9.336557858518055e-05}
+{"ts": "2025-12-24T06:26:09", "event": "train_log", "step": 3238, "epoch": 1.3969795037756203, "progress_pct": 23.28, "epoch_pct": 23.28, "eta": "52:52:35", "max_grad_norm": 0.8, "loss": 0.6180470585823059, "grad_norm": 1.1031018495559692, "learning_rate": 9.335335157584473e-05}
+{"ts": "2025-12-24T06:26:28", "event": "train_log", "step": 3240, "epoch": 1.3978425026968715, "progress_pct": 23.3, "epoch_pct": 23.3, "eta": "52:51:05", "max_grad_norm": 0.8, "loss": 0.7097930908203125, "grad_norm": 1.1876481771469116, "learning_rate": 9.334111411204363e-05}
+{"ts": "2025-12-24T06:26:47", "event": "train_log", "step": 3242, "epoch": 1.398705501618123, "progress_pct": 23.31, "epoch_pct": 23.31, "eta": "52:49:33", "max_grad_norm": 0.8, "loss": 0.6521070003509521, "grad_norm": 1.1510672569274902, "learning_rate": 9.332886619672823e-05}
+{"ts": "2025-12-24T06:27:05", "event": "train_log", "step": 3244, "epoch": 1.3995685005393743, "progress_pct": 23.32, "epoch_pct": 23.33, "eta": "52:48:01", "max_grad_norm": 0.8, "loss": 0.6738741397857666, "grad_norm": 1.2715339660644531, "learning_rate": 9.331660783285208e-05}
+{"ts": "2025-12-24T06:27:24", "event": "train_log", "step": 3246, "epoch": 1.4004314994606257, "progress_pct": 23.34, "epoch_pct": 23.34, "eta": "52:46:30", "max_grad_norm": 0.8, "loss": 0.6261277794837952, "grad_norm": 1.1528351306915283, "learning_rate": 9.330433902337124e-05}
+{"ts": "2025-12-24T06:27:43", "event": "train_log", "step": 3248, "epoch": 1.4012944983818771, "progress_pct": 23.35, "epoch_pct": 23.35, "eta": "52:45:01", "max_grad_norm": 0.8, "loss": 0.7071060538291931, "grad_norm": 1.1715319156646729, "learning_rate": 9.329205977124424e-05}
+{"ts": "2025-12-24T06:28:03", "event": "train_log", "step": 3250, "epoch": 1.4021574973031283, "progress_pct": 23.37, "epoch_pct": 23.37, "eta": "52:43:32", "max_grad_norm": 0.8, "loss": 0.6339676380157471, "grad_norm": 1.1527135372161865, "learning_rate": 9.32797700794322e-05}
+{"ts": "2025-12-24T06:28:23", "event": "train_log", "step": 3252, "epoch": 1.4030204962243797, "progress_pct": 23.38, "epoch_pct": 23.38, "eta": "52:42:04", "max_grad_norm": 0.8, "loss": 0.6572067737579346, "grad_norm": 1.1546231508255005, "learning_rate": 9.32674699508987e-05}
+{"ts": "2025-12-24T06:28:42", "event": "train_log", "step": 3254, "epoch": 1.4038834951456312, "progress_pct": 23.4, "epoch_pct": 23.4, "eta": "52:40:35", "max_grad_norm": 0.8, "loss": 0.6182453036308289, "grad_norm": 1.1085712909698486, "learning_rate": 9.325515938860988e-05}
+{"ts": "2025-12-24T06:29:00", "event": "train_log", "step": 3256, "epoch": 1.4047464940668823, "progress_pct": 23.41, "epoch_pct": 23.41, "eta": "52:39:03", "max_grad_norm": 0.8, "loss": 0.6558552980422974, "grad_norm": 1.2346831560134888, "learning_rate": 9.324283839553437e-05}
+{"ts": "2025-12-24T06:29:20", "event": "train_log", "step": 3258, "epoch": 1.4056094929881338, "progress_pct": 23.43, "epoch_pct": 23.43, "eta": "52:37:36", "max_grad_norm": 0.8, "loss": 0.6406651735305786, "grad_norm": 1.1873912811279297, "learning_rate": 9.323050697464331e-05}
+{"ts": "2025-12-24T06:29:38", "event": "train_log", "step": 3260, "epoch": 1.4064724919093852, "progress_pct": 23.44, "epoch_pct": 23.44, "eta": "52:36:02", "max_grad_norm": 0.8, "loss": 0.6869475841522217, "grad_norm": 1.5326859951019287, "learning_rate": 9.32181651289104e-05}
+{"ts": "2025-12-24T06:29:55", "event": "train_log", "step": 3262, "epoch": 1.4073354908306364, "progress_pct": 23.45, "epoch_pct": 23.46, "eta": "52:34:27", "max_grad_norm": 0.8, "loss": 0.6805770397186279, "grad_norm": 1.2141598463058472, "learning_rate": 9.320581286131178e-05}
+{"ts": "2025-12-24T06:30:14", "event": "train_log", "step": 3264, "epoch": 1.4081984897518878, "progress_pct": 23.47, "epoch_pct": 23.47, "eta": "52:32:58", "max_grad_norm": 0.8, "loss": 0.6193398833274841, "grad_norm": 1.232690453529358, "learning_rate": 9.319345017482615e-05}
+{"ts": "2025-12-24T06:30:33", "event": "train_log", "step": 3266, "epoch": 1.4090614886731392, "progress_pct": 23.48, "epoch_pct": 23.48, "eta": "52:31:29", "max_grad_norm": 0.8, "loss": 0.5868768095970154, "grad_norm": 1.0616317987442017, "learning_rate": 9.318107707243475e-05}
+{"ts": "2025-12-24T06:30:53", "event": "train_log", "step": 3268, "epoch": 1.4099244875943904, "progress_pct": 23.5, "epoch_pct": 23.5, "eta": "52:30:01", "max_grad_norm": 0.8, "loss": 0.6360217928886414, "grad_norm": 1.091909408569336, "learning_rate": 9.316869355712127e-05}
+{"ts": "2025-12-24T06:31:13", "event": "train_log", "step": 3270, "epoch": 1.4107874865156418, "progress_pct": 23.51, "epoch_pct": 23.51, "eta": "52:28:36", "max_grad_norm": 0.8, "loss": 0.6082828044891357, "grad_norm": 1.0247875452041626, "learning_rate": 9.315629963187194e-05}
+{"ts": "2025-12-24T06:31:33", "event": "train_log", "step": 3272, "epoch": 1.4116504854368932, "progress_pct": 23.53, "epoch_pct": 23.53, "eta": "52:27:10", "max_grad_norm": 0.8, "loss": 0.6340673565864563, "grad_norm": 1.2529716491699219, "learning_rate": 9.314389529967551e-05}
+{"ts": "2025-12-24T06:31:53", "event": "train_log", "step": 3274, "epoch": 1.4125134843581446, "progress_pct": 23.54, "epoch_pct": 23.54, "eta": "52:25:43", "max_grad_norm": 0.8, "loss": 0.6333131790161133, "grad_norm": 0.9816965460777283, "learning_rate": 9.31314805635232e-05}
+{"ts": "2025-12-24T06:32:11", "event": "train_log", "step": 3276, "epoch": 1.413376483279396, "progress_pct": 23.55, "epoch_pct": 23.56, "eta": "52:24:11", "max_grad_norm": 0.8, "loss": 0.7256702780723572, "grad_norm": 1.2140034437179565, "learning_rate": 9.311905542640882e-05}
+{"ts": "2025-12-24T06:32:29", "event": "train_log", "step": 3278, "epoch": 1.4142394822006472, "progress_pct": 23.57, "epoch_pct": 23.57, "eta": "52:22:39", "max_grad_norm": 0.8, "loss": 0.7062730193138123, "grad_norm": 1.2171046733856201, "learning_rate": 9.310661989132859e-05}
+{"ts": "2025-12-24T06:32:49", "event": "train_log", "step": 3280, "epoch": 1.4151024811218986, "progress_pct": 23.58, "epoch_pct": 23.59, "eta": "52:21:13", "max_grad_norm": 0.8, "loss": 0.6192088723182678, "grad_norm": 1.0224506855010986, "learning_rate": 9.309417396128131e-05}
+{"ts": "2025-12-24T06:33:08", "event": "train_log", "step": 3282, "epoch": 1.41596548004315, "progress_pct": 23.6, "epoch_pct": 23.6, "eta": "52:19:44", "max_grad_norm": 0.8, "loss": 0.6522074341773987, "grad_norm": 1.1904140710830688, "learning_rate": 9.308171763926823e-05}
+{"ts": "2025-12-24T06:33:27", "event": "train_log", "step": 3284, "epoch": 1.4168284789644012, "progress_pct": 23.61, "epoch_pct": 23.61, "eta": "52:18:15", "max_grad_norm": 0.8, "loss": 0.6472417712211609, "grad_norm": 1.1734132766723633, "learning_rate": 9.306925092829317e-05}
+{"ts": "2025-12-24T06:33:45", "event": "train_log", "step": 3286, "epoch": 1.4176914778856526, "progress_pct": 23.63, "epoch_pct": 23.63, "eta": "52:16:45", "max_grad_norm": 0.8, "loss": 0.6990411877632141, "grad_norm": 1.2349748611450195, "learning_rate": 9.30567738313624e-05}
+{"ts": "2025-12-24T06:34:05", "event": "train_log", "step": 3288, "epoch": 1.418554476806904, "progress_pct": 23.64, "epoch_pct": 23.64, "eta": "52:15:19", "max_grad_norm": 0.8, "loss": 0.6181937456130981, "grad_norm": 1.1732749938964844, "learning_rate": 9.304428635148472e-05}
+{"ts": "2025-12-24T06:34:23", "event": "train_log", "step": 3290, "epoch": 1.4194174757281552, "progress_pct": 23.66, "epoch_pct": 23.66, "eta": "52:13:47", "max_grad_norm": 0.8, "loss": 0.7471824288368225, "grad_norm": 1.3659436702728271, "learning_rate": 9.303178849167142e-05}
+{"ts": "2025-12-24T06:34:42", "event": "train_log", "step": 3292, "epoch": 1.4202804746494067, "progress_pct": 23.67, "epoch_pct": 23.67, "eta": "52:12:20", "max_grad_norm": 0.8, "loss": 0.5541732907295227, "grad_norm": 1.0865631103515625, "learning_rate": 9.301928025493634e-05}
+{"ts": "2025-12-24T06:35:00", "event": "train_log", "step": 3294, "epoch": 1.421143473570658, "progress_pct": 23.68, "epoch_pct": 23.69, "eta": "52:10:46", "max_grad_norm": 0.8, "loss": 0.6934989094734192, "grad_norm": 1.2114793062210083, "learning_rate": 9.300676164429574e-05}
+{"ts": "2025-12-24T06:35:18", "event": "train_log", "step": 3296, "epoch": 1.4220064724919093, "progress_pct": 23.7, "epoch_pct": 23.7, "eta": "52:09:16", "max_grad_norm": 0.8, "loss": 0.6887713670730591, "grad_norm": 1.1783509254455566, "learning_rate": 9.299423266276846e-05}
+{"ts": "2025-12-24T06:35:37", "event": "train_log", "step": 3298, "epoch": 1.4228694714131607, "progress_pct": 23.71, "epoch_pct": 23.71, "eta": "52:07:49", "max_grad_norm": 0.8, "loss": 0.6956127882003784, "grad_norm": 1.3001595735549927, "learning_rate": 9.298169331337577e-05}
+{"ts": "2025-12-24T06:35:55", "event": "train_log", "step": 3300, "epoch": 1.423732470334412, "progress_pct": 23.73, "epoch_pct": 23.73, "eta": "52:06:18", "max_grad_norm": 0.8, "loss": 0.690113365650177, "grad_norm": 1.1657462120056152, "learning_rate": 9.296914359914151e-05}
+{"ts": "2025-12-24T06:49:54", "event": "train_log", "step": 3300, "epoch": 1.423732470334412, "progress_pct": 23.73, "epoch_pct": 23.73, "eta": "52:51:15", "max_grad_norm": 0.8, "eval_loss": 0.7086145281791687, "eval_runtime": 838.887, "eval_samples_per_second": 2.456, "eval_steps_per_second": 2.456}
+{"ts": "2025-12-24T06:50:13", "event": "train_log", "step": 3302, "epoch": 1.4245954692556635, "progress_pct": 23.74, "epoch_pct": 23.74, "eta": "52:49:43", "max_grad_norm": 0.8, "loss": 0.6801355481147766, "grad_norm": 1.1632269620895386, "learning_rate": 9.295658352309198e-05}
+{"ts": "2025-12-24T06:50:31", "event": "train_log", "step": 3304, "epoch": 1.425458468176915, "progress_pct": 23.76, "epoch_pct": 23.76, "eta": "52:48:11", "max_grad_norm": 0.8, "loss": 0.6801818013191223, "grad_norm": 1.3355532884597778, "learning_rate": 9.2944013088256e-05}
+{"ts": "2025-12-24T06:50:50", "event": "train_log", "step": 3306, "epoch": 1.426321467098166, "progress_pct": 23.77, "epoch_pct": 23.77, "eta": "52:46:39", "max_grad_norm": 0.8, "loss": 0.633465051651001, "grad_norm": 1.0911835432052612, "learning_rate": 9.293143229766484e-05}
+{"ts": "2025-12-24T06:51:09", "event": "train_log", "step": 3308, "epoch": 1.4271844660194175, "progress_pct": 23.78, "epoch_pct": 23.79, "eta": "52:45:11", "max_grad_norm": 0.8, "loss": 0.6677361130714417, "grad_norm": 1.1873823404312134, "learning_rate": 9.29188411543523e-05}
+{"ts": "2025-12-24T06:51:29", "event": "train_log", "step": 3310, "epoch": 1.428047464940669, "progress_pct": 23.8, "epoch_pct": 23.8, "eta": "52:43:44", "max_grad_norm": 0.8, "loss": 0.6741982102394104, "grad_norm": 1.2115873098373413, "learning_rate": 9.290623966135472e-05}
+{"ts": "2025-12-24T06:51:47", "event": "train_log", "step": 3312, "epoch": 1.42891046386192, "progress_pct": 23.81, "epoch_pct": 23.82, "eta": "52:42:11", "max_grad_norm": 0.8, "loss": 0.7169438600540161, "grad_norm": 1.2389469146728516, "learning_rate": 9.289362782171084e-05}
+{"ts": "2025-12-24T06:52:06", "event": "train_log", "step": 3314, "epoch": 1.4297734627831715, "progress_pct": 23.83, "epoch_pct": 23.83, "eta": "52:40:42", "max_grad_norm": 0.8, "loss": 0.7112365961074829, "grad_norm": 1.2088500261306763, "learning_rate": 9.288100563846201e-05}
+{"ts": "2025-12-24T06:52:25", "event": "train_log", "step": 3316, "epoch": 1.430636461704423, "progress_pct": 23.84, "epoch_pct": 23.84, "eta": "52:39:13", "max_grad_norm": 0.8, "loss": 0.6910304427146912, "grad_norm": 1.1279292106628418, "learning_rate": 9.286837311465196e-05}
+{"ts": "2025-12-24T06:52:43", "event": "train_log", "step": 3318, "epoch": 1.4314994606256741, "progress_pct": 23.86, "epoch_pct": 23.86, "eta": "52:37:42", "max_grad_norm": 0.8, "loss": 0.7288362979888916, "grad_norm": 1.1812585592269897, "learning_rate": 9.285573025332698e-05}
+{"ts": "2025-12-24T06:53:03", "event": "train_log", "step": 3320, "epoch": 1.4323624595469255, "progress_pct": 23.87, "epoch_pct": 23.87, "eta": "52:36:14", "max_grad_norm": 0.8, "loss": 0.6918224692344666, "grad_norm": 1.5522732734680176, "learning_rate": 9.284307705753587e-05}
+{"ts": "2025-12-24T06:53:22", "event": "train_log", "step": 3322, "epoch": 1.433225458468177, "progress_pct": 23.89, "epoch_pct": 23.89, "eta": "52:34:46", "max_grad_norm": 0.8, "loss": 0.6756533980369568, "grad_norm": 1.2825061082839966, "learning_rate": 9.283041353032984e-05}
+{"ts": "2025-12-24T06:53:42", "event": "train_log", "step": 3324, "epoch": 1.4340884573894281, "progress_pct": 23.9, "epoch_pct": 23.9, "eta": "52:33:20", "max_grad_norm": 0.8, "loss": 0.6005351543426514, "grad_norm": 1.0962797403335571, "learning_rate": 9.281773967476269e-05}
+{"ts": "2025-12-24T06:54:02", "event": "train_log", "step": 3326, "epoch": 1.4349514563106796, "progress_pct": 23.91, "epoch_pct": 23.92, "eta": "52:31:53", "max_grad_norm": 0.8, "loss": 0.6576942205429077, "grad_norm": 1.249868631362915, "learning_rate": 9.280505549389063e-05}
+{"ts": "2025-12-24T06:54:19", "event": "train_log", "step": 3328, "epoch": 1.435814455231931, "progress_pct": 23.93, "epoch_pct": 23.93, "eta": "52:30:19", "max_grad_norm": 0.8, "loss": 0.6686707735061646, "grad_norm": 1.2587590217590332, "learning_rate": 9.279236099077239e-05}
+{"ts": "2025-12-24T06:54:40", "event": "train_log", "step": 3330, "epoch": 1.4366774541531824, "progress_pct": 23.94, "epoch_pct": 23.94, "eta": "52:28:54", "max_grad_norm": 0.8, "loss": 0.5858182311058044, "grad_norm": 1.0961607694625854, "learning_rate": 9.277965616846922e-05}
+{"ts": "2025-12-24T06:54:58", "event": "train_log", "step": 3332, "epoch": 1.4375404530744338, "progress_pct": 23.96, "epoch_pct": 23.96, "eta": "52:27:25", "max_grad_norm": 0.8, "loss": 0.6661836504936218, "grad_norm": 1.199666142463684, "learning_rate": 9.276694103004481e-05}
+{"ts": "2025-12-24T06:55:17", "event": "train_log", "step": 3334, "epoch": 1.438403451995685, "progress_pct": 23.97, "epoch_pct": 23.97, "eta": "52:25:53", "max_grad_norm": 0.8, "loss": 0.6922421455383301, "grad_norm": 1.2444990873336792, "learning_rate": 9.275421557856536e-05}
+{"ts": "2025-12-24T06:55:35", "event": "train_log", "step": 3336, "epoch": 1.4392664509169364, "progress_pct": 23.99, "epoch_pct": 23.99, "eta": "52:24:22", "max_grad_norm": 0.8, "loss": 0.750586986541748, "grad_norm": 1.3198295831680298, "learning_rate": 9.274147981709957e-05}
+{"ts": "2025-12-24T06:55:53", "event": "train_log", "step": 3338, "epoch": 1.4401294498381878, "progress_pct": 24.0, "epoch_pct": 24.0, "eta": "52:22:51", "max_grad_norm": 0.8, "loss": 0.6817078590393066, "grad_norm": 1.0782018899917603, "learning_rate": 9.27287337487186e-05}
+{"ts": "2025-12-24T06:56:13", "event": "train_log", "step": 3340, "epoch": 1.440992448759439, "progress_pct": 24.01, "epoch_pct": 24.02, "eta": "52:21:27", "max_grad_norm": 0.8, "loss": 0.6163688898086548, "grad_norm": 1.1463382244110107, "learning_rate": 9.271597737649611e-05}
+{"ts": "2025-12-24T06:56:32", "event": "train_log", "step": 3342, "epoch": 1.4418554476806904, "progress_pct": 24.03, "epoch_pct": 24.03, "eta": "52:19:57", "max_grad_norm": 0.8, "loss": 0.6624670028686523, "grad_norm": 1.0618164539337158, "learning_rate": 9.270321070350825e-05}
+{"ts": "2025-12-24T06:56:51", "event": "train_log", "step": 3344, "epoch": 1.4427184466019418, "progress_pct": 24.04, "epoch_pct": 24.05, "eta": "52:18:29", "max_grad_norm": 0.8, "loss": 0.6802749037742615, "grad_norm": 1.1389427185058594, "learning_rate": 9.269043373283362e-05}
+{"ts": "2025-12-24T06:57:10", "event": "train_log", "step": 3346, "epoch": 1.443581445523193, "progress_pct": 24.06, "epoch_pct": 24.06, "eta": "52:17:02", "max_grad_norm": 0.8, "loss": 0.6428707838058472, "grad_norm": 1.082173466682434, "learning_rate": 9.267764646755335e-05}
+{"ts": "2025-12-24T06:57:29", "event": "train_log", "step": 3348, "epoch": 1.4444444444444444, "progress_pct": 24.07, "epoch_pct": 24.07, "eta": "52:15:31", "max_grad_norm": 0.8, "loss": 0.6853904128074646, "grad_norm": 1.1741442680358887, "learning_rate": 9.266484891075103e-05}
+{"ts": "2025-12-24T06:57:47", "event": "train_log", "step": 3350, "epoch": 1.4453074433656958, "progress_pct": 24.09, "epoch_pct": 24.09, "eta": "52:14:02", "max_grad_norm": 0.8, "loss": 0.6889524459838867, "grad_norm": 1.1328803300857544, "learning_rate": 9.265204106551273e-05}
+{"ts": "2025-12-24T06:58:06", "event": "train_log", "step": 3352, "epoch": 1.446170442286947, "progress_pct": 24.1, "epoch_pct": 24.1, "eta": "52:12:34", "max_grad_norm": 0.8, "loss": 0.6037911176681519, "grad_norm": 1.0607874393463135, "learning_rate": 9.2639222934927e-05}
+{"ts": "2025-12-24T06:58:25", "event": "train_log", "step": 3354, "epoch": 1.4470334412081984, "progress_pct": 24.12, "epoch_pct": 24.12, "eta": "52:11:07", "max_grad_norm": 0.8, "loss": 0.6671991944313049, "grad_norm": 1.1584129333496094, "learning_rate": 9.262639452208487e-05}
+{"ts": "2025-12-24T06:58:44", "event": "train_log", "step": 3356, "epoch": 1.4478964401294498, "progress_pct": 24.13, "epoch_pct": 24.13, "eta": "52:09:38", "max_grad_norm": 0.8, "loss": 0.7165482044219971, "grad_norm": 1.1881403923034668, "learning_rate": 9.261355583007987e-05}
+{"ts": "2025-12-24T06:59:03", "event": "train_log", "step": 3358, "epoch": 1.4487594390507013, "progress_pct": 24.14, "epoch_pct": 24.15, "eta": "52:08:10", "max_grad_norm": 0.8, "loss": 0.7108280062675476, "grad_norm": 1.2093054056167603, "learning_rate": 9.260070686200797e-05}
+{"ts": "2025-12-24T06:59:21", "event": "train_log", "step": 3360, "epoch": 1.4496224379719527, "progress_pct": 24.16, "epoch_pct": 24.16, "eta": "52:06:41", "max_grad_norm": 0.8, "loss": 0.718152642250061, "grad_norm": 1.1105186939239502, "learning_rate": 9.258784762096768e-05}
+{"ts": "2025-12-24T06:59:42", "event": "train_log", "step": 3362, "epoch": 1.4504854368932039, "progress_pct": 24.17, "epoch_pct": 24.17, "eta": "52:05:16", "max_grad_norm": 0.8, "loss": 0.6449500322341919, "grad_norm": 1.028255820274353, "learning_rate": 9.25749781100599e-05}
+{"ts": "2025-12-24T07:00:01", "event": "train_log", "step": 3364, "epoch": 1.4513484358144553, "progress_pct": 24.19, "epoch_pct": 24.19, "eta": "52:03:51", "max_grad_norm": 0.8, "loss": 0.6895458102226257, "grad_norm": 1.235719919204712, "learning_rate": 9.25620983323881e-05}
+{"ts": "2025-12-24T07:00:20", "event": "train_log", "step": 3366, "epoch": 1.4522114347357067, "progress_pct": 24.2, "epoch_pct": 24.2, "eta": "52:02:22", "max_grad_norm": 0.8, "loss": 0.7243028879165649, "grad_norm": 1.3693572282791138, "learning_rate": 9.254920829105816e-05}
+{"ts": "2025-12-24T07:00:39", "event": "train_log", "step": 3368, "epoch": 1.4530744336569579, "progress_pct": 24.22, "epoch_pct": 24.22, "eta": "52:00:57", "max_grad_norm": 0.8, "loss": 0.6177746057510376, "grad_norm": 1.1795028448104858, "learning_rate": 9.253630798917843e-05}
+{"ts": "2025-12-24T07:00:58", "event": "train_log", "step": 3370, "epoch": 1.4539374325782093, "progress_pct": 24.23, "epoch_pct": 24.23, "eta": "51:59:29", "max_grad_norm": 0.8, "loss": 0.6603507399559021, "grad_norm": 1.0759676694869995, "learning_rate": 9.25233974298598e-05}
+{"ts": "2025-12-24T07:01:19", "event": "train_log", "step": 3372, "epoch": 1.4548004314994607, "progress_pct": 24.25, "epoch_pct": 24.25, "eta": "51:58:08", "max_grad_norm": 0.8, "loss": 0.5989331603050232, "grad_norm": 0.9495601058006287, "learning_rate": 9.251047661621556e-05}
+{"ts": "2025-12-24T07:01:39", "event": "train_log", "step": 3374, "epoch": 1.455663430420712, "progress_pct": 24.26, "epoch_pct": 24.26, "eta": "51:56:43", "max_grad_norm": 0.8, "loss": 0.5743664503097534, "grad_norm": 1.1334787607192993, "learning_rate": 9.249754555136154e-05}
+{"ts": "2025-12-24T07:02:00", "event": "train_log", "step": 3376, "epoch": 1.4565264293419633, "progress_pct": 24.27, "epoch_pct": 24.28, "eta": "51:55:24", "max_grad_norm": 0.8, "loss": 0.5905263423919678, "grad_norm": 1.114450216293335, "learning_rate": 9.248460423841598e-05}
+{"ts": "2025-12-24T07:02:19", "event": "train_log", "step": 3378, "epoch": 1.4573894282632147, "progress_pct": 24.29, "epoch_pct": 24.29, "eta": "51:53:57", "max_grad_norm": 0.8, "loss": 0.6751091480255127, "grad_norm": 1.2400883436203003, "learning_rate": 9.247165268049963e-05}
+{"ts": "2025-12-24T07:02:40", "event": "train_log", "step": 3380, "epoch": 1.458252427184466, "progress_pct": 24.3, "epoch_pct": 24.3, "eta": "51:52:34", "max_grad_norm": 0.8, "loss": 0.6872944235801697, "grad_norm": 1.2845438718795776, "learning_rate": 9.245869088073569e-05}
+{"ts": "2025-12-24T07:03:01", "event": "train_log", "step": 3382, "epoch": 1.4591154261057173, "progress_pct": 24.32, "epoch_pct": 24.32, "eta": "51:51:14", "max_grad_norm": 0.8, "loss": 0.6467806100845337, "grad_norm": 1.054388165473938, "learning_rate": 9.244571884224987e-05}
+{"ts": "2025-12-24T07:03:20", "event": "train_log", "step": 3384, "epoch": 1.4599784250269687, "progress_pct": 24.33, "epoch_pct": 24.33, "eta": "51:49:49", "max_grad_norm": 0.8, "loss": 0.6486450433731079, "grad_norm": 1.0959769487380981, "learning_rate": 9.243273656817029e-05}
+{"ts": "2025-12-24T07:03:37", "event": "train_log", "step": 3386, "epoch": 1.4608414239482201, "progress_pct": 24.35, "epoch_pct": 24.35, "eta": "51:48:16", "max_grad_norm": 0.8, "loss": 0.7024005651473999, "grad_norm": 1.6487104892730713, "learning_rate": 9.241974406162756e-05}
+{"ts": "2025-12-24T07:03:58", "event": "train_log", "step": 3388, "epoch": 1.4617044228694713, "progress_pct": 24.36, "epoch_pct": 24.36, "eta": "51:46:54", "max_grad_norm": 0.8, "loss": 0.7081685662269592, "grad_norm": 1.3018934726715088, "learning_rate": 9.240674132575481e-05}
+{"ts": "2025-12-24T07:04:16", "event": "train_log", "step": 3390, "epoch": 1.4625674217907227, "progress_pct": 24.37, "epoch_pct": 24.38, "eta": "51:45:26", "max_grad_norm": 0.8, "loss": 0.5842618346214294, "grad_norm": 0.9295057058334351, "learning_rate": 9.239372836368754e-05}
+{"ts": "2025-12-24T07:04:37", "event": "train_log", "step": 3392, "epoch": 1.4634304207119742, "progress_pct": 24.39, "epoch_pct": 24.39, "eta": "51:44:06", "max_grad_norm": 0.8, "loss": 0.6355351209640503, "grad_norm": 1.0556540489196777, "learning_rate": 9.238070517856378e-05}
+{"ts": "2025-12-24T07:04:55", "event": "train_log", "step": 3394, "epoch": 1.4642934196332256, "progress_pct": 24.4, "epoch_pct": 24.4, "eta": "51:42:37", "max_grad_norm": 0.8, "loss": 0.6340579390525818, "grad_norm": 1.2247962951660156, "learning_rate": 9.236767177352403e-05}
+{"ts": "2025-12-24T07:05:16", "event": "train_log", "step": 3396, "epoch": 1.4651564185544768, "progress_pct": 24.42, "epoch_pct": 24.42, "eta": "51:41:17", "max_grad_norm": 0.8, "loss": 0.6405525803565979, "grad_norm": 1.1625045537948608, "learning_rate": 9.235462815171123e-05}
+{"ts": "2025-12-24T07:05:37", "event": "train_log", "step": 3398, "epoch": 1.4660194174757282, "progress_pct": 24.43, "epoch_pct": 24.43, "eta": "51:39:56", "max_grad_norm": 0.8, "loss": 0.5884681940078735, "grad_norm": 1.009028673171997, "learning_rate": 9.234157431627077e-05}
+{"ts": "2025-12-24T07:05:58", "event": "train_log", "step": 3400, "epoch": 1.4668824163969796, "progress_pct": 24.45, "epoch_pct": 24.45, "eta": "51:38:36", "max_grad_norm": 0.8, "loss": 0.6532622575759888, "grad_norm": 1.174851894378662, "learning_rate": 9.232851027035056e-05}
+{"ts": "2025-12-24T07:19:48", "event": "train_log", "step": 3400, "epoch": 1.4668824163969796, "progress_pct": 24.45, "epoch_pct": 24.45, "eta": "52:21:21", "max_grad_norm": 0.8, "eval_loss": 0.7067069411277771, "eval_runtime": 829.9411, "eval_samples_per_second": 2.482, "eval_steps_per_second": 2.482}
+{"ts": "2025-12-24T07:20:07", "event": "train_log", "step": 3402, "epoch": 1.4677454153182308, "progress_pct": 24.46, "epoch_pct": 24.46, "eta": "52:19:54", "max_grad_norm": 0.8, "loss": 0.6505036354064941, "grad_norm": 0.9843166470527649, "learning_rate": 9.231543601710091e-05}
+{"ts": "2025-12-24T07:20:26", "event": "train_log", "step": 3404, "epoch": 1.4686084142394822, "progress_pct": 24.48, "epoch_pct": 24.48, "eta": "52:18:24", "max_grad_norm": 0.8, "loss": 0.5809159874916077, "grad_norm": 1.1260782480239868, "learning_rate": 9.230235155967461e-05}
+{"ts": "2025-12-24T07:20:46", "event": "train_log", "step": 3406, "epoch": 1.4694714131607336, "progress_pct": 24.49, "epoch_pct": 24.49, "eta": "52:17:01", "max_grad_norm": 0.8, "loss": 0.6431854367256165, "grad_norm": 1.3986146450042725, "learning_rate": 9.228925690122696e-05}
+{"ts": "2025-12-24T07:21:04", "event": "train_log", "step": 3408, "epoch": 1.4703344120819848, "progress_pct": 24.5, "epoch_pct": 24.51, "eta": "52:15:31", "max_grad_norm": 0.8, "loss": 0.6126174330711365, "grad_norm": 1.145108938217163, "learning_rate": 9.227615204491562e-05}
+{"ts": "2025-12-24T07:21:23", "event": "train_log", "step": 3410, "epoch": 1.4711974110032362, "progress_pct": 24.52, "epoch_pct": 24.52, "eta": "52:14:01", "max_grad_norm": 0.8, "loss": 0.6271735429763794, "grad_norm": 1.0973509550094604, "learning_rate": 9.22630369939008e-05}
+{"ts": "2025-12-24T07:21:41", "event": "train_log", "step": 3412, "epoch": 1.4720604099244876, "progress_pct": 24.53, "epoch_pct": 24.53, "eta": "52:12:33", "max_grad_norm": 0.8, "loss": 0.5894950032234192, "grad_norm": 1.0548614263534546, "learning_rate": 9.224991175134515e-05}
+{"ts": "2025-12-24T07:22:00", "event": "train_log", "step": 3414, "epoch": 1.472923408845739, "progress_pct": 24.55, "epoch_pct": 24.55, "eta": "52:11:05", "max_grad_norm": 0.8, "loss": 0.6621272563934326, "grad_norm": 1.3288220167160034, "learning_rate": 9.223677632041374e-05}
+{"ts": "2025-12-24T07:22:20", "event": "train_log", "step": 3416, "epoch": 1.4737864077669902, "progress_pct": 24.56, "epoch_pct": 24.56, "eta": "52:09:38", "max_grad_norm": 0.8, "loss": 0.6911466717720032, "grad_norm": 1.0312891006469727, "learning_rate": 9.222363070427411e-05}
+{"ts": "2025-12-24T07:22:40", "event": "train_log", "step": 3418, "epoch": 1.4746494066882416, "progress_pct": 24.58, "epoch_pct": 24.58, "eta": "52:08:14", "max_grad_norm": 0.8, "loss": 0.5945214033126831, "grad_norm": 1.1039631366729736, "learning_rate": 9.221047490609629e-05}
+{"ts": "2025-12-24T07:22:59", "event": "train_log", "step": 3420, "epoch": 1.475512405609493, "progress_pct": 24.59, "epoch_pct": 24.59, "eta": "52:06:48", "max_grad_norm": 0.8, "loss": 0.6231427192687988, "grad_norm": 1.1555835008621216, "learning_rate": 9.219730892905273e-05}
+{"ts": "2025-12-24T07:23:18", "event": "train_log", "step": 3422, "epoch": 1.4763754045307445, "progress_pct": 24.6, "epoch_pct": 24.61, "eta": "52:05:21", "max_grad_norm": 0.8, "loss": 0.6413036584854126, "grad_norm": 1.0504940748214722, "learning_rate": 9.218413277631835e-05}
+{"ts": "2025-12-24T07:23:37", "event": "train_log", "step": 3424, "epoch": 1.4772384034519956, "progress_pct": 24.62, "epoch_pct": 24.62, "eta": "52:03:55", "max_grad_norm": 0.8, "loss": 0.663884699344635, "grad_norm": 1.0855275392532349, "learning_rate": 9.217094645107053e-05}
+{"ts": "2025-12-24T07:23:57", "event": "train_log", "step": 3426, "epoch": 1.478101402373247, "progress_pct": 24.63, "epoch_pct": 24.64, "eta": "52:02:30", "max_grad_norm": 0.8, "loss": 0.6161345839500427, "grad_norm": 1.068493366241455, "learning_rate": 9.215774995648907e-05}
+{"ts": "2025-12-24T07:24:17", "event": "train_log", "step": 3428, "epoch": 1.4789644012944985, "progress_pct": 24.65, "epoch_pct": 24.65, "eta": "52:01:06", "max_grad_norm": 0.8, "loss": 0.6693077087402344, "grad_norm": 1.0962451696395874, "learning_rate": 9.214454329575626e-05}
+{"ts": "2025-12-24T07:24:38", "event": "train_log", "step": 3430, "epoch": 1.4798274002157497, "progress_pct": 24.66, "epoch_pct": 24.66, "eta": "51:59:45", "max_grad_norm": 0.8, "loss": 0.6362645030021667, "grad_norm": 1.070582389831543, "learning_rate": 9.213132647205682e-05}
+{"ts": "2025-12-24T07:24:57", "event": "train_log", "step": 3432, "epoch": 1.480690399137001, "progress_pct": 24.68, "epoch_pct": 24.68, "eta": "51:58:17", "max_grad_norm": 0.8, "loss": 0.7378971576690674, "grad_norm": 1.201202154159546, "learning_rate": 9.211809948857793e-05}
+{"ts": "2025-12-24T07:25:16", "event": "train_log", "step": 3434, "epoch": 1.4815533980582525, "progress_pct": 24.69, "epoch_pct": 24.69, "eta": "51:56:53", "max_grad_norm": 0.8, "loss": 0.686621367931366, "grad_norm": 1.1988805532455444, "learning_rate": 9.210486234850921e-05}
+{"ts": "2025-12-24T07:25:36", "event": "train_log", "step": 3436, "epoch": 1.4824163969795037, "progress_pct": 24.71, "epoch_pct": 24.71, "eta": "51:55:28", "max_grad_norm": 0.8, "loss": 0.6873248219490051, "grad_norm": 1.2886041402816772, "learning_rate": 9.209161505504274e-05}
+{"ts": "2025-12-24T07:25:57", "event": "train_log", "step": 3438, "epoch": 1.483279395900755, "progress_pct": 24.72, "epoch_pct": 24.72, "eta": "51:54:07", "max_grad_norm": 0.8, "loss": 0.624991774559021, "grad_norm": 1.0028132200241089, "learning_rate": 9.207835761137307e-05}
+{"ts": "2025-12-24T07:26:18", "event": "train_log", "step": 3440, "epoch": 1.4841423948220065, "progress_pct": 24.73, "epoch_pct": 24.74, "eta": "51:52:47", "max_grad_norm": 0.8, "loss": 0.6064573526382446, "grad_norm": 1.139201045036316, "learning_rate": 9.206509002069715e-05}
+{"ts": "2025-12-24T07:26:37", "event": "train_log", "step": 3442, "epoch": 1.4850053937432577, "progress_pct": 24.75, "epoch_pct": 24.75, "eta": "51:51:21", "max_grad_norm": 0.8, "loss": 0.6139655709266663, "grad_norm": 1.0371732711791992, "learning_rate": 9.205181228621438e-05}
+{"ts": "2025-12-24T07:26:58", "event": "train_log", "step": 3444, "epoch": 1.485868392664509, "progress_pct": 24.76, "epoch_pct": 24.76, "eta": "51:50:00", "max_grad_norm": 0.8, "loss": 0.6176325082778931, "grad_norm": 1.1208233833312988, "learning_rate": 9.203852441112665e-05}
+{"ts": "2025-12-24T07:27:18", "event": "train_log", "step": 3446, "epoch": 1.4867313915857605, "progress_pct": 24.78, "epoch_pct": 24.78, "eta": "51:48:38", "max_grad_norm": 0.8, "loss": 0.6920754313468933, "grad_norm": 1.2438154220581055, "learning_rate": 9.20252263986383e-05}
+{"ts": "2025-12-24T07:27:38", "event": "train_log", "step": 3448, "epoch": 1.487594390507012, "progress_pct": 24.79, "epoch_pct": 24.79, "eta": "51:47:15", "max_grad_norm": 0.8, "loss": 0.6859762072563171, "grad_norm": 1.2117916345596313, "learning_rate": 9.201191825195604e-05}
+{"ts": "2025-12-24T07:27:58", "event": "train_log", "step": 3450, "epoch": 1.4884573894282633, "progress_pct": 24.81, "epoch_pct": 24.81, "eta": "51:45:52", "max_grad_norm": 0.8, "loss": 0.7358171939849854, "grad_norm": 1.2274118661880493, "learning_rate": 9.199859997428908e-05}
+{"ts": "2025-12-24T07:28:17", "event": "train_log", "step": 3452, "epoch": 1.4893203883495145, "progress_pct": 24.82, "epoch_pct": 24.82, "eta": "51:44:26", "max_grad_norm": 0.8, "loss": 0.6878830194473267, "grad_norm": 1.2810527086257935, "learning_rate": 9.198527156884907e-05}
+{"ts": "2025-12-24T07:28:37", "event": "train_log", "step": 3454, "epoch": 1.490183387270766, "progress_pct": 24.83, "epoch_pct": 24.84, "eta": "51:43:02", "max_grad_norm": 0.8, "loss": 0.6995986104011536, "grad_norm": 1.2344684600830078, "learning_rate": 9.197193303885008e-05}
+{"ts": "2025-12-24T07:28:56", "event": "train_log", "step": 3456, "epoch": 1.4910463861920173, "progress_pct": 24.85, "epoch_pct": 24.85, "eta": "51:41:37", "max_grad_norm": 0.8, "loss": 0.6423236131668091, "grad_norm": 1.233686923980713, "learning_rate": 9.195858438750866e-05}
+{"ts": "2025-12-24T07:29:15", "event": "train_log", "step": 3458, "epoch": 1.4919093851132685, "progress_pct": 24.86, "epoch_pct": 24.87, "eta": "51:40:10", "max_grad_norm": 0.8, "loss": 0.7513291835784912, "grad_norm": 1.3485876321792603, "learning_rate": 9.194522561804377e-05}
+{"ts": "2025-12-24T07:29:35", "event": "train_log", "step": 3460, "epoch": 1.49277238403452, "progress_pct": 24.88, "epoch_pct": 24.88, "eta": "51:38:47", "max_grad_norm": 0.8, "loss": 0.6667183637619019, "grad_norm": 0.9913640022277832, "learning_rate": 9.19318567336768e-05}
+{"ts": "2025-12-24T07:29:54", "event": "train_log", "step": 3462, "epoch": 1.4936353829557714, "progress_pct": 24.89, "epoch_pct": 24.89, "eta": "51:37:23", "max_grad_norm": 0.8, "loss": 0.643630862236023, "grad_norm": 1.146099328994751, "learning_rate": 9.191847773763161e-05}
+{"ts": "2025-12-24T07:30:15", "event": "train_log", "step": 3464, "epoch": 1.4944983818770226, "progress_pct": 24.91, "epoch_pct": 24.91, "eta": "51:36:01", "max_grad_norm": 0.8, "loss": 0.655176043510437, "grad_norm": 1.0300167798995972, "learning_rate": 9.190508863313448e-05}
+{"ts": "2025-12-24T07:30:34", "event": "train_log", "step": 3466, "epoch": 1.495361380798274, "progress_pct": 24.92, "epoch_pct": 24.92, "eta": "51:34:37", "max_grad_norm": 0.8, "loss": 0.6792252659797668, "grad_norm": 1.285008430480957, "learning_rate": 9.189168942341416e-05}
+{"ts": "2025-12-24T07:30:52", "event": "train_log", "step": 3468, "epoch": 1.4962243797195254, "progress_pct": 24.94, "epoch_pct": 24.94, "eta": "51:33:09", "max_grad_norm": 0.8, "loss": 0.7081018686294556, "grad_norm": 1.1455490589141846, "learning_rate": 9.187828011170175e-05}
+{"ts": "2025-12-24T07:31:11", "event": "train_log", "step": 3470, "epoch": 1.4970873786407766, "progress_pct": 24.95, "epoch_pct": 24.95, "eta": "51:31:42", "max_grad_norm": 0.8, "loss": 0.6613248586654663, "grad_norm": 1.2560410499572754, "learning_rate": 9.186486070123091e-05}
+{"ts": "2025-12-24T07:31:31", "event": "train_log", "step": 3472, "epoch": 1.497950377562028, "progress_pct": 24.96, "epoch_pct": 24.97, "eta": "51:30:22", "max_grad_norm": 0.8, "loss": 0.6299360990524292, "grad_norm": 1.2391352653503418, "learning_rate": 9.185143119523761e-05}
+{"ts": "2025-12-24T07:31:51", "event": "train_log", "step": 3474, "epoch": 1.4988133764832794, "progress_pct": 24.98, "epoch_pct": 24.98, "eta": "51:28:59", "max_grad_norm": 0.8, "loss": 0.5910941362380981, "grad_norm": 1.134585976600647, "learning_rate": 9.183799159696035e-05}
+{"ts": "2025-12-24T07:32:11", "event": "train_log", "step": 3476, "epoch": 1.4996763754045308, "progress_pct": 24.99, "epoch_pct": 24.99, "eta": "51:27:38", "max_grad_norm": 0.8, "loss": 0.7077166438102722, "grad_norm": 1.1744983196258545, "learning_rate": 9.182454190964005e-05}
+{"ts": "2025-12-24T07:32:32", "event": "train_log", "step": 3478, "epoch": 1.5005393743257822, "progress_pct": 25.01, "epoch_pct": 25.01, "eta": "51:26:17", "max_grad_norm": 0.8, "loss": 0.6168769598007202, "grad_norm": 1.1142816543579102, "learning_rate": 9.181108213651999e-05}
+{"ts": "2025-12-24T07:32:50", "event": "train_log", "step": 3480, "epoch": 1.5014023732470334, "progress_pct": 25.02, "epoch_pct": 25.02, "eta": "51:24:49", "max_grad_norm": 0.8, "loss": 0.6587874293327332, "grad_norm": 1.1463894844055176, "learning_rate": 9.179761228084597e-05}
+{"ts": "2025-12-24T07:33:07", "event": "train_log", "step": 3482, "epoch": 1.5022653721682848, "progress_pct": 25.04, "epoch_pct": 25.04, "eta": "51:23:18", "max_grad_norm": 0.8, "loss": 0.7696266770362854, "grad_norm": 1.3945616483688354, "learning_rate": 9.178413234586616e-05}
+{"ts": "2025-12-24T07:33:26", "event": "train_log", "step": 3484, "epoch": 1.5031283710895362, "progress_pct": 25.05, "epoch_pct": 25.05, "eta": "51:21:55", "max_grad_norm": 0.8, "loss": 0.7142559885978699, "grad_norm": 1.3266242742538452, "learning_rate": 9.177064233483121e-05}
+{"ts": "2025-12-24T07:33:45", "event": "train_log", "step": 3486, "epoch": 1.5039913700107874, "progress_pct": 25.06, "epoch_pct": 25.07, "eta": "51:20:30", "max_grad_norm": 0.8, "loss": 0.7085118293762207, "grad_norm": 1.1857004165649414, "learning_rate": 9.175714225099417e-05}
+{"ts": "2025-12-24T07:34:04", "event": "train_log", "step": 3488, "epoch": 1.5048543689320388, "progress_pct": 25.08, "epoch_pct": 25.08, "eta": "51:19:05", "max_grad_norm": 0.8, "loss": 0.6643039584159851, "grad_norm": 1.1070430278778076, "learning_rate": 9.174363209761049e-05}
+{"ts": "2025-12-24T07:34:25", "event": "train_log", "step": 3490, "epoch": 1.5057173678532902, "progress_pct": 25.09, "epoch_pct": 25.1, "eta": "51:17:45", "max_grad_norm": 0.8, "loss": 0.6282537579536438, "grad_norm": 1.0877243280410767, "learning_rate": 9.173011187793814e-05}
+{"ts": "2025-12-24T07:34:45", "event": "train_log", "step": 3492, "epoch": 1.5065803667745414, "progress_pct": 25.11, "epoch_pct": 25.11, "eta": "51:16:24", "max_grad_norm": 0.8, "loss": 0.6811800599098206, "grad_norm": 1.2648247480392456, "learning_rate": 9.171658159523741e-05}
+{"ts": "2025-12-24T07:35:05", "event": "train_log", "step": 3494, "epoch": 1.5074433656957928, "progress_pct": 25.12, "epoch_pct": 25.12, "eta": "51:15:02", "max_grad_norm": 0.8, "loss": 0.6442922353744507, "grad_norm": 1.1535227298736572, "learning_rate": 9.170304125277111e-05}
+{"ts": "2025-12-24T07:35:25", "event": "train_log", "step": 3496, "epoch": 1.5083063646170443, "progress_pct": 25.14, "epoch_pct": 25.14, "eta": "51:13:40", "max_grad_norm": 0.8, "loss": 0.717490553855896, "grad_norm": 1.1175323724746704, "learning_rate": 9.16894908538044e-05}
+{"ts": "2025-12-24T07:35:44", "event": "train_log", "step": 3498, "epoch": 1.5091693635382954, "progress_pct": 25.15, "epoch_pct": 25.15, "eta": "51:12:17", "max_grad_norm": 0.8, "loss": 0.6537191867828369, "grad_norm": 1.1923470497131348, "learning_rate": 9.167593040160491e-05}
+{"ts": "2025-12-24T07:36:01", "event": "train_log", "step": 3500, "epoch": 1.510032362459547, "progress_pct": 25.17, "epoch_pct": 25.17, "eta": "51:10:48", "max_grad_norm": 0.8, "loss": 0.7314199805259705, "grad_norm": 1.3302042484283447, "learning_rate": 9.166235989944268e-05}
+{"ts": "2025-12-24T07:49:43", "event": "train_log", "step": 3500, "epoch": 1.510032362459547, "progress_pct": 25.17, "epoch_pct": 25.17, "eta": "51:51:32", "max_grad_norm": 0.8, "eval_loss": 0.7047909498214722, "eval_runtime": 821.8811, "eval_samples_per_second": 2.506, "eval_steps_per_second": 2.506}
+{"ts": "2025-12-24T07:50:04", "event": "train_log", "step": 3502, "epoch": 1.5108953613807983, "progress_pct": 25.18, "epoch_pct": 25.18, "eta": "51:50:10", "max_grad_norm": 0.8, "loss": 0.6363071203231812, "grad_norm": 1.1480292081832886, "learning_rate": 9.164877935059017e-05}
+{"ts": "2025-12-24T07:50:23", "event": "train_log", "step": 3504, "epoch": 1.5117583603020495, "progress_pct": 25.19, "epoch_pct": 25.2, "eta": "51:48:45", "max_grad_norm": 0.8, "loss": 0.6020049452781677, "grad_norm": 1.1068446636199951, "learning_rate": 9.163518875832227e-05}
+{"ts": "2025-12-24T07:50:43", "event": "train_log", "step": 3506, "epoch": 1.512621359223301, "progress_pct": 25.21, "epoch_pct": 25.21, "eta": "51:47:21", "max_grad_norm": 0.8, "loss": 0.7053841352462769, "grad_norm": 1.2131398916244507, "learning_rate": 9.16215881259163e-05}
+{"ts": "2025-12-24T07:51:03", "event": "train_log", "step": 3508, "epoch": 1.5134843581445523, "progress_pct": 25.22, "epoch_pct": 25.22, "eta": "51:46:00", "max_grad_norm": 0.8, "loss": 0.6994143724441528, "grad_norm": 1.1875312328338623, "learning_rate": 9.160797745665199e-05}
+{"ts": "2025-12-24T07:51:22", "event": "train_log", "step": 3510, "epoch": 1.5143473570658037, "progress_pct": 25.24, "epoch_pct": 25.24, "eta": "51:44:35", "max_grad_norm": 0.8, "loss": 0.7001288533210754, "grad_norm": 1.301967740058899, "learning_rate": 9.159435675381147e-05}
+{"ts": "2025-12-24T07:51:42", "event": "train_log", "step": 3512, "epoch": 1.5152103559870551, "progress_pct": 25.25, "epoch_pct": 25.25, "eta": "51:43:12", "max_grad_norm": 0.8, "loss": 0.6847118139266968, "grad_norm": 1.1110293865203857, "learning_rate": 9.158072602067933e-05}
+{"ts": "2025-12-24T07:52:03", "event": "train_log", "step": 3514, "epoch": 1.5160733549083063, "progress_pct": 25.27, "epoch_pct": 25.27, "eta": "51:41:52", "max_grad_norm": 0.8, "loss": 0.711972713470459, "grad_norm": 1.2346487045288086, "learning_rate": 9.156708526054256e-05}
+{"ts": "2025-12-24T07:52:23", "event": "train_log", "step": 3516, "epoch": 1.5169363538295577, "progress_pct": 25.28, "epoch_pct": 25.28, "eta": "51:40:30", "max_grad_norm": 0.8, "loss": 0.6406894326210022, "grad_norm": 1.079206109046936, "learning_rate": 9.155343447669058e-05}
+{"ts": "2025-12-24T07:52:45", "event": "train_log", "step": 3518, "epoch": 1.5177993527508091, "progress_pct": 25.29, "epoch_pct": 25.3, "eta": "51:39:11", "max_grad_norm": 0.8, "loss": 0.6568963527679443, "grad_norm": 1.0806539058685303, "learning_rate": 9.153977367241519e-05}
+{"ts": "2025-12-24T07:53:04", "event": "train_log", "step": 3520, "epoch": 1.5186623516720603, "progress_pct": 25.31, "epoch_pct": 25.31, "eta": "51:37:47", "max_grad_norm": 0.8, "loss": 0.6841240525245667, "grad_norm": 1.244635820388794, "learning_rate": 9.152610285101064e-05}
+{"ts": "2025-12-24T07:53:24", "event": "train_log", "step": 3522, "epoch": 1.5195253505933117, "progress_pct": 25.32, "epoch_pct": 25.33, "eta": "51:36:24", "max_grad_norm": 0.8, "loss": 0.619236946105957, "grad_norm": 1.1596359014511108, "learning_rate": 9.151242201577359e-05}
+{"ts": "2025-12-24T07:53:44", "event": "train_log", "step": 3524, "epoch": 1.5203883495145631, "progress_pct": 25.34, "epoch_pct": 25.34, "eta": "51:35:03", "max_grad_norm": 0.8, "loss": 0.6059816479682922, "grad_norm": 1.2404531240463257, "learning_rate": 9.14987311700031e-05}
+{"ts": "2025-12-24T07:54:04", "event": "train_log", "step": 3526, "epoch": 1.5212513484358143, "progress_pct": 25.35, "epoch_pct": 25.35, "eta": "51:33:41", "max_grad_norm": 0.8, "loss": 0.606432318687439, "grad_norm": 1.0919517278671265, "learning_rate": 9.148503031700067e-05}
+{"ts": "2025-12-24T07:54:22", "event": "train_log", "step": 3528, "epoch": 1.522114347357066, "progress_pct": 25.37, "epoch_pct": 25.37, "eta": "51:32:13", "max_grad_norm": 0.8, "loss": 0.6507619023323059, "grad_norm": 1.113590121269226, "learning_rate": 9.147131946007019e-05}
+{"ts": "2025-12-24T07:54:40", "event": "train_log", "step": 3530, "epoch": 1.5229773462783172, "progress_pct": 25.38, "epoch_pct": 25.38, "eta": "51:30:46", "max_grad_norm": 0.8, "loss": 0.7071492075920105, "grad_norm": 1.309885025024414, "learning_rate": 9.145759860251799e-05}
+{"ts": "2025-12-24T07:55:01", "event": "train_log", "step": 3532, "epoch": 1.5238403451995683, "progress_pct": 25.4, "epoch_pct": 25.4, "eta": "51:29:24", "max_grad_norm": 0.8, "loss": 0.6297637224197388, "grad_norm": 1.1807539463043213, "learning_rate": 9.144386774765278e-05}
+{"ts": "2025-12-24T07:55:19", "event": "train_log", "step": 3534, "epoch": 1.52470334412082, "progress_pct": 25.41, "epoch_pct": 25.41, "eta": "51:27:58", "max_grad_norm": 0.8, "loss": 0.6588859558105469, "grad_norm": 1.1587409973144531, "learning_rate": 9.143012689878568e-05}
+{"ts": "2025-12-24T07:55:39", "event": "train_log", "step": 3536, "epoch": 1.5255663430420712, "progress_pct": 25.42, "epoch_pct": 25.43, "eta": "51:26:35", "max_grad_norm": 0.8, "loss": 0.6460821628570557, "grad_norm": 1.2401988506317139, "learning_rate": 9.141637605923026e-05}
+{"ts": "2025-12-24T07:55:58", "event": "train_log", "step": 3538, "epoch": 1.5264293419633226, "progress_pct": 25.44, "epoch_pct": 25.44, "eta": "51:25:12", "max_grad_norm": 0.8, "loss": 0.6757200956344604, "grad_norm": 1.0708749294281006, "learning_rate": 9.140261523230244e-05}
+{"ts": "2025-12-24T07:56:17", "event": "train_log", "step": 3540, "epoch": 1.527292340884574, "progress_pct": 25.45, "epoch_pct": 25.45, "eta": "51:23:47", "max_grad_norm": 0.8, "loss": 0.63701993227005, "grad_norm": 1.2433289289474487, "learning_rate": 9.138884442132061e-05}
+{"ts": "2025-12-24T07:56:35", "event": "train_log", "step": 3542, "epoch": 1.5281553398058252, "progress_pct": 25.47, "epoch_pct": 25.47, "eta": "51:22:19", "max_grad_norm": 0.8, "loss": 0.6708084344863892, "grad_norm": 1.376205563545227, "learning_rate": 9.137506362960553e-05}
+{"ts": "2025-12-24T07:56:55", "event": "train_log", "step": 3544, "epoch": 1.5290183387270766, "progress_pct": 25.48, "epoch_pct": 25.48, "eta": "51:20:57", "max_grad_norm": 0.8, "loss": 0.6108639240264893, "grad_norm": 1.1187406778335571, "learning_rate": 9.136127286048038e-05}
+{"ts": "2025-12-24T07:57:16", "event": "train_log", "step": 3546, "epoch": 1.529881337648328, "progress_pct": 25.5, "epoch_pct": 25.5, "eta": "51:19:38", "max_grad_norm": 0.8, "loss": 0.58659827709198, "grad_norm": 1.0509378910064697, "learning_rate": 9.134747211727074e-05}
+{"ts": "2025-12-24T07:57:34", "event": "train_log", "step": 3548, "epoch": 1.5307443365695792, "progress_pct": 25.51, "epoch_pct": 25.51, "eta": "51:18:11", "max_grad_norm": 0.8, "loss": 0.6567258238792419, "grad_norm": 1.1126946210861206, "learning_rate": 9.133366140330458e-05}
+{"ts": "2025-12-24T07:57:53", "event": "train_log", "step": 3550, "epoch": 1.5316073354908306, "progress_pct": 25.52, "epoch_pct": 25.53, "eta": "51:16:47", "max_grad_norm": 0.8, "loss": 0.7188405990600586, "grad_norm": 1.235511302947998, "learning_rate": 9.131984072191232e-05}
+{"ts": "2025-12-24T07:58:13", "event": "train_log", "step": 3552, "epoch": 1.532470334412082, "progress_pct": 25.54, "epoch_pct": 25.54, "eta": "51:15:26", "max_grad_norm": 0.8, "loss": 0.7415963411331177, "grad_norm": 1.1522574424743652, "learning_rate": 9.130601007642674e-05}
+{"ts": "2025-12-24T07:58:33", "event": "train_log", "step": 3554, "epoch": 1.5333333333333332, "progress_pct": 25.55, "epoch_pct": 25.56, "eta": "51:14:06", "max_grad_norm": 0.8, "loss": 0.6286011934280396, "grad_norm": 1.1485437154769897, "learning_rate": 9.129216947018302e-05}
+{"ts": "2025-12-24T07:58:52", "event": "train_log", "step": 3556, "epoch": 1.5341963322545846, "progress_pct": 25.57, "epoch_pct": 25.57, "eta": "51:12:41", "max_grad_norm": 0.8, "loss": 0.6919670701026917, "grad_norm": 1.1583795547485352, "learning_rate": 9.127831890651879e-05}
+{"ts": "2025-12-24T07:59:11", "event": "train_log", "step": 3558, "epoch": 1.535059331175836, "progress_pct": 25.58, "epoch_pct": 25.58, "eta": "51:11:17", "max_grad_norm": 0.8, "loss": 0.6787450313568115, "grad_norm": 1.1450461149215698, "learning_rate": 9.126445838877404e-05}
+{"ts": "2025-12-24T07:59:31", "event": "train_log", "step": 3560, "epoch": 1.5359223300970872, "progress_pct": 25.6, "epoch_pct": 25.6, "eta": "51:09:56", "max_grad_norm": 0.8, "loss": 0.6718940138816833, "grad_norm": 1.0954205989837646, "learning_rate": 9.125058792029118e-05}
+{"ts": "2025-12-24T07:59:50", "event": "train_log", "step": 3562, "epoch": 1.5367853290183389, "progress_pct": 25.61, "epoch_pct": 25.61, "eta": "51:08:32", "max_grad_norm": 0.8, "loss": 0.7359552383422852, "grad_norm": 1.2366043329238892, "learning_rate": 9.1236707504415e-05}
+{"ts": "2025-12-24T08:00:08", "event": "train_log", "step": 3564, "epoch": 1.53764832793959, "progress_pct": 25.63, "epoch_pct": 25.63, "eta": "51:07:06", "max_grad_norm": 0.8, "loss": 0.654550313949585, "grad_norm": 1.1417096853256226, "learning_rate": 9.122281714449268e-05}
+{"ts": "2025-12-24T08:00:26", "event": "train_log", "step": 3566, "epoch": 1.5385113268608415, "progress_pct": 25.64, "epoch_pct": 25.64, "eta": "51:05:40", "max_grad_norm": 0.8, "loss": 0.7129375338554382, "grad_norm": 1.006664752960205, "learning_rate": 9.120891684387386e-05}
+{"ts": "2025-12-24T08:00:44", "event": "train_log", "step": 3568, "epoch": 1.5393743257820929, "progress_pct": 25.65, "epoch_pct": 25.66, "eta": "51:04:13", "max_grad_norm": 0.8, "loss": 0.728861391544342, "grad_norm": 1.2887771129608154, "learning_rate": 9.11950066059105e-05}
+{"ts": "2025-12-24T08:01:03", "event": "train_log", "step": 3570, "epoch": 1.540237324703344, "progress_pct": 25.67, "epoch_pct": 25.67, "eta": "51:02:50", "max_grad_norm": 0.8, "loss": 0.6207047700881958, "grad_norm": 1.194985270500183, "learning_rate": 9.118108643395702e-05}
+{"ts": "2025-12-24T08:01:24", "event": "train_log", "step": 3572, "epoch": 1.5411003236245955, "progress_pct": 25.68, "epoch_pct": 25.69, "eta": "51:01:30", "max_grad_norm": 0.8, "loss": 0.6507744789123535, "grad_norm": 1.1913846731185913, "learning_rate": 9.116715633137019e-05}
+{"ts": "2025-12-24T08:01:43", "event": "train_log", "step": 3574, "epoch": 1.541963322545847, "progress_pct": 25.7, "epoch_pct": 25.7, "eta": "51:00:07", "max_grad_norm": 0.8, "loss": 0.6736630201339722, "grad_norm": 1.1695647239685059, "learning_rate": 9.115321630150919e-05}
+{"ts": "2025-12-24T08:02:01", "event": "train_log", "step": 3576, "epoch": 1.542826321467098, "progress_pct": 25.71, "epoch_pct": 25.71, "eta": "50:58:43", "max_grad_norm": 0.8, "loss": 0.6998549699783325, "grad_norm": 1.2866519689559937, "learning_rate": 9.113926634773558e-05}
+{"ts": "2025-12-24T08:02:20", "event": "train_log", "step": 3578, "epoch": 1.5436893203883495, "progress_pct": 25.73, "epoch_pct": 25.73, "eta": "50:57:18", "max_grad_norm": 0.8, "loss": 0.6986453533172607, "grad_norm": 1.1755015850067139, "learning_rate": 9.112530647341337e-05}
+{"ts": "2025-12-24T08:02:39", "event": "train_log", "step": 3580, "epoch": 1.544552319309601, "progress_pct": 25.74, "epoch_pct": 25.74, "eta": "50:55:56", "max_grad_norm": 0.8, "loss": 0.6959735155105591, "grad_norm": 1.2085564136505127, "learning_rate": 9.111133668190889e-05}
+{"ts": "2025-12-24T08:02:57", "event": "train_log", "step": 3582, "epoch": 1.545415318230852, "progress_pct": 25.75, "epoch_pct": 25.76, "eta": "50:54:30", "max_grad_norm": 0.8, "loss": 0.6336050033569336, "grad_norm": 1.2213492393493652, "learning_rate": 9.109735697659089e-05}
+{"ts": "2025-12-24T08:03:16", "event": "train_log", "step": 3584, "epoch": 1.5462783171521035, "progress_pct": 25.77, "epoch_pct": 25.77, "eta": "50:53:05", "max_grad_norm": 0.8, "loss": 0.652992844581604, "grad_norm": 1.1609338521957397, "learning_rate": 9.108336736083054e-05}
+{"ts": "2025-12-24T08:03:35", "event": "train_log", "step": 3586, "epoch": 1.547141316073355, "progress_pct": 25.78, "epoch_pct": 25.79, "eta": "50:51:42", "max_grad_norm": 0.8, "loss": 0.5593333840370178, "grad_norm": 1.1483949422836304, "learning_rate": 9.106936783800136e-05}
+{"ts": "2025-12-24T08:03:53", "event": "train_log", "step": 3588, "epoch": 1.5480043149946061, "progress_pct": 25.8, "epoch_pct": 25.8, "eta": "50:50:19", "max_grad_norm": 0.8, "loss": 0.6488125920295715, "grad_norm": 1.0852224826812744, "learning_rate": 9.105535841147925e-05}
+{"ts": "2025-12-24T08:04:13", "event": "train_log", "step": 3590, "epoch": 1.5488673139158577, "progress_pct": 25.81, "epoch_pct": 25.81, "eta": "50:48:59", "max_grad_norm": 0.8, "loss": 0.5877764225006104, "grad_norm": 1.1804794073104858, "learning_rate": 9.104133908464254e-05}
+{"ts": "2025-12-24T08:04:33", "event": "train_log", "step": 3592, "epoch": 1.549730312837109, "progress_pct": 25.83, "epoch_pct": 25.83, "eta": "50:47:37", "max_grad_norm": 0.8, "loss": 0.5924050211906433, "grad_norm": 1.0295292139053345, "learning_rate": 9.102730986087194e-05}
+{"ts": "2025-12-24T08:04:53", "event": "train_log", "step": 3594, "epoch": 1.5505933117583603, "progress_pct": 25.84, "epoch_pct": 25.84, "eta": "50:46:17", "max_grad_norm": 0.8, "loss": 0.6224870085716248, "grad_norm": 1.524871587753296, "learning_rate": 9.101327074355051e-05}
+{"ts": "2025-12-24T08:05:11", "event": "train_log", "step": 3596, "epoch": 1.5514563106796118, "progress_pct": 25.86, "epoch_pct": 25.86, "eta": "50:44:54", "max_grad_norm": 0.8, "loss": 0.5918852090835571, "grad_norm": 1.3207813501358032, "learning_rate": 9.099922173606373e-05}
+{"ts": "2025-12-24T08:05:30", "event": "train_log", "step": 3598, "epoch": 1.552319309600863, "progress_pct": 25.87, "epoch_pct": 25.87, "eta": "50:43:29", "max_grad_norm": 0.8, "loss": 0.6756706237792969, "grad_norm": 1.2524348497390747, "learning_rate": 9.098516284179945e-05}
+{"ts": "2025-12-24T08:05:48", "event": "train_log", "step": 3600, "epoch": 1.5531823085221144, "progress_pct": 25.88, "epoch_pct": 25.89, "eta": "50:42:05", "max_grad_norm": 0.8, "loss": 0.7416443824768066, "grad_norm": 1.2434089183807373, "learning_rate": 9.097109406414792e-05}
+{"ts": "2025-12-24T08:19:44", "event": "train_log", "step": 3600, "epoch": 1.5531823085221144, "progress_pct": 25.88, "epoch_pct": 25.89, "eta": "51:21:59", "max_grad_norm": 0.8, "eval_loss": 0.701831042766571, "eval_runtime": 836.0199, "eval_samples_per_second": 2.464, "eval_steps_per_second": 2.464}
+{"ts": "2025-12-24T08:20:05", "event": "train_log", "step": 3602, "epoch": 1.5540453074433658, "progress_pct": 25.9, "epoch_pct": 25.9, "eta": "51:20:39", "max_grad_norm": 0.8, "loss": 0.5911263227462769, "grad_norm": 1.0528122186660767, "learning_rate": 9.095701540650175e-05}
+{"ts": "2025-12-24T08:20:25", "event": "train_log", "step": 3604, "epoch": 1.554908306364617, "progress_pct": 25.91, "epoch_pct": 25.92, "eta": "51:19:20", "max_grad_norm": 0.8, "loss": 0.5925720930099487, "grad_norm": 1.1165716648101807, "learning_rate": 9.094292687225595e-05}
+{"ts": "2025-12-24T08:20:46", "event": "train_log", "step": 3606, "epoch": 1.5557713052858684, "progress_pct": 25.93, "epoch_pct": 25.93, "eta": "51:18:01", "max_grad_norm": 0.8, "loss": 0.6038441061973572, "grad_norm": 1.1127078533172607, "learning_rate": 9.09288284648079e-05}
+{"ts": "2025-12-24T08:21:04", "event": "train_log", "step": 3608, "epoch": 1.5566343042071198, "progress_pct": 25.94, "epoch_pct": 25.94, "eta": "51:16:34", "max_grad_norm": 0.8, "loss": 0.6781054735183716, "grad_norm": 1.195662498474121, "learning_rate": 9.091472018755738e-05}
+{"ts": "2025-12-24T08:21:25", "event": "train_log", "step": 3610, "epoch": 1.557497303128371, "progress_pct": 25.96, "epoch_pct": 25.96, "eta": "51:15:15", "max_grad_norm": 0.8, "loss": 0.7171973586082458, "grad_norm": 1.1671538352966309, "learning_rate": 9.090060204390652e-05}
+{"ts": "2025-12-24T08:21:45", "event": "train_log", "step": 3612, "epoch": 1.5583603020496224, "progress_pct": 25.97, "epoch_pct": 25.97, "eta": "51:13:55", "max_grad_norm": 0.8, "loss": 0.6465495228767395, "grad_norm": 1.2814654111862183, "learning_rate": 9.088647403725987e-05}
+{"ts": "2025-12-24T08:22:04", "event": "train_log", "step": 3614, "epoch": 1.5592233009708738, "progress_pct": 25.99, "epoch_pct": 25.99, "eta": "51:12:30", "max_grad_norm": 0.8, "loss": 0.7105259299278259, "grad_norm": 1.2238248586654663, "learning_rate": 9.087233617102433e-05}
+{"ts": "2025-12-24T08:22:24", "event": "train_log", "step": 3616, "epoch": 1.560086299892125, "progress_pct": 26.0, "epoch_pct": 26.0, "eta": "51:11:09", "max_grad_norm": 0.8, "loss": 0.6721329092979431, "grad_norm": 1.1868736743927002, "learning_rate": 9.085818844860916e-05}
+{"ts": "2025-12-24T08:22:45", "event": "train_log", "step": 3618, "epoch": 1.5609492988133766, "progress_pct": 26.01, "epoch_pct": 26.02, "eta": "51:09:53", "max_grad_norm": 0.8, "loss": 0.6684948801994324, "grad_norm": 1.1358642578125, "learning_rate": 9.084403087342606e-05}
+{"ts": "2025-12-24T08:23:04", "event": "train_log", "step": 3620, "epoch": 1.5618122977346278, "progress_pct": 26.03, "epoch_pct": 26.03, "eta": "51:08:29", "max_grad_norm": 0.8, "loss": 0.653505802154541, "grad_norm": 1.1101531982421875, "learning_rate": 9.082986344888901e-05}
+{"ts": "2025-12-24T08:23:23", "event": "train_log", "step": 3622, "epoch": 1.5626752966558792, "progress_pct": 26.04, "epoch_pct": 26.04, "eta": "51:07:04", "max_grad_norm": 0.8, "loss": 0.673059344291687, "grad_norm": 1.3343961238861084, "learning_rate": 9.081568617841447e-05}
+{"ts": "2025-12-24T08:23:42", "event": "train_log", "step": 3624, "epoch": 1.5635382955771306, "progress_pct": 26.06, "epoch_pct": 26.06, "eta": "51:05:41", "max_grad_norm": 0.8, "loss": 0.6429969668388367, "grad_norm": 1.0680615901947021, "learning_rate": 9.080149906542119e-05}
+{"ts": "2025-12-24T08:24:02", "event": "train_log", "step": 3626, "epoch": 1.5644012944983818, "progress_pct": 26.07, "epoch_pct": 26.07, "eta": "51:04:20", "max_grad_norm": 0.8, "loss": 0.694593608379364, "grad_norm": 1.0026768445968628, "learning_rate": 9.078730211333038e-05}
+{"ts": "2025-12-24T08:24:21", "event": "train_log", "step": 3628, "epoch": 1.5652642934196332, "progress_pct": 26.09, "epoch_pct": 26.09, "eta": "51:02:57", "max_grad_norm": 0.8, "loss": 0.6931326985359192, "grad_norm": 1.100269079208374, "learning_rate": 9.077309532556551e-05}
+{"ts": "2025-12-24T08:24:41", "event": "train_log", "step": 3630, "epoch": 1.5661272923408847, "progress_pct": 26.1, "epoch_pct": 26.1, "eta": "51:01:36", "max_grad_norm": 0.8, "loss": 0.6182643175125122, "grad_norm": 1.1724740266799927, "learning_rate": 9.07588787055525e-05}
+{"ts": "2025-12-24T08:25:00", "event": "train_log", "step": 3632, "epoch": 1.5669902912621358, "progress_pct": 26.11, "epoch_pct": 26.12, "eta": "51:00:15", "max_grad_norm": 0.8, "loss": 0.6307282447814941, "grad_norm": 1.0137040615081787, "learning_rate": 9.074465225671965e-05}
+{"ts": "2025-12-24T08:25:20", "event": "train_log", "step": 3634, "epoch": 1.5678532901833873, "progress_pct": 26.13, "epoch_pct": 26.13, "eta": "50:58:53", "max_grad_norm": 0.8, "loss": 0.6049362421035767, "grad_norm": 1.0997183322906494, "learning_rate": 9.073041598249756e-05}
+{"ts": "2025-12-24T08:25:39", "event": "train_log", "step": 3636, "epoch": 1.5687162891046387, "progress_pct": 26.14, "epoch_pct": 26.15, "eta": "50:57:31", "max_grad_norm": 0.8, "loss": 0.6583153009414673, "grad_norm": 1.1688816547393799, "learning_rate": 9.07161698863193e-05}
+{"ts": "2025-12-24T08:25:58", "event": "train_log", "step": 3638, "epoch": 1.5695792880258899, "progress_pct": 26.16, "epoch_pct": 26.16, "eta": "50:56:08", "max_grad_norm": 0.8, "loss": 0.6346786022186279, "grad_norm": 1.2558770179748535, "learning_rate": 9.07019139716202e-05}
+{"ts": "2025-12-24T08:26:17", "event": "train_log", "step": 3640, "epoch": 1.5704422869471413, "progress_pct": 26.17, "epoch_pct": 26.17, "eta": "50:54:46", "max_grad_norm": 0.8, "loss": 0.6214386224746704, "grad_norm": 1.037838101387024, "learning_rate": 9.068764824183804e-05}
+{"ts": "2025-12-24T08:26:36", "event": "train_log", "step": 3642, "epoch": 1.5713052858683927, "progress_pct": 26.19, "epoch_pct": 26.19, "eta": "50:53:23", "max_grad_norm": 0.8, "loss": 0.6745668649673462, "grad_norm": 1.2514159679412842, "learning_rate": 9.067337270041293e-05}
+{"ts": "2025-12-24T08:26:54", "event": "train_log", "step": 3644, "epoch": 1.5721682847896439, "progress_pct": 26.2, "epoch_pct": 26.2, "eta": "50:51:57", "max_grad_norm": 0.8, "loss": 0.6409499049186707, "grad_norm": 1.2718034982681274, "learning_rate": 9.065908735078736e-05}
+{"ts": "2025-12-24T08:27:13", "event": "train_log", "step": 3646, "epoch": 1.5730312837108955, "progress_pct": 26.22, "epoch_pct": 26.22, "eta": "50:50:36", "max_grad_norm": 0.8, "loss": 0.5885956883430481, "grad_norm": 1.1754566431045532, "learning_rate": 9.064479219640615e-05}
+{"ts": "2025-12-24T08:27:31", "event": "train_log", "step": 3648, "epoch": 1.5738942826321467, "progress_pct": 26.23, "epoch_pct": 26.23, "eta": "50:49:10", "max_grad_norm": 0.8, "loss": 0.6472923159599304, "grad_norm": 1.3022228479385376, "learning_rate": 9.063048724071653e-05}
+{"ts": "2025-12-24T08:27:50", "event": "train_log", "step": 3650, "epoch": 1.574757281553398, "progress_pct": 26.24, "epoch_pct": 26.25, "eta": "50:47:46", "max_grad_norm": 0.8, "loss": 0.6977139711380005, "grad_norm": 1.2492727041244507, "learning_rate": 9.061617248716808e-05}
+{"ts": "2025-12-24T08:28:10", "event": "train_log", "step": 3652, "epoch": 1.5756202804746495, "progress_pct": 26.26, "epoch_pct": 26.26, "eta": "50:46:26", "max_grad_norm": 0.8, "loss": 0.6403328776359558, "grad_norm": 1.0432982444763184, "learning_rate": 9.060184793921274e-05}
+{"ts": "2025-12-24T08:28:29", "event": "train_log", "step": 3654, "epoch": 1.5764832793959007, "progress_pct": 26.27, "epoch_pct": 26.27, "eta": "50:45:04", "max_grad_norm": 0.8, "loss": 0.6726073026657104, "grad_norm": 1.0898420810699463, "learning_rate": 9.05875136003048e-05}
+{"ts": "2025-12-24T08:28:48", "event": "train_log", "step": 3656, "epoch": 1.5773462783171521, "progress_pct": 26.29, "epoch_pct": 26.29, "eta": "50:43:41", "max_grad_norm": 0.8, "loss": 0.6861021518707275, "grad_norm": 1.1885254383087158, "learning_rate": 9.057316947390091e-05}
+{"ts": "2025-12-24T08:29:07", "event": "train_log", "step": 3658, "epoch": 1.5782092772384035, "progress_pct": 26.3, "epoch_pct": 26.3, "eta": "50:42:21", "max_grad_norm": 0.8, "loss": 0.6786487102508545, "grad_norm": 1.0795915126800537, "learning_rate": 9.055881556346012e-05}
+{"ts": "2025-12-24T08:29:27", "event": "train_log", "step": 3660, "epoch": 1.5790722761596547, "progress_pct": 26.32, "epoch_pct": 26.32, "eta": "50:41:01", "max_grad_norm": 0.8, "loss": 0.6757274866104126, "grad_norm": 1.0702277421951294, "learning_rate": 9.054445187244379e-05}
+{"ts": "2025-12-24T08:29:46", "event": "train_log", "step": 3662, "epoch": 1.5799352750809061, "progress_pct": 26.33, "epoch_pct": 26.33, "eta": "50:39:39", "max_grad_norm": 0.8, "loss": 0.6596087217330933, "grad_norm": 0.9469023942947388, "learning_rate": 9.053007840431566e-05}
+{"ts": "2025-12-24T08:30:06", "event": "train_log", "step": 3664, "epoch": 1.5807982740021576, "progress_pct": 26.34, "epoch_pct": 26.35, "eta": "50:38:19", "max_grad_norm": 0.8, "loss": 0.6254066228866577, "grad_norm": 0.979669988155365, "learning_rate": 9.051569516254185e-05}
+{"ts": "2025-12-24T08:30:24", "event": "train_log", "step": 3666, "epoch": 1.5816612729234087, "progress_pct": 26.36, "epoch_pct": 26.36, "eta": "50:36:56", "max_grad_norm": 0.8, "loss": 0.6463974118232727, "grad_norm": 1.294577717781067, "learning_rate": 9.050130215059081e-05}
+{"ts": "2025-12-24T08:30:44", "event": "train_log", "step": 3668, "epoch": 1.5825242718446602, "progress_pct": 26.37, "epoch_pct": 26.38, "eta": "50:35:36", "max_grad_norm": 0.8, "loss": 0.6496619582176208, "grad_norm": 1.1566381454467773, "learning_rate": 9.04868993719333e-05}
+{"ts": "2025-12-24T08:31:03", "event": "train_log", "step": 3670, "epoch": 1.5833872707659116, "progress_pct": 26.39, "epoch_pct": 26.39, "eta": "50:34:15", "max_grad_norm": 0.8, "loss": 0.5987569093704224, "grad_norm": 1.0207090377807617, "learning_rate": 9.047248683004254e-05}
+{"ts": "2025-12-24T08:31:22", "event": "train_log", "step": 3672, "epoch": 1.5842502696871628, "progress_pct": 26.4, "epoch_pct": 26.4, "eta": "50:32:53", "max_grad_norm": 0.8, "loss": 0.6392318606376648, "grad_norm": 1.1533408164978027, "learning_rate": 9.045806452839404e-05}
+{"ts": "2025-12-24T08:31:42", "event": "train_log", "step": 3674, "epoch": 1.5851132686084144, "progress_pct": 26.42, "epoch_pct": 26.42, "eta": "50:31:33", "max_grad_norm": 0.8, "loss": 0.6197013854980469, "grad_norm": 1.0433615446090698, "learning_rate": 9.044363247046565e-05}
+{"ts": "2025-12-24T08:32:01", "event": "train_log", "step": 3676, "epoch": 1.5859762675296656, "progress_pct": 26.43, "epoch_pct": 26.43, "eta": "50:30:13", "max_grad_norm": 0.8, "loss": 0.620261549949646, "grad_norm": 1.2047847509384155, "learning_rate": 9.042919065973763e-05}
+{"ts": "2025-12-24T08:32:20", "event": "train_log", "step": 3678, "epoch": 1.5868392664509168, "progress_pct": 26.45, "epoch_pct": 26.45, "eta": "50:28:49", "max_grad_norm": 0.8, "loss": 0.6751506924629211, "grad_norm": 1.232506513595581, "learning_rate": 9.041473909969252e-05}
+{"ts": "2025-12-24T08:32:38", "event": "train_log", "step": 3680, "epoch": 1.5877022653721684, "progress_pct": 26.46, "epoch_pct": 26.46, "eta": "50:27:26", "max_grad_norm": 0.8, "loss": 0.6619004011154175, "grad_norm": 1.1659306287765503, "learning_rate": 9.040027779381528e-05}
+{"ts": "2025-12-24T08:32:57", "event": "train_log", "step": 3682, "epoch": 1.5885652642934196, "progress_pct": 26.47, "epoch_pct": 26.48, "eta": "50:26:05", "max_grad_norm": 0.8, "loss": 0.6120786666870117, "grad_norm": 1.0183755159378052, "learning_rate": 9.038580674559316e-05}
+{"ts": "2025-12-24T08:33:17", "event": "train_log", "step": 3684, "epoch": 1.589428263214671, "progress_pct": 26.49, "epoch_pct": 26.49, "eta": "50:24:45", "max_grad_norm": 0.8, "loss": 0.616247296333313, "grad_norm": 1.194318175315857, "learning_rate": 9.037132595851581e-05}
+{"ts": "2025-12-24T08:33:35", "event": "train_log", "step": 3686, "epoch": 1.5902912621359224, "progress_pct": 26.5, "epoch_pct": 26.5, "eta": "50:23:23", "max_grad_norm": 0.8, "loss": 0.7229626774787903, "grad_norm": 1.500475525856018, "learning_rate": 9.035683543607522e-05}
+{"ts": "2025-12-24T08:33:55", "event": "train_log", "step": 3688, "epoch": 1.5911542610571736, "progress_pct": 26.52, "epoch_pct": 26.52, "eta": "50:22:04", "max_grad_norm": 0.8, "loss": 0.6193928718566895, "grad_norm": 1.047954797744751, "learning_rate": 9.034233518176567e-05}
+{"ts": "2025-12-24T08:34:13", "event": "train_log", "step": 3690, "epoch": 1.592017259978425, "progress_pct": 26.53, "epoch_pct": 26.53, "eta": "50:20:40", "max_grad_norm": 0.8, "loss": 0.7276740074157715, "grad_norm": 1.3852946758270264, "learning_rate": 9.032782519908386e-05}
+{"ts": "2025-12-24T08:34:32", "event": "train_log", "step": 3692, "epoch": 1.5928802588996764, "progress_pct": 26.55, "epoch_pct": 26.55, "eta": "50:19:18", "max_grad_norm": 0.8, "loss": 0.6635852456092834, "grad_norm": 1.2648470401763916, "learning_rate": 9.031330549152882e-05}
+{"ts": "2025-12-24T08:34:51", "event": "train_log", "step": 3694, "epoch": 1.5937432578209276, "progress_pct": 26.56, "epoch_pct": 26.56, "eta": "50:17:57", "max_grad_norm": 0.8, "loss": 0.6905328631401062, "grad_norm": 1.1195251941680908, "learning_rate": 9.029877606260186e-05}
+{"ts": "2025-12-24T08:35:11", "event": "train_log", "step": 3696, "epoch": 1.594606256742179, "progress_pct": 26.57, "epoch_pct": 26.58, "eta": "50:16:40", "max_grad_norm": 0.8, "loss": 0.6474795341491699, "grad_norm": 1.0823792219161987, "learning_rate": 9.028423691580675e-05}
+{"ts": "2025-12-24T08:35:30", "event": "train_log", "step": 3698, "epoch": 1.5954692556634305, "progress_pct": 26.59, "epoch_pct": 26.59, "eta": "50:15:18", "max_grad_norm": 0.8, "loss": 0.6740748882293701, "grad_norm": 1.1743394136428833, "learning_rate": 9.026968805464948e-05}
+{"ts": "2025-12-24T08:35:50", "event": "train_log", "step": 3700, "epoch": 1.5963322545846816, "progress_pct": 26.6, "epoch_pct": 26.61, "eta": "50:13:59", "max_grad_norm": 0.8, "loss": 0.5662298798561096, "grad_norm": 1.19667387008667, "learning_rate": 9.02551294826385e-05}
+{"ts": "2025-12-24T08:49:46", "event": "train_log", "step": 3700, "epoch": 1.5963322545846816, "progress_pct": 26.6, "epoch_pct": 26.61, "eta": "50:52:27", "max_grad_norm": 0.8, "eval_loss": 0.6999035477638245, "eval_runtime": 836.4944, "eval_samples_per_second": 2.463, "eval_steps_per_second": 2.463}
+{"ts": "2025-12-24T08:50:06", "event": "train_log", "step": 3702, "epoch": 1.5971952535059333, "progress_pct": 26.62, "epoch_pct": 26.62, "eta": "50:51:07", "max_grad_norm": 0.8, "loss": 0.6141433119773865, "grad_norm": 0.9947987794876099, "learning_rate": 9.02405612032845e-05}
+{"ts": "2025-12-24T08:50:25", "event": "train_log", "step": 3704, "epoch": 1.5980582524271845, "progress_pct": 26.63, "epoch_pct": 26.63, "eta": "50:49:45", "max_grad_norm": 0.8, "loss": 0.6011156439781189, "grad_norm": 1.0307772159576416, "learning_rate": 9.022598322010059e-05}
+{"ts": "2025-12-24T08:50:44", "event": "train_log", "step": 3706, "epoch": 1.5989212513484357, "progress_pct": 26.65, "epoch_pct": 26.65, "eta": "50:48:22", "max_grad_norm": 0.8, "loss": 0.7671760320663452, "grad_norm": 1.3516873121261597, "learning_rate": 9.021139553660214e-05}
+{"ts": "2025-12-24T08:51:04", "event": "train_log", "step": 3708, "epoch": 1.5997842502696873, "progress_pct": 26.66, "epoch_pct": 26.66, "eta": "50:47:03", "max_grad_norm": 0.8, "loss": 0.5630212426185608, "grad_norm": 1.0319154262542725, "learning_rate": 9.019679815630695e-05}
+{"ts": "2025-12-24T08:51:22", "event": "train_log", "step": 3710, "epoch": 1.6006472491909385, "progress_pct": 26.68, "epoch_pct": 26.68, "eta": "50:45:37", "max_grad_norm": 0.8, "loss": 0.7143195271492004, "grad_norm": 1.34807550907135, "learning_rate": 9.01821910827351e-05}
+{"ts": "2025-12-24T08:51:41", "event": "train_log", "step": 3712, "epoch": 1.60151024811219, "progress_pct": 26.69, "epoch_pct": 26.69, "eta": "50:44:17", "max_grad_norm": 0.8, "loss": 0.6624904274940491, "grad_norm": 1.0618840456008911, "learning_rate": 9.016757431940898e-05}
+{"ts": "2025-12-24T08:52:00", "event": "train_log", "step": 3714, "epoch": 1.6023732470334413, "progress_pct": 26.7, "epoch_pct": 26.71, "eta": "50:42:53", "max_grad_norm": 0.8, "loss": 0.631399929523468, "grad_norm": 0.987445056438446, "learning_rate": 9.015294786985342e-05}
+{"ts": "2025-12-24T08:52:18", "event": "train_log", "step": 3716, "epoch": 1.6032362459546925, "progress_pct": 26.72, "epoch_pct": 26.72, "eta": "50:41:30", "max_grad_norm": 0.8, "loss": 0.646812379360199, "grad_norm": 1.0872961282730103, "learning_rate": 9.013831173759548e-05}
+{"ts": "2025-12-24T08:52:38", "event": "train_log", "step": 3718, "epoch": 1.604099244875944, "progress_pct": 26.73, "epoch_pct": 26.73, "eta": "50:40:11", "max_grad_norm": 0.8, "loss": 0.6353676915168762, "grad_norm": 0.9798976182937622, "learning_rate": 9.012366592616461e-05}
+{"ts": "2025-12-24T08:52:59", "event": "train_log", "step": 3720, "epoch": 1.6049622437971953, "progress_pct": 26.75, "epoch_pct": 26.75, "eta": "50:38:53", "max_grad_norm": 0.8, "loss": 0.5802496671676636, "grad_norm": 1.3835123777389526, "learning_rate": 9.010901043909258e-05}
+{"ts": "2025-12-24T08:53:18", "event": "train_log", "step": 3722, "epoch": 1.6058252427184465, "progress_pct": 26.76, "epoch_pct": 26.76, "eta": "50:37:32", "max_grad_norm": 0.8, "loss": 0.6140236854553223, "grad_norm": 1.0870599746704102, "learning_rate": 9.009434527991349e-05}
+{"ts": "2025-12-24T08:53:37", "event": "train_log", "step": 3724, "epoch": 1.606688241639698, "progress_pct": 26.78, "epoch_pct": 26.78, "eta": "50:36:10", "max_grad_norm": 0.8, "loss": 0.6562411785125732, "grad_norm": 1.0582176446914673, "learning_rate": 9.007967045216379e-05}
+{"ts": "2025-12-24T08:53:56", "event": "train_log", "step": 3726, "epoch": 1.6075512405609493, "progress_pct": 26.79, "epoch_pct": 26.79, "eta": "50:34:49", "max_grad_norm": 0.8, "loss": 0.5995528697967529, "grad_norm": 1.0486522912979126, "learning_rate": 9.006498595938224e-05}
+{"ts": "2025-12-24T08:54:16", "event": "train_log", "step": 3728, "epoch": 1.6084142394822005, "progress_pct": 26.8, "epoch_pct": 26.81, "eta": "50:33:29", "max_grad_norm": 0.8, "loss": 0.6296200752258301, "grad_norm": 1.1152576208114624, "learning_rate": 9.005029180510994e-05}
+{"ts": "2025-12-24T08:54:34", "event": "train_log", "step": 3730, "epoch": 1.6092772384034522, "progress_pct": 26.82, "epoch_pct": 26.82, "eta": "50:32:06", "max_grad_norm": 0.8, "loss": 0.7050777077674866, "grad_norm": 1.2114241123199463, "learning_rate": 9.003558799289031e-05}
+{"ts": "2025-12-24T08:54:54", "event": "train_log", "step": 3732, "epoch": 1.6101402373247033, "progress_pct": 26.83, "epoch_pct": 26.84, "eta": "50:30:47", "max_grad_norm": 0.8, "loss": 0.6323828101158142, "grad_norm": 1.1112302541732788, "learning_rate": 9.002087452626913e-05}
+{"ts": "2025-12-24T08:55:13", "event": "train_log", "step": 3734, "epoch": 1.6110032362459545, "progress_pct": 26.85, "epoch_pct": 26.85, "eta": "50:29:25", "max_grad_norm": 0.8, "loss": 0.7021811008453369, "grad_norm": 1.102839469909668, "learning_rate": 9.000615140879446e-05}
+{"ts": "2025-12-24T08:55:34", "event": "train_log", "step": 3736, "epoch": 1.6118662351672062, "progress_pct": 26.86, "epoch_pct": 26.86, "eta": "50:28:10", "max_grad_norm": 0.8, "loss": 0.5094663500785828, "grad_norm": 0.8860642910003662, "learning_rate": 8.999141864401672e-05}
+{"ts": "2025-12-24T08:55:52", "event": "train_log", "step": 3738, "epoch": 1.6127292340884574, "progress_pct": 26.88, "epoch_pct": 26.88, "eta": "50:26:47", "max_grad_norm": 0.8, "loss": 0.6481591463088989, "grad_norm": 1.1873432397842407, "learning_rate": 8.997667623548869e-05}
+{"ts": "2025-12-24T08:56:11", "event": "train_log", "step": 3740, "epoch": 1.6135922330097088, "progress_pct": 26.89, "epoch_pct": 26.89, "eta": "50:25:26", "max_grad_norm": 0.8, "loss": 0.6846200227737427, "grad_norm": 0.9258447885513306, "learning_rate": 8.996192418676537e-05}
+{"ts": "2025-12-24T08:56:30", "event": "train_log", "step": 3742, "epoch": 1.6144552319309602, "progress_pct": 26.91, "epoch_pct": 26.91, "eta": "50:24:04", "max_grad_norm": 0.8, "loss": 0.7081074118614197, "grad_norm": 1.1108704805374146, "learning_rate": 8.994716250140419e-05}
+{"ts": "2025-12-24T08:56:49", "event": "train_log", "step": 3744, "epoch": 1.6153182308522114, "progress_pct": 26.92, "epoch_pct": 26.92, "eta": "50:22:42", "max_grad_norm": 0.8, "loss": 0.683017373085022, "grad_norm": 1.01946222782135, "learning_rate": 8.993239118296486e-05}
+{"ts": "2025-12-24T08:57:08", "event": "train_log", "step": 3746, "epoch": 1.6161812297734628, "progress_pct": 26.93, "epoch_pct": 26.94, "eta": "50:21:22", "max_grad_norm": 0.8, "loss": 0.6625820994377136, "grad_norm": 1.0794068574905396, "learning_rate": 8.991761023500942e-05}
+{"ts": "2025-12-24T08:57:28", "event": "train_log", "step": 3748, "epoch": 1.6170442286947142, "progress_pct": 26.95, "epoch_pct": 26.95, "eta": "50:20:03", "max_grad_norm": 0.8, "loss": 0.6714970469474792, "grad_norm": 1.0697723627090454, "learning_rate": 8.990281966110221e-05}
+{"ts": "2025-12-24T08:57:48", "event": "train_log", "step": 3750, "epoch": 1.6179072276159654, "progress_pct": 26.96, "epoch_pct": 26.97, "eta": "50:18:46", "max_grad_norm": 0.8, "loss": 0.6702249646186829, "grad_norm": 1.118213176727295, "learning_rate": 8.988801946480991e-05}
+{"ts": "2025-12-24T08:58:08", "event": "train_log", "step": 3752, "epoch": 1.6187702265372168, "progress_pct": 26.98, "epoch_pct": 26.98, "eta": "50:17:27", "max_grad_norm": 0.8, "loss": 0.6783145070075989, "grad_norm": 1.3541207313537598, "learning_rate": 8.987320964970156e-05}
+{"ts": "2025-12-24T08:58:28", "event": "train_log", "step": 3754, "epoch": 1.6196332254584682, "progress_pct": 26.99, "epoch_pct": 26.99, "eta": "50:16:09", "max_grad_norm": 0.8, "loss": 0.6774528622627258, "grad_norm": 1.2015430927276611, "learning_rate": 8.985839021934843e-05}
+{"ts": "2025-12-24T08:58:46", "event": "train_log", "step": 3756, "epoch": 1.6204962243797194, "progress_pct": 27.01, "epoch_pct": 27.01, "eta": "50:14:45", "max_grad_norm": 0.8, "loss": 0.6122720241546631, "grad_norm": 1.2218681573867798, "learning_rate": 8.984356117732419e-05}
+{"ts": "2025-12-24T08:59:08", "event": "train_log", "step": 3758, "epoch": 1.6213592233009708, "progress_pct": 27.02, "epoch_pct": 27.02, "eta": "50:13:34", "max_grad_norm": 0.8, "loss": 0.6658767461776733, "grad_norm": 1.0959049463272095, "learning_rate": 8.982872252720478e-05}
+{"ts": "2025-12-24T08:59:28", "event": "train_log", "step": 3760, "epoch": 1.6222222222222222, "progress_pct": 27.03, "epoch_pct": 27.04, "eta": "50:12:16", "max_grad_norm": 0.8, "loss": 0.6943061947822571, "grad_norm": 1.0399792194366455, "learning_rate": 8.981387427256849e-05}
+{"ts": "2025-12-24T08:59:49", "event": "train_log", "step": 3762, "epoch": 1.6230852211434734, "progress_pct": 27.05, "epoch_pct": 27.05, "eta": "50:10:59", "max_grad_norm": 0.8, "loss": 0.6100694537162781, "grad_norm": 1.1107885837554932, "learning_rate": 8.979901641699591e-05}
+{"ts": "2025-12-24T09:00:07", "event": "train_log", "step": 3764, "epoch": 1.623948220064725, "progress_pct": 27.06, "epoch_pct": 27.07, "eta": "50:09:39", "max_grad_norm": 0.8, "loss": 0.7148857712745667, "grad_norm": 1.2478004693984985, "learning_rate": 8.978414896406993e-05}
+{"ts": "2025-12-24T09:00:26", "event": "train_log", "step": 3766, "epoch": 1.6248112189859762, "progress_pct": 27.08, "epoch_pct": 27.08, "eta": "50:08:19", "max_grad_norm": 0.8, "loss": 0.6626213788986206, "grad_norm": 1.333235502243042, "learning_rate": 8.976927191737577e-05}
+{"ts": "2025-12-24T09:00:46", "event": "train_log", "step": 3768, "epoch": 1.6256742179072277, "progress_pct": 27.09, "epoch_pct": 27.09, "eta": "50:06:59", "max_grad_norm": 0.8, "loss": 0.6073925495147705, "grad_norm": 1.2001140117645264, "learning_rate": 8.975438528050095e-05}
+{"ts": "2025-12-24T09:01:07", "event": "train_log", "step": 3770, "epoch": 1.626537216828479, "progress_pct": 27.11, "epoch_pct": 27.11, "eta": "50:05:46", "max_grad_norm": 0.8, "loss": 0.5861713290214539, "grad_norm": 0.8740705847740173, "learning_rate": 8.973948905703535e-05}
+{"ts": "2025-12-24T09:01:28", "event": "train_log", "step": 3772, "epoch": 1.6274002157497303, "progress_pct": 27.12, "epoch_pct": 27.12, "eta": "50:04:30", "max_grad_norm": 0.8, "loss": 0.7104416489601135, "grad_norm": 1.0697330236434937, "learning_rate": 8.972458325057111e-05}
+{"ts": "2025-12-24T09:01:47", "event": "train_log", "step": 3774, "epoch": 1.6282632146709817, "progress_pct": 27.14, "epoch_pct": 27.14, "eta": "50:03:11", "max_grad_norm": 0.8, "loss": 0.652777373790741, "grad_norm": 1.2699353694915771, "learning_rate": 8.970966786470271e-05}
+{"ts": "2025-12-24T09:02:08", "event": "train_log", "step": 3776, "epoch": 1.629126213592233, "progress_pct": 27.15, "epoch_pct": 27.15, "eta": "50:01:56", "max_grad_norm": 0.8, "loss": 0.6303225159645081, "grad_norm": 1.2045143842697144, "learning_rate": 8.969474290302691e-05}
+{"ts": "2025-12-24T09:02:28", "event": "train_log", "step": 3778, "epoch": 1.6299892125134843, "progress_pct": 27.16, "epoch_pct": 27.17, "eta": "50:00:38", "max_grad_norm": 0.8, "loss": 0.594299852848053, "grad_norm": 1.2175726890563965, "learning_rate": 8.96798083691428e-05}
+{"ts": "2025-12-24T09:02:49", "event": "train_log", "step": 3780, "epoch": 1.6308522114347357, "progress_pct": 27.18, "epoch_pct": 27.18, "eta": "49:59:23", "max_grad_norm": 0.8, "loss": 0.566745400428772, "grad_norm": 0.9873425960540771, "learning_rate": 8.966486426665178e-05}
+{"ts": "2025-12-24T09:03:08", "event": "train_log", "step": 3782, "epoch": 1.631715210355987, "progress_pct": 27.19, "epoch_pct": 27.2, "eta": "49:58:03", "max_grad_norm": 0.8, "loss": 0.76564621925354, "grad_norm": 1.4304969310760498, "learning_rate": 8.964991059915756e-05}
+{"ts": "2025-12-24T09:03:27", "event": "train_log", "step": 3784, "epoch": 1.6325782092772383, "progress_pct": 27.21, "epoch_pct": 27.21, "eta": "49:56:45", "max_grad_norm": 0.8, "loss": 0.7026722431182861, "grad_norm": 1.1225755214691162, "learning_rate": 8.963494737026611e-05}
+{"ts": "2025-12-24T09:03:47", "event": "train_log", "step": 3786, "epoch": 1.6334412081984897, "progress_pct": 27.22, "epoch_pct": 27.22, "eta": "49:55:28", "max_grad_norm": 0.8, "loss": 0.6778496503829956, "grad_norm": 1.251551628112793, "learning_rate": 8.961997458358581e-05}
+{"ts": "2025-12-24T09:04:09", "event": "train_log", "step": 3788, "epoch": 1.6343042071197411, "progress_pct": 27.24, "epoch_pct": 27.24, "eta": "49:54:15", "max_grad_norm": 0.8, "loss": 0.5866316556930542, "grad_norm": 0.9573827981948853, "learning_rate": 8.960499224272725e-05}
+{"ts": "2025-12-24T09:04:29", "event": "train_log", "step": 3790, "epoch": 1.6351672060409923, "progress_pct": 27.25, "epoch_pct": 27.25, "eta": "49:52:59", "max_grad_norm": 0.8, "loss": 0.6244695782661438, "grad_norm": 1.0466986894607544, "learning_rate": 8.959000035130334e-05}
+{"ts": "2025-12-24T09:04:51", "event": "train_log", "step": 3792, "epoch": 1.636030204962244, "progress_pct": 27.26, "epoch_pct": 27.27, "eta": "49:51:48", "max_grad_norm": 0.8, "loss": 0.6380400061607361, "grad_norm": 1.0724520683288574, "learning_rate": 8.957499891292931e-05}
+{"ts": "2025-12-24T09:05:09", "event": "train_log", "step": 3794, "epoch": 1.6368932038834951, "progress_pct": 27.28, "epoch_pct": 27.28, "eta": "49:50:24", "max_grad_norm": 0.8, "loss": 0.625150203704834, "grad_norm": 1.1590946912765503, "learning_rate": 8.955998793122271e-05}
+{"ts": "2025-12-24T09:05:27", "event": "train_log", "step": 3796, "epoch": 1.6377562028047465, "progress_pct": 27.29, "epoch_pct": 27.3, "eta": "49:49:03", "max_grad_norm": 0.8, "loss": 0.7554174065589905, "grad_norm": 1.3405407667160034, "learning_rate": 8.954496740980334e-05}
+{"ts": "2025-12-24T09:05:46", "event": "train_log", "step": 3798, "epoch": 1.638619201725998, "progress_pct": 27.31, "epoch_pct": 27.31, "eta": "49:47:45", "max_grad_norm": 0.8, "loss": 0.6710763573646545, "grad_norm": 1.2695956230163574, "learning_rate": 8.952993735229338e-05}
+{"ts": "2025-12-24T09:06:03", "event": "train_log", "step": 3800, "epoch": 1.6394822006472491, "progress_pct": 27.32, "epoch_pct": 27.32, "eta": "49:46:19", "max_grad_norm": 0.8, "loss": 0.7102031111717224, "grad_norm": 1.4330781698226929, "learning_rate": 8.95148977623172e-05}
+{"ts": "2025-12-24T09:19:47", "event": "train_log", "step": 3800, "epoch": 1.6394822006472491, "progress_pct": 27.32, "epoch_pct": 27.32, "eta": "50:22:50", "max_grad_norm": 0.8, "eval_loss": 0.6974969506263733, "eval_runtime": 823.7017, "eval_samples_per_second": 2.501, "eval_steps_per_second": 2.501}
+{"ts": "2025-12-24T09:20:07", "event": "train_log", "step": 3802, "epoch": 1.6403451995685006, "progress_pct": 27.34, "epoch_pct": 27.34, "eta": "50:21:33", "max_grad_norm": 0.8, "loss": 0.6028781533241272, "grad_norm": 0.9818692803382874, "learning_rate": 8.949984864350157e-05}
+{"ts": "2025-12-24T09:20:26", "event": "train_log", "step": 3804, "epoch": 1.641208198489752, "progress_pct": 27.35, "epoch_pct": 27.35, "eta": "50:20:13", "max_grad_norm": 0.8, "loss": 0.649263322353363, "grad_norm": 1.0779424905776978, "learning_rate": 8.94847899994755e-05}
+{"ts": "2025-12-24T09:20:44", "event": "train_log", "step": 3806, "epoch": 1.6420711974110032, "progress_pct": 27.37, "epoch_pct": 27.37, "eta": "50:18:49", "max_grad_norm": 0.8, "loss": 0.6424242258071899, "grad_norm": 1.2096810340881348, "learning_rate": 8.946972183387033e-05}
+{"ts": "2025-12-24T09:21:02", "event": "train_log", "step": 3808, "epoch": 1.6429341963322546, "progress_pct": 27.38, "epoch_pct": 27.38, "eta": "50:17:27", "max_grad_norm": 0.8, "loss": 0.6751719117164612, "grad_norm": 1.1697514057159424, "learning_rate": 8.945464415031966e-05}
+{"ts": "2025-12-24T09:21:22", "event": "train_log", "step": 3810, "epoch": 1.643797195253506, "progress_pct": 27.39, "epoch_pct": 27.4, "eta": "50:16:08", "max_grad_norm": 0.8, "loss": 0.616973340511322, "grad_norm": 1.6524088382720947, "learning_rate": 8.943955695245942e-05}
+{"ts": "2025-12-24T09:21:39", "event": "train_log", "step": 3812, "epoch": 1.6446601941747572, "progress_pct": 27.41, "epoch_pct": 27.41, "eta": "50:14:41", "max_grad_norm": 0.8, "loss": 0.6577548980712891, "grad_norm": 1.1154218912124634, "learning_rate": 8.942446024392782e-05}
+{"ts": "2025-12-24T09:21:59", "event": "train_log", "step": 3814, "epoch": 1.6455231930960086, "progress_pct": 27.42, "epoch_pct": 27.43, "eta": "50:13:24", "max_grad_norm": 0.8, "loss": 0.6534224152565002, "grad_norm": 1.1747835874557495, "learning_rate": 8.940935402836535e-05}
+{"ts": "2025-12-24T09:22:19", "event": "train_log", "step": 3816, "epoch": 1.64638619201726, "progress_pct": 27.44, "epoch_pct": 27.44, "eta": "50:12:08", "max_grad_norm": 0.8, "loss": 0.6356330513954163, "grad_norm": 1.189562439918518, "learning_rate": 8.939423830941482e-05}
+{"ts": "2025-12-24T09:22:39", "event": "train_log", "step": 3818, "epoch": 1.6472491909385112, "progress_pct": 27.45, "epoch_pct": 27.45, "eta": "50:10:49", "max_grad_norm": 0.8, "loss": 0.6097415685653687, "grad_norm": 1.0842949151992798, "learning_rate": 8.937911309072132e-05}
+{"ts": "2025-12-24T09:22:58", "event": "train_log", "step": 3820, "epoch": 1.6481121898597628, "progress_pct": 27.47, "epoch_pct": 27.47, "eta": "50:09:30", "max_grad_norm": 0.8, "loss": 0.6339239478111267, "grad_norm": 1.1931639909744263, "learning_rate": 8.936397837593221e-05}
+{"ts": "2025-12-24T09:23:17", "event": "train_log", "step": 3822, "epoch": 1.648975188781014, "progress_pct": 27.48, "epoch_pct": 27.48, "eta": "50:08:08", "max_grad_norm": 0.8, "loss": 0.6588500142097473, "grad_norm": 1.4980956315994263, "learning_rate": 8.934883416869717e-05}
+{"ts": "2025-12-24T09:23:36", "event": "train_log", "step": 3824, "epoch": 1.6498381877022654, "progress_pct": 27.49, "epoch_pct": 27.5, "eta": "50:06:49", "max_grad_norm": 0.8, "loss": 0.6418336033821106, "grad_norm": 1.2506407499313354, "learning_rate": 8.933368047266814e-05}
+{"ts": "2025-12-24T09:23:56", "event": "train_log", "step": 3826, "epoch": 1.6507011866235168, "progress_pct": 27.51, "epoch_pct": 27.51, "eta": "50:05:32", "max_grad_norm": 0.8, "loss": 0.6475774049758911, "grad_norm": 1.3003007173538208, "learning_rate": 8.93185172914994e-05}
+{"ts": "2025-12-24T09:24:15", "event": "train_log", "step": 3828, "epoch": 1.651564185544768, "progress_pct": 27.52, "epoch_pct": 27.53, "eta": "50:04:11", "max_grad_norm": 0.8, "loss": 0.7352058291435242, "grad_norm": 1.2024277448654175, "learning_rate": 8.930334462884747e-05}
+{"ts": "2025-12-24T09:24:33", "event": "train_log", "step": 3830, "epoch": 1.6524271844660194, "progress_pct": 27.54, "epoch_pct": 27.54, "eta": "50:02:48", "max_grad_norm": 0.8, "loss": 0.6267798542976379, "grad_norm": 1.0713937282562256, "learning_rate": 8.928816248837115e-05}
+{"ts": "2025-12-24T09:24:51", "event": "train_log", "step": 3832, "epoch": 1.6532901833872709, "progress_pct": 27.55, "epoch_pct": 27.55, "eta": "50:01:26", "max_grad_norm": 0.8, "loss": 0.7097233533859253, "grad_norm": 1.3538615703582764, "learning_rate": 8.927297087373157e-05}
+{"ts": "2025-12-24T09:25:11", "event": "train_log", "step": 3834, "epoch": 1.654153182308522, "progress_pct": 27.57, "epoch_pct": 27.57, "eta": "50:00:09", "max_grad_norm": 0.8, "loss": 0.6012749671936035, "grad_norm": 1.1811957359313965, "learning_rate": 8.92577697885921e-05}
+{"ts": "2025-12-24T09:25:31", "event": "train_log", "step": 3836, "epoch": 1.6550161812297735, "progress_pct": 27.58, "epoch_pct": 27.58, "eta": "49:58:52", "max_grad_norm": 0.8, "loss": 0.6257606744766235, "grad_norm": 1.064659595489502, "learning_rate": 8.924255923661841e-05}
+{"ts": "2025-12-24T09:25:50", "event": "train_log", "step": 3838, "epoch": 1.6558791801510249, "progress_pct": 27.6, "epoch_pct": 27.6, "eta": "49:57:33", "max_grad_norm": 0.8, "loss": 0.7108950018882751, "grad_norm": 1.1258097887039185, "learning_rate": 8.922733922147848e-05}
+{"ts": "2025-12-24T09:26:11", "event": "train_log", "step": 3840, "epoch": 1.656742179072276, "progress_pct": 27.61, "epoch_pct": 27.61, "eta": "49:56:18", "max_grad_norm": 0.8, "loss": 0.6432359218597412, "grad_norm": 1.4515002965927124, "learning_rate": 8.921210974684253e-05}
+{"ts": "2025-12-24T09:26:31", "event": "train_log", "step": 3842, "epoch": 1.6576051779935275, "progress_pct": 27.62, "epoch_pct": 27.63, "eta": "49:55:02", "max_grad_norm": 0.8, "loss": 0.5962746739387512, "grad_norm": 1.040401816368103, "learning_rate": 8.919687081638308e-05}
+{"ts": "2025-12-24T09:26:50", "event": "train_log", "step": 3844, "epoch": 1.6584681769147789, "progress_pct": 27.64, "epoch_pct": 27.64, "eta": "49:53:41", "max_grad_norm": 0.8, "loss": 0.6396839022636414, "grad_norm": 1.0090585947036743, "learning_rate": 8.918162243377493e-05}
+{"ts": "2025-12-24T09:27:07", "event": "train_log", "step": 3846, "epoch": 1.65933117583603, "progress_pct": 27.65, "epoch_pct": 27.66, "eta": "49:52:17", "max_grad_norm": 0.8, "loss": 0.7058185935020447, "grad_norm": 1.2226985692977905, "learning_rate": 8.916636460269517e-05}
+{"ts": "2025-12-24T09:27:29", "event": "train_log", "step": 3848, "epoch": 1.6601941747572817, "progress_pct": 27.67, "epoch_pct": 27.67, "eta": "49:51:06", "max_grad_norm": 0.8, "loss": 0.6138123273849487, "grad_norm": 1.117598295211792, "learning_rate": 8.915109732682316e-05}
+{"ts": "2025-12-24T09:27:49", "event": "train_log", "step": 3850, "epoch": 1.661057173678533, "progress_pct": 27.68, "epoch_pct": 27.68, "eta": "49:49:49", "max_grad_norm": 0.8, "loss": 0.6389812231063843, "grad_norm": 1.1237187385559082, "learning_rate": 8.913582060984052e-05}
+{"ts": "2025-12-24T09:28:08", "event": "train_log", "step": 3852, "epoch": 1.661920172599784, "progress_pct": 27.7, "epoch_pct": 27.7, "eta": "49:48:30", "max_grad_norm": 0.8, "loss": 0.7189341187477112, "grad_norm": 1.148144245147705, "learning_rate": 8.912053445543114e-05}
+{"ts": "2025-12-24T09:28:27", "event": "train_log", "step": 3854, "epoch": 1.6627831715210357, "progress_pct": 27.71, "epoch_pct": 27.71, "eta": "49:47:11", "max_grad_norm": 0.8, "loss": 0.671591579914093, "grad_norm": 1.3115259408950806, "learning_rate": 8.910523886728126e-05}
+{"ts": "2025-12-24T09:28:47", "event": "train_log", "step": 3856, "epoch": 1.663646170442287, "progress_pct": 27.73, "epoch_pct": 27.73, "eta": "49:45:54", "max_grad_norm": 0.8, "loss": 0.6695008873939514, "grad_norm": 1.0031025409698486, "learning_rate": 8.908993384907928e-05}
+{"ts": "2025-12-24T09:29:07", "event": "train_log", "step": 3858, "epoch": 1.6645091693635383, "progress_pct": 27.74, "epoch_pct": 27.74, "eta": "49:44:37", "max_grad_norm": 0.8, "loss": 0.6557230353355408, "grad_norm": 1.2426995038986206, "learning_rate": 8.907461940451601e-05}
+{"ts": "2025-12-24T09:29:27", "event": "train_log", "step": 3860, "epoch": 1.6653721682847897, "progress_pct": 27.75, "epoch_pct": 27.76, "eta": "49:43:22", "max_grad_norm": 0.8, "loss": 0.6126149296760559, "grad_norm": 1.1471003293991089, "learning_rate": 8.90592955372844e-05}
+{"ts": "2025-12-24T09:29:47", "event": "train_log", "step": 3862, "epoch": 1.666235167206041, "progress_pct": 27.77, "epoch_pct": 27.77, "eta": "49:42:04", "max_grad_norm": 0.8, "loss": 0.742831826210022, "grad_norm": 1.2763721942901611, "learning_rate": 8.904396225107976e-05}
+{"ts": "2025-12-24T09:30:06", "event": "train_log", "step": 3864, "epoch": 1.6670981661272923, "progress_pct": 27.78, "epoch_pct": 27.78, "eta": "49:40:47", "max_grad_norm": 0.8, "loss": 0.6506689190864563, "grad_norm": 1.0612274408340454, "learning_rate": 8.902861954959963e-05}
+{"ts": "2025-12-24T09:30:25", "event": "train_log", "step": 3866, "epoch": 1.6679611650485437, "progress_pct": 27.8, "epoch_pct": 27.8, "eta": "49:39:27", "max_grad_norm": 0.8, "loss": 0.6939412355422974, "grad_norm": 1.160914421081543, "learning_rate": 8.901326743654384e-05}
+{"ts": "2025-12-24T09:30:46", "event": "train_log", "step": 3868, "epoch": 1.668824163969795, "progress_pct": 27.81, "epoch_pct": 27.81, "eta": "49:38:15", "max_grad_norm": 0.8, "loss": 0.6331133842468262, "grad_norm": 1.2517154216766357, "learning_rate": 8.899790591561447e-05}
+{"ts": "2025-12-24T09:31:06", "event": "train_log", "step": 3870, "epoch": 1.6696871628910464, "progress_pct": 27.83, "epoch_pct": 27.83, "eta": "49:36:58", "max_grad_norm": 0.8, "loss": 0.6425275206565857, "grad_norm": 0.9491620063781738, "learning_rate": 8.89825349905159e-05}
+{"ts": "2025-12-24T09:31:28", "event": "train_log", "step": 3872, "epoch": 1.6705501618122978, "progress_pct": 27.84, "epoch_pct": 27.84, "eta": "49:35:46", "max_grad_norm": 0.8, "loss": 0.5937016606330872, "grad_norm": 0.9615546464920044, "learning_rate": 8.896715466495476e-05}
+{"ts": "2025-12-24T09:31:47", "event": "train_log", "step": 3874, "epoch": 1.671413160733549, "progress_pct": 27.85, "epoch_pct": 27.86, "eta": "49:34:28", "max_grad_norm": 0.8, "loss": 0.7523217797279358, "grad_norm": 1.1985108852386475, "learning_rate": 8.895176494263992e-05}
+{"ts": "2025-12-24T09:32:07", "event": "train_log", "step": 3876, "epoch": 1.6722761596548006, "progress_pct": 27.87, "epoch_pct": 27.87, "eta": "49:33:12", "max_grad_norm": 0.8, "loss": 0.6035639047622681, "grad_norm": 1.1150988340377808, "learning_rate": 8.893636582728257e-05}
+{"ts": "2025-12-24T09:32:26", "event": "train_log", "step": 3878, "epoch": 1.6731391585760518, "progress_pct": 27.88, "epoch_pct": 27.89, "eta": "49:31:55", "max_grad_norm": 0.8, "loss": 0.6141317486763, "grad_norm": 0.9545241594314575, "learning_rate": 8.892095732259612e-05}
+{"ts": "2025-12-24T09:32:46", "event": "train_log", "step": 3880, "epoch": 1.674002157497303, "progress_pct": 27.9, "epoch_pct": 27.9, "eta": "49:30:39", "max_grad_norm": 0.8, "loss": 0.6015914678573608, "grad_norm": 1.2937493324279785, "learning_rate": 8.890553943229627e-05}
+{"ts": "2025-12-24T09:33:07", "event": "train_log", "step": 3882, "epoch": 1.6748651564185546, "progress_pct": 27.91, "epoch_pct": 27.91, "eta": "49:29:25", "max_grad_norm": 0.8, "loss": 0.6010603904724121, "grad_norm": 1.0606268644332886, "learning_rate": 8.889011216010099e-05}
+{"ts": "2025-12-24T09:33:28", "event": "train_log", "step": 3884, "epoch": 1.6757281553398058, "progress_pct": 27.93, "epoch_pct": 27.93, "eta": "49:28:12", "max_grad_norm": 0.8, "loss": 0.6256402134895325, "grad_norm": 1.0271397829055786, "learning_rate": 8.887467550973046e-05}
+{"ts": "2025-12-24T09:33:48", "event": "train_log", "step": 3886, "epoch": 1.6765911542610572, "progress_pct": 27.94, "epoch_pct": 27.94, "eta": "49:26:56", "max_grad_norm": 0.8, "loss": 0.6119387745857239, "grad_norm": 1.096360683441162, "learning_rate": 8.885922948490719e-05}
+{"ts": "2025-12-24T09:34:07", "event": "train_log", "step": 3888, "epoch": 1.6774541531823086, "progress_pct": 27.96, "epoch_pct": 27.96, "eta": "49:25:39", "max_grad_norm": 0.8, "loss": 0.6526403427124023, "grad_norm": 1.1860841512680054, "learning_rate": 8.884377408935589e-05}
+{"ts": "2025-12-24T09:34:26", "event": "train_log", "step": 3890, "epoch": 1.6783171521035598, "progress_pct": 27.97, "epoch_pct": 27.97, "eta": "49:24:20", "max_grad_norm": 0.8, "loss": 0.5966535806655884, "grad_norm": 1.1262441873550415, "learning_rate": 8.882830932680359e-05}
+{"ts": "2025-12-24T09:34:45", "event": "train_log", "step": 3892, "epoch": 1.6791801510248112, "progress_pct": 27.98, "epoch_pct": 27.99, "eta": "49:23:02", "max_grad_norm": 0.8, "loss": 0.6420881748199463, "grad_norm": 1.0326828956604004, "learning_rate": 8.881283520097953e-05}
+{"ts": "2025-12-24T09:35:05", "event": "train_log", "step": 3894, "epoch": 1.6800431499460626, "progress_pct": 28.0, "epoch_pct": 28.0, "eta": "49:21:46", "max_grad_norm": 0.8, "loss": 0.6054600477218628, "grad_norm": 1.1928532123565674, "learning_rate": 8.879735171561525e-05}
+{"ts": "2025-12-24T09:35:22", "event": "train_log", "step": 3896, "epoch": 1.6809061488673138, "progress_pct": 28.01, "epoch_pct": 28.02, "eta": "49:20:26", "max_grad_norm": 0.8, "loss": 0.7106166481971741, "grad_norm": 1.2887117862701416, "learning_rate": 8.878185887444448e-05}
+{"ts": "2025-12-24T09:35:41", "event": "train_log", "step": 3898, "epoch": 1.6817691477885652, "progress_pct": 28.03, "epoch_pct": 28.03, "eta": "49:19:06", "max_grad_norm": 0.8, "loss": 0.7001441717147827, "grad_norm": 1.1483163833618164, "learning_rate": 8.876635668120327e-05}
+{"ts": "2025-12-24T09:36:00", "event": "train_log", "step": 3900, "epoch": 1.6826321467098166, "progress_pct": 28.04, "epoch_pct": 28.04, "eta": "49:17:48", "max_grad_norm": 0.8, "loss": 0.6607111692428589, "grad_norm": 1.2908320426940918, "learning_rate": 8.875084513962993e-05}
+{"ts": "2025-12-24T09:49:37", "event": "train_log", "step": 3900, "epoch": 1.6826321467098166, "progress_pct": 28.04, "epoch_pct": 28.04, "eta": "49:52:45", "max_grad_norm": 0.8, "eval_loss": 0.6949095726013184, "eval_runtime": 817.4497, "eval_samples_per_second": 2.52, "eval_steps_per_second": 2.52}
+{"ts": "2025-12-24T09:49:57", "event": "train_log", "step": 3902, "epoch": 1.6834951456310678, "progress_pct": 28.06, "epoch_pct": 28.06, "eta": "49:51:28", "max_grad_norm": 0.8, "loss": 0.6463292837142944, "grad_norm": 1.00212824344635, "learning_rate": 8.873532425346496e-05}
+{"ts": "2025-12-24T09:50:15", "event": "train_log", "step": 3904, "epoch": 1.6843581445523195, "progress_pct": 28.07, "epoch_pct": 28.07, "eta": "49:50:07", "max_grad_norm": 0.8, "loss": 0.6627839207649231, "grad_norm": 1.2223109006881714, "learning_rate": 8.871979402645116e-05}
+{"ts": "2025-12-24T09:50:34", "event": "train_log", "step": 3906, "epoch": 1.6852211434735707, "progress_pct": 28.08, "epoch_pct": 28.09, "eta": "49:48:48", "max_grad_norm": 0.8, "loss": 0.6311076879501343, "grad_norm": 1.047956109046936, "learning_rate": 8.870425446233359e-05}
+{"ts": "2025-12-24T09:50:53", "event": "train_log", "step": 3908, "epoch": 1.6860841423948218, "progress_pct": 28.1, "epoch_pct": 28.1, "eta": "49:47:28", "max_grad_norm": 0.8, "loss": 0.6916288137435913, "grad_norm": 1.178468108177185, "learning_rate": 8.868870556485954e-05}
+{"ts": "2025-12-24T09:51:11", "event": "train_log", "step": 3910, "epoch": 1.6869471413160735, "progress_pct": 28.11, "epoch_pct": 28.12, "eta": "49:46:08", "max_grad_norm": 0.8, "loss": 0.696479856967926, "grad_norm": 6.569681644439697, "learning_rate": 8.867314733777854e-05}
+{"ts": "2025-12-24T09:51:30", "event": "train_log", "step": 3912, "epoch": 1.6878101402373247, "progress_pct": 28.13, "epoch_pct": 28.13, "eta": "49:44:50", "max_grad_norm": 0.8, "loss": 0.6245654225349426, "grad_norm": 1.100183129310608, "learning_rate": 8.865757978484241e-05}
+{"ts": "2025-12-24T09:51:50", "event": "train_log", "step": 3914, "epoch": 1.688673139158576, "progress_pct": 28.14, "epoch_pct": 28.14, "eta": "49:43:33", "max_grad_norm": 0.8, "loss": 0.6178512573242188, "grad_norm": 1.222562313079834, "learning_rate": 8.864200290980518e-05}
+{"ts": "2025-12-24T09:52:10", "event": "train_log", "step": 3916, "epoch": 1.6895361380798275, "progress_pct": 28.16, "epoch_pct": 28.16, "eta": "49:42:15", "max_grad_norm": 0.8, "loss": 0.6598679423332214, "grad_norm": 1.235471487045288, "learning_rate": 8.862641671642314e-05}
+{"ts": "2025-12-24T09:52:29", "event": "train_log", "step": 3918, "epoch": 1.6903991370010787, "progress_pct": 28.17, "epoch_pct": 28.17, "eta": "49:40:56", "max_grad_norm": 0.8, "loss": 0.6503698825836182, "grad_norm": 1.2539454698562622, "learning_rate": 8.861082120845486e-05}
+{"ts": "2025-12-24T09:52:48", "event": "train_log", "step": 3920, "epoch": 1.69126213592233, "progress_pct": 28.19, "epoch_pct": 28.19, "eta": "49:39:38", "max_grad_norm": 0.8, "loss": 0.6858329176902771, "grad_norm": 1.1766268014907837, "learning_rate": 8.859521638966107e-05}
+{"ts": "2025-12-24T09:53:06", "event": "train_log", "step": 3922, "epoch": 1.6921251348435815, "progress_pct": 28.2, "epoch_pct": 28.2, "eta": "49:38:18", "max_grad_norm": 0.8, "loss": 0.679460883140564, "grad_norm": 1.3463754653930664, "learning_rate": 8.857960226380486e-05}
+{"ts": "2025-12-24T09:53:24", "event": "train_log", "step": 3924, "epoch": 1.6929881337648327, "progress_pct": 28.21, "epoch_pct": 28.22, "eta": "49:36:58", "max_grad_norm": 0.8, "loss": 0.6901042461395264, "grad_norm": 1.3011246919631958, "learning_rate": 8.856397883465149e-05}
+{"ts": "2025-12-24T09:53:45", "event": "train_log", "step": 3926, "epoch": 1.6938511326860841, "progress_pct": 28.23, "epoch_pct": 28.23, "eta": "49:35:44", "max_grad_norm": 0.8, "loss": 0.6756444573402405, "grad_norm": 0.9615590572357178, "learning_rate": 8.854834610596845e-05}
+{"ts": "2025-12-24T09:54:05", "event": "train_log", "step": 3928, "epoch": 1.6947141316073355, "progress_pct": 28.24, "epoch_pct": 28.25, "eta": "49:34:27", "max_grad_norm": 0.8, "loss": 0.6690155863761902, "grad_norm": 0.9635165333747864, "learning_rate": 8.853270408152552e-05}
+{"ts": "2025-12-24T09:54:23", "event": "train_log", "step": 3930, "epoch": 1.6955771305285867, "progress_pct": 28.26, "epoch_pct": 28.26, "eta": "49:33:07", "max_grad_norm": 0.8, "loss": 0.6412895321846008, "grad_norm": 1.2924212217330933, "learning_rate": 8.851705276509472e-05}
+{"ts": "2025-12-24T09:54:43", "event": "train_log", "step": 3932, "epoch": 1.6964401294498384, "progress_pct": 28.27, "epoch_pct": 28.27, "eta": "49:31:51", "max_grad_norm": 0.8, "loss": 0.6250708699226379, "grad_norm": 1.0281823873519897, "learning_rate": 8.850139216045028e-05}
+{"ts": "2025-12-24T09:55:01", "event": "train_log", "step": 3934, "epoch": 1.6973031283710895, "progress_pct": 28.29, "epoch_pct": 28.29, "eta": "49:30:30", "max_grad_norm": 0.8, "loss": 0.6545571684837341, "grad_norm": 1.321605920791626, "learning_rate": 8.848572227136868e-05}
+{"ts": "2025-12-24T09:55:20", "event": "train_log", "step": 3936, "epoch": 1.6981661272923407, "progress_pct": 28.3, "epoch_pct": 28.3, "eta": "49:29:12", "max_grad_norm": 0.8, "loss": 0.6466772556304932, "grad_norm": 1.1515092849731445, "learning_rate": 8.847004310162866e-05}
+{"ts": "2025-12-24T09:55:39", "event": "train_log", "step": 3938, "epoch": 1.6990291262135924, "progress_pct": 28.31, "epoch_pct": 28.32, "eta": "49:27:53", "max_grad_norm": 0.8, "loss": 0.679786741733551, "grad_norm": 1.1318165063858032, "learning_rate": 8.845435465501116e-05}
+{"ts": "2025-12-24T09:55:52", "event": "train_log", "step": 3940, "epoch": 1.6998921251348436, "progress_pct": 28.33, "epoch_pct": 28.33, "eta": "49:26:21", "max_grad_norm": 0.8, "loss": 0.6546611189842224, "grad_norm": 0.973469078540802, "learning_rate": 8.84386569352994e-05}
+{"ts": "2025-12-24T09:56:06", "event": "train_log", "step": 3942, "epoch": 1.700755124056095, "progress_pct": 28.34, "epoch_pct": 28.35, "eta": "49:24:49", "max_grad_norm": 0.8, "loss": 0.6207357048988342, "grad_norm": 1.0992401838302612, "learning_rate": 8.842294994627879e-05}
+{"ts": "2025-12-24T09:56:18", "event": "train_log", "step": 3944, "epoch": 1.7016181229773464, "progress_pct": 28.36, "epoch_pct": 28.36, "eta": "49:23:15", "max_grad_norm": 0.8, "loss": 0.6113827228546143, "grad_norm": 1.1114588975906372, "learning_rate": 8.840723369173702e-05}
+{"ts": "2025-12-24T09:56:31", "event": "train_log", "step": 3946, "epoch": 1.7024811218985976, "progress_pct": 28.37, "epoch_pct": 28.37, "eta": "49:21:41", "max_grad_norm": 0.8, "loss": 0.6246283650398254, "grad_norm": 1.171692132949829, "learning_rate": 8.839150817546398e-05}
+{"ts": "2025-12-24T09:56:42", "event": "train_log", "step": 3948, "epoch": 1.703344120819849, "progress_pct": 28.39, "epoch_pct": 28.39, "eta": "49:20:04", "max_grad_norm": 0.8, "loss": 0.7027832269668579, "grad_norm": 1.249851942062378, "learning_rate": 8.837577340125181e-05}
+{"ts": "2025-12-24T09:56:55", "event": "train_log", "step": 3950, "epoch": 1.7042071197411004, "progress_pct": 28.4, "epoch_pct": 28.4, "eta": "49:18:30", "max_grad_norm": 0.8, "loss": 0.7121281027793884, "grad_norm": 1.2475496530532837, "learning_rate": 8.836002937289489e-05}
+{"ts": "2025-12-24T09:57:07", "event": "train_log", "step": 3952, "epoch": 1.7050701186623516, "progress_pct": 28.42, "epoch_pct": 28.42, "eta": "49:16:56", "max_grad_norm": 0.8, "loss": 0.6521060466766357, "grad_norm": 1.048726201057434, "learning_rate": 8.834427609418981e-05}
+{"ts": "2025-12-24T09:57:19", "event": "train_log", "step": 3954, "epoch": 1.705933117583603, "progress_pct": 28.43, "epoch_pct": 28.43, "eta": "49:15:21", "max_grad_norm": 0.8, "loss": 0.6735731959342957, "grad_norm": 1.1689599752426147, "learning_rate": 8.832851356893541e-05}
+{"ts": "2025-12-24T09:57:31", "event": "train_log", "step": 3956, "epoch": 1.7067961165048544, "progress_pct": 28.44, "epoch_pct": 28.45, "eta": "49:13:45", "max_grad_norm": 0.8, "loss": 0.6530178189277649, "grad_norm": 1.4230015277862549, "learning_rate": 8.831274180093275e-05}
+{"ts": "2025-12-24T09:57:43", "event": "train_log", "step": 3958, "epoch": 1.7076591154261056, "progress_pct": 28.46, "epoch_pct": 28.46, "eta": "49:12:11", "max_grad_norm": 0.8, "loss": 0.6858299970626831, "grad_norm": 1.2237123250961304, "learning_rate": 8.829696079398511e-05}
+{"ts": "2025-12-24T09:57:56", "event": "train_log", "step": 3960, "epoch": 1.708522114347357, "progress_pct": 28.47, "epoch_pct": 28.48, "eta": "49:10:37", "max_grad_norm": 0.8, "loss": 0.6567219495773315, "grad_norm": 1.1330037117004395, "learning_rate": 8.828117055189803e-05}
+{"ts": "2025-12-24T09:58:08", "event": "train_log", "step": 3962, "epoch": 1.7093851132686084, "progress_pct": 28.49, "epoch_pct": 28.49, "eta": "49:09:03", "max_grad_norm": 0.8, "loss": 0.6207150816917419, "grad_norm": 1.2441636323928833, "learning_rate": 8.826537107847925e-05}
+{"ts": "2025-12-24T09:58:20", "event": "train_log", "step": 3964, "epoch": 1.7102481121898596, "progress_pct": 28.5, "epoch_pct": 28.5, "eta": "49:07:29", "max_grad_norm": 0.8, "loss": 0.6192324161529541, "grad_norm": 0.9952784776687622, "learning_rate": 8.824956237753872e-05}
+{"ts": "2025-12-24T09:58:33", "event": "train_log", "step": 3966, "epoch": 1.7111111111111112, "progress_pct": 28.52, "epoch_pct": 28.52, "eta": "49:05:56", "max_grad_norm": 0.8, "loss": 0.6485670208930969, "grad_norm": 1.0648247003555298, "learning_rate": 8.823374445288865e-05}
+{"ts": "2025-12-24T09:58:45", "event": "train_log", "step": 3968, "epoch": 1.7119741100323624, "progress_pct": 28.53, "epoch_pct": 28.53, "eta": "49:04:21", "max_grad_norm": 0.8, "loss": 0.6591274738311768, "grad_norm": 1.2040860652923584, "learning_rate": 8.821791730834347e-05}
+{"ts": "2025-12-24T09:58:57", "event": "train_log", "step": 3970, "epoch": 1.7128371089536139, "progress_pct": 28.54, "epoch_pct": 28.55, "eta": "49:02:48", "max_grad_norm": 0.8, "loss": 0.6427604556083679, "grad_norm": 1.2346793413162231, "learning_rate": 8.820208094771982e-05}
+{"ts": "2025-12-24T09:59:10", "event": "train_log", "step": 3972, "epoch": 1.7137001078748653, "progress_pct": 28.56, "epoch_pct": 28.56, "eta": "49:01:15", "max_grad_norm": 0.8, "loss": 0.5952412486076355, "grad_norm": 1.0716660022735596, "learning_rate": 8.818623537483658e-05}
+{"ts": "2025-12-24T09:59:23", "event": "train_log", "step": 3974, "epoch": 1.7145631067961165, "progress_pct": 28.57, "epoch_pct": 28.58, "eta": "48:59:43", "max_grad_norm": 0.8, "loss": 0.6808739900588989, "grad_norm": 1.1809104681015015, "learning_rate": 8.817038059351481e-05}
+{"ts": "2025-12-24T09:59:35", "event": "train_log", "step": 3976, "epoch": 1.7154261057173679, "progress_pct": 28.59, "epoch_pct": 28.59, "eta": "48:58:10", "max_grad_norm": 0.8, "loss": 0.6927714347839355, "grad_norm": 1.224974513053894, "learning_rate": 8.815451660757783e-05}
+{"ts": "2025-12-24T09:59:48", "event": "train_log", "step": 3978, "epoch": 1.7162891046386193, "progress_pct": 28.6, "epoch_pct": 28.6, "eta": "48:56:37", "max_grad_norm": 0.8, "loss": 0.6450885534286499, "grad_norm": 1.141288161277771, "learning_rate": 8.813864342085117e-05}
+{"ts": "2025-12-24T10:00:01", "event": "train_log", "step": 3980, "epoch": 1.7171521035598705, "progress_pct": 28.62, "epoch_pct": 28.62, "eta": "48:55:05", "max_grad_norm": 0.8, "loss": 0.6775641441345215, "grad_norm": 1.0749722719192505, "learning_rate": 8.812276103716259e-05}
+{"ts": "2025-12-24T10:00:13", "event": "train_log", "step": 3982, "epoch": 1.7180151024811219, "progress_pct": 28.63, "epoch_pct": 28.63, "eta": "48:53:32", "max_grad_norm": 0.8, "loss": 0.6083413362503052, "grad_norm": 1.13978111743927, "learning_rate": 8.810686946034203e-05}
+{"ts": "2025-12-24T10:00:25", "event": "train_log", "step": 3984, "epoch": 1.7188781014023733, "progress_pct": 28.65, "epoch_pct": 28.65, "eta": "48:51:58", "max_grad_norm": 0.8, "loss": 0.6711311936378479, "grad_norm": 1.3158389329910278, "learning_rate": 8.80909686942217e-05}
+{"ts": "2025-12-24T10:00:37", "event": "train_log", "step": 3986, "epoch": 1.7197411003236245, "progress_pct": 28.66, "epoch_pct": 28.66, "eta": "48:50:25", "max_grad_norm": 0.8, "loss": 0.6818822026252747, "grad_norm": 1.0368914604187012, "learning_rate": 8.807505874263597e-05}
+{"ts": "2025-12-24T10:00:49", "event": "train_log", "step": 3988, "epoch": 1.720604099244876, "progress_pct": 28.67, "epoch_pct": 28.68, "eta": "48:48:51", "max_grad_norm": 0.8, "loss": 0.6346645355224609, "grad_norm": 1.249948263168335, "learning_rate": 8.805913960942148e-05}
+{"ts": "2025-12-24T10:01:02", "event": "train_log", "step": 3990, "epoch": 1.7214670981661273, "progress_pct": 28.69, "epoch_pct": 28.69, "eta": "48:47:19", "max_grad_norm": 0.8, "loss": 0.6550438404083252, "grad_norm": 1.0894579887390137, "learning_rate": 8.804321129841703e-05}
+{"ts": "2025-12-24T10:01:14", "event": "train_log", "step": 3992, "epoch": 1.7223300970873785, "progress_pct": 28.7, "epoch_pct": 28.71, "eta": "48:45:45", "max_grad_norm": 0.8, "loss": 0.6257092952728271, "grad_norm": 1.2956905364990234, "learning_rate": 8.80272738134637e-05}
+{"ts": "2025-12-24T10:01:27", "event": "train_log", "step": 3994, "epoch": 1.7231930960086301, "progress_pct": 28.72, "epoch_pct": 28.72, "eta": "48:44:14", "max_grad_norm": 0.8, "loss": 0.5877772569656372, "grad_norm": 1.0089573860168457, "learning_rate": 8.80113271584047e-05}
+{"ts": "2025-12-24T10:01:39", "event": "train_log", "step": 3996, "epoch": 1.7240560949298813, "progress_pct": 28.73, "epoch_pct": 28.73, "eta": "48:42:42", "max_grad_norm": 0.8, "loss": 0.6440973281860352, "grad_norm": 1.0698174238204956, "learning_rate": 8.799537133708552e-05}
+{"ts": "2025-12-24T10:01:53", "event": "train_log", "step": 3998, "epoch": 1.7249190938511327, "progress_pct": 28.75, "epoch_pct": 28.75, "eta": "48:41:12", "max_grad_norm": 0.8, "loss": 0.6270571351051331, "grad_norm": 1.1145644187927246, "learning_rate": 8.797940635335382e-05}
+{"ts": "2025-12-24T10:02:06", "event": "train_log", "step": 4000, "epoch": 1.7257820927723841, "progress_pct": 28.76, "epoch_pct": 28.76, "eta": "48:39:43", "max_grad_norm": 0.8, "loss": 0.6463025212287903, "grad_norm": 1.0673385858535767, "learning_rate": 8.796343221105949e-05}
+{"ts": "2025-12-24T10:10:24", "event": "train_log", "step": 4000, "epoch": 1.7257820927723841, "progress_pct": 28.76, "epoch_pct": 28.76, "eta": "49:00:17", "max_grad_norm": 0.8, "eval_loss": 0.6941319704055786, "eval_runtime": 498.1423, "eval_samples_per_second": 4.135, "eval_steps_per_second": 4.135}
+{"ts": "2025-12-24T10:10:37", "event": "train_log", "step": 4002, "epoch": 1.7266450916936353, "progress_pct": 28.77, "epoch_pct": 28.78, "eta": "48:58:44", "max_grad_norm": 0.8, "loss": 0.6483243107795715, "grad_norm": 1.1285523176193237, "learning_rate": 8.794744891405461e-05}
+{"ts": "2025-12-24T10:10:50", "event": "train_log", "step": 4004, "epoch": 1.7275080906148867, "progress_pct": 28.79, "epoch_pct": 28.79, "eta": "48:57:12", "max_grad_norm": 0.8, "loss": 0.6334620714187622, "grad_norm": 1.0747114419937134, "learning_rate": 8.793145646619352e-05}
+{"ts": "2025-12-24T10:11:02", "event": "train_log", "step": 4006, "epoch": 1.7283710895361382, "progress_pct": 28.8, "epoch_pct": 28.81, "eta": "48:55:40", "max_grad_norm": 0.8, "loss": 0.6604806780815125, "grad_norm": 1.0277562141418457, "learning_rate": 8.791545487133269e-05}
+{"ts": "2025-12-24T10:11:16", "event": "train_log", "step": 4008, "epoch": 1.7292340884573894, "progress_pct": 28.82, "epoch_pct": 28.82, "eta": "48:54:09", "max_grad_norm": 0.8, "loss": 0.6033409237861633, "grad_norm": 1.0783002376556396, "learning_rate": 8.789944413333083e-05}
+{"ts": "2025-12-24T10:11:28", "event": "train_log", "step": 4010, "epoch": 1.7300970873786408, "progress_pct": 28.83, "epoch_pct": 28.83, "eta": "48:52:35", "max_grad_norm": 0.8, "loss": 0.6360779404640198, "grad_norm": 1.2508432865142822, "learning_rate": 8.788342425604887e-05}
+{"ts": "2025-12-24T10:11:40", "event": "train_log", "step": 4012, "epoch": 1.7309600862998922, "progress_pct": 28.85, "epoch_pct": 28.85, "eta": "48:51:03", "max_grad_norm": 0.8, "loss": 0.5815097689628601, "grad_norm": 1.0905060768127441, "learning_rate": 8.786739524334993e-05}
+{"ts": "2025-12-24T10:11:52", "event": "train_log", "step": 4014, "epoch": 1.7318230852211434, "progress_pct": 28.86, "epoch_pct": 28.86, "eta": "48:49:30", "max_grad_norm": 0.8, "loss": 0.7228885889053345, "grad_norm": 1.1219831705093384, "learning_rate": 8.785135709909935e-05}
+{"ts": "2025-12-24T10:12:04", "event": "train_log", "step": 4016, "epoch": 1.7326860841423948, "progress_pct": 28.88, "epoch_pct": 28.88, "eta": "48:47:56", "max_grad_norm": 0.8, "loss": 0.6539638638496399, "grad_norm": 1.2444658279418945, "learning_rate": 8.783530982716463e-05}
+{"ts": "2025-12-24T10:12:17", "event": "train_log", "step": 4018, "epoch": 1.7335490830636462, "progress_pct": 28.89, "epoch_pct": 28.89, "eta": "48:46:24", "max_grad_norm": 0.8, "loss": 0.6482524275779724, "grad_norm": 1.158237338066101, "learning_rate": 8.781925343141551e-05}
+{"ts": "2025-12-24T10:12:29", "event": "train_log", "step": 4020, "epoch": 1.7344120819848974, "progress_pct": 28.9, "epoch_pct": 28.91, "eta": "48:44:50", "max_grad_norm": 0.8, "loss": 0.699881374835968, "grad_norm": 1.288036584854126, "learning_rate": 8.78031879157239e-05}
+{"ts": "2025-12-24T10:12:42", "event": "train_log", "step": 4022, "epoch": 1.735275080906149, "progress_pct": 28.92, "epoch_pct": 28.92, "eta": "48:43:20", "max_grad_norm": 0.8, "loss": 0.6012337803840637, "grad_norm": 1.0680478811264038, "learning_rate": 8.778711328396396e-05}
+{"ts": "2025-12-24T10:12:54", "event": "train_log", "step": 4024, "epoch": 1.7361380798274002, "progress_pct": 28.93, "epoch_pct": 28.94, "eta": "48:41:46", "max_grad_norm": 0.8, "loss": 0.7336176037788391, "grad_norm": 1.3618063926696777, "learning_rate": 8.7771029540012e-05}
+{"ts": "2025-12-24T10:13:06", "event": "train_log", "step": 4026, "epoch": 1.7370010787486516, "progress_pct": 28.95, "epoch_pct": 28.95, "eta": "48:40:14", "max_grad_norm": 0.8, "loss": 0.6370863914489746, "grad_norm": 1.1720085144042969, "learning_rate": 8.775493668774652e-05}
+{"ts": "2025-12-24T10:13:17", "event": "train_log", "step": 4028, "epoch": 1.737864077669903, "progress_pct": 28.96, "epoch_pct": 28.96, "eta": "48:38:40", "max_grad_norm": 0.8, "loss": 0.6755601763725281, "grad_norm": 1.1800546646118164, "learning_rate": 8.773883473104827e-05}
+{"ts": "2025-12-24T10:13:30", "event": "train_log", "step": 4030, "epoch": 1.7387270765911542, "progress_pct": 28.98, "epoch_pct": 28.98, "eta": "48:37:08", "max_grad_norm": 0.8, "loss": 0.6209036707878113, "grad_norm": 1.1262768507003784, "learning_rate": 8.772272367380015e-05}
+{"ts": "2025-12-24T10:13:42", "event": "train_log", "step": 4032, "epoch": 1.7395900755124056, "progress_pct": 28.99, "epoch_pct": 28.99, "eta": "48:35:35", "max_grad_norm": 0.8, "loss": 0.6727437973022461, "grad_norm": 1.2891435623168945, "learning_rate": 8.770660351988727e-05}
+{"ts": "2025-12-24T10:13:55", "event": "train_log", "step": 4034, "epoch": 1.740453074433657, "progress_pct": 29.0, "epoch_pct": 29.01, "eta": "48:34:04", "max_grad_norm": 0.8, "loss": 0.6565355658531189, "grad_norm": 1.1410391330718994, "learning_rate": 8.769047427319694e-05}
+{"ts": "2025-12-24T10:14:07", "event": "train_log", "step": 4036, "epoch": 1.7413160733549082, "progress_pct": 29.02, "epoch_pct": 29.02, "eta": "48:32:33", "max_grad_norm": 0.8, "loss": 0.7123345136642456, "grad_norm": 1.075622797012329, "learning_rate": 8.767433593761864e-05}
+{"ts": "2025-12-24T10:14:19", "event": "train_log", "step": 4038, "epoch": 1.7421790722761596, "progress_pct": 29.03, "epoch_pct": 29.04, "eta": "48:31:01", "max_grad_norm": 0.8, "loss": 0.6391130685806274, "grad_norm": 1.1938068866729736, "learning_rate": 8.765818851704408e-05}
+{"ts": "2025-12-24T10:14:32", "event": "train_log", "step": 4040, "epoch": 1.743042071197411, "progress_pct": 29.05, "epoch_pct": 29.05, "eta": "48:29:30", "max_grad_norm": 0.8, "loss": 0.7138327956199646, "grad_norm": 1.2302309274673462, "learning_rate": 8.76420320153671e-05}
+{"ts": "2025-12-24T10:14:44", "event": "train_log", "step": 4042, "epoch": 1.7439050701186622, "progress_pct": 29.06, "epoch_pct": 29.07, "eta": "48:27:57", "max_grad_norm": 0.8, "loss": 0.6010884046554565, "grad_norm": 1.3204374313354492, "learning_rate": 8.76258664364838e-05}
+{"ts": "2025-12-24T10:14:56", "event": "train_log", "step": 4044, "epoch": 1.7447680690399137, "progress_pct": 29.08, "epoch_pct": 29.08, "eta": "48:26:25", "max_grad_norm": 0.8, "loss": 0.6665774583816528, "grad_norm": 1.2324718236923218, "learning_rate": 8.760969178429244e-05}
+{"ts": "2025-12-24T10:15:08", "event": "train_log", "step": 4046, "epoch": 1.745631067961165, "progress_pct": 29.09, "epoch_pct": 29.09, "eta": "48:24:53", "max_grad_norm": 0.8, "loss": 0.644280731678009, "grad_norm": 1.0836756229400635, "learning_rate": 8.759350806269342e-05}
+{"ts": "2025-12-24T10:15:22", "event": "train_log", "step": 4048, "epoch": 1.7464940668824163, "progress_pct": 29.11, "epoch_pct": 29.11, "eta": "48:23:25", "max_grad_norm": 0.8, "loss": 0.5871413946151733, "grad_norm": 1.00897216796875, "learning_rate": 8.757731527558943e-05}
+{"ts": "2025-12-24T10:15:35", "event": "train_log", "step": 4050, "epoch": 1.747357065803668, "progress_pct": 29.12, "epoch_pct": 29.12, "eta": "48:21:56", "max_grad_norm": 0.8, "loss": 0.5935707688331604, "grad_norm": 0.9774143695831299, "learning_rate": 8.756111342688527e-05}
+{"ts": "2025-12-24T10:15:48", "event": "train_log", "step": 4052, "epoch": 1.748220064724919, "progress_pct": 29.13, "epoch_pct": 29.14, "eta": "48:20:25", "max_grad_norm": 0.8, "loss": 0.6011680960655212, "grad_norm": 1.0634222030639648, "learning_rate": 8.75449025204879e-05}
+{"ts": "2025-12-24T10:16:00", "event": "train_log", "step": 4054, "epoch": 1.7490830636461703, "progress_pct": 29.15, "epoch_pct": 29.15, "eta": "48:18:55", "max_grad_norm": 0.8, "loss": 0.6434821486473083, "grad_norm": 1.2054167985916138, "learning_rate": 8.752868256030658e-05}
+{"ts": "2025-12-24T10:16:12", "event": "train_log", "step": 4056, "epoch": 1.749946062567422, "progress_pct": 29.16, "epoch_pct": 29.17, "eta": "48:17:22", "max_grad_norm": 0.8, "loss": 0.6644792556762695, "grad_norm": 1.2121766805648804, "learning_rate": 8.751245355025261e-05}
+{"ts": "2025-12-24T10:16:24", "event": "train_log", "step": 4058, "epoch": 1.750809061488673, "progress_pct": 29.18, "epoch_pct": 29.18, "eta": "48:15:50", "max_grad_norm": 0.8, "loss": 0.6489604711532593, "grad_norm": 1.186061143875122, "learning_rate": 8.749621549423959e-05}
+{"ts": "2025-12-24T10:16:37", "event": "train_log", "step": 4060, "epoch": 1.7516720604099245, "progress_pct": 29.19, "epoch_pct": 29.19, "eta": "48:14:20", "max_grad_norm": 0.8, "loss": 0.6018257737159729, "grad_norm": 1.074569821357727, "learning_rate": 8.747996839618322e-05}
+{"ts": "2025-12-24T10:16:50", "event": "train_log", "step": 4062, "epoch": 1.752535059331176, "progress_pct": 29.21, "epoch_pct": 29.21, "eta": "48:12:51", "max_grad_norm": 0.8, "loss": 0.6242427825927734, "grad_norm": 1.0770072937011719, "learning_rate": 8.746371226000145e-05}
+{"ts": "2025-12-24T10:17:02", "event": "train_log", "step": 4064, "epoch": 1.7533980582524271, "progress_pct": 29.22, "epoch_pct": 29.22, "eta": "48:11:20", "max_grad_norm": 0.8, "loss": 0.6011215448379517, "grad_norm": 1.070820927619934, "learning_rate": 8.744744708961435e-05}
+{"ts": "2025-12-24T10:17:14", "event": "train_log", "step": 4066, "epoch": 1.7542610571736785, "progress_pct": 29.23, "epoch_pct": 29.24, "eta": "48:09:49", "max_grad_norm": 0.8, "loss": 0.6660047769546509, "grad_norm": 1.163550853729248, "learning_rate": 8.743117288894419e-05}
+{"ts": "2025-12-24T10:17:27", "event": "train_log", "step": 4068, "epoch": 1.75512405609493, "progress_pct": 29.25, "epoch_pct": 29.25, "eta": "48:08:19", "max_grad_norm": 0.8, "loss": 0.6795720458030701, "grad_norm": 1.2277700901031494, "learning_rate": 8.741488966191544e-05}
+{"ts": "2025-12-24T10:17:39", "event": "train_log", "step": 4070, "epoch": 1.7559870550161811, "progress_pct": 29.26, "epoch_pct": 29.27, "eta": "48:06:49", "max_grad_norm": 0.8, "loss": 0.635612428188324, "grad_norm": 1.0723415613174438, "learning_rate": 8.73985974124547e-05}
+{"ts": "2025-12-24T10:17:51", "event": "train_log", "step": 4072, "epoch": 1.7568500539374325, "progress_pct": 29.28, "epoch_pct": 29.28, "eta": "48:05:18", "max_grad_norm": 0.8, "loss": 0.5903881788253784, "grad_norm": 1.1380949020385742, "learning_rate": 8.73822961444908e-05}
+{"ts": "2025-12-24T10:18:03", "event": "train_log", "step": 4074, "epoch": 1.757713052858684, "progress_pct": 29.29, "epoch_pct": 29.3, "eta": "48:03:46", "max_grad_norm": 0.8, "loss": 0.6036412119865417, "grad_norm": 1.1686267852783203, "learning_rate": 8.736598586195471e-05}
+{"ts": "2025-12-24T10:18:16", "event": "train_log", "step": 4076, "epoch": 1.7585760517799351, "progress_pct": 29.31, "epoch_pct": 29.31, "eta": "48:02:16", "max_grad_norm": 0.8, "loss": 0.6732221245765686, "grad_norm": 1.1719619035720825, "learning_rate": 8.734966656877956e-05}
+{"ts": "2025-12-24T10:18:28", "event": "train_log", "step": 4078, "epoch": 1.7594390507011868, "progress_pct": 29.32, "epoch_pct": 29.32, "eta": "48:00:46", "max_grad_norm": 0.8, "loss": 0.6788899898529053, "grad_norm": 1.0898852348327637, "learning_rate": 8.733333826890071e-05}
+{"ts": "2025-12-24T10:18:40", "event": "train_log", "step": 4080, "epoch": 1.760302049622438, "progress_pct": 29.34, "epoch_pct": 29.34, "eta": "47:59:14", "max_grad_norm": 0.8, "loss": 0.7186161875724792, "grad_norm": 1.405272126197815, "learning_rate": 8.731700096625563e-05}
+{"ts": "2025-12-24T10:18:52", "event": "train_log", "step": 4082, "epoch": 1.7611650485436892, "progress_pct": 29.35, "epoch_pct": 29.35, "eta": "47:57:44", "max_grad_norm": 0.8, "loss": 0.7305195331573486, "grad_norm": 1.287775993347168, "learning_rate": 8.7300654664784e-05}
+{"ts": "2025-12-24T10:19:05", "event": "train_log", "step": 4084, "epoch": 1.7620280474649408, "progress_pct": 29.36, "epoch_pct": 29.37, "eta": "47:56:15", "max_grad_norm": 0.8, "loss": 0.6772475838661194, "grad_norm": 1.2288155555725098, "learning_rate": 8.728429936842762e-05}
+{"ts": "2025-12-24T10:19:17", "event": "train_log", "step": 4086, "epoch": 1.762891046386192, "progress_pct": 29.38, "epoch_pct": 29.38, "eta": "47:54:45", "max_grad_norm": 0.8, "loss": 0.6882839202880859, "grad_norm": 1.2059677839279175, "learning_rate": 8.726793508113054e-05}
+{"ts": "2025-12-24T10:19:28", "event": "train_log", "step": 4088, "epoch": 1.7637540453074434, "progress_pct": 29.39, "epoch_pct": 29.4, "eta": "47:53:12", "max_grad_norm": 0.8, "loss": 0.72773277759552, "grad_norm": 1.3448551893234253, "learning_rate": 8.725156180683893e-05}
+{"ts": "2025-12-24T10:19:41", "event": "train_log", "step": 4090, "epoch": 1.7646170442286948, "progress_pct": 29.41, "epoch_pct": 29.41, "eta": "47:51:44", "max_grad_norm": 0.8, "loss": 0.6122981309890747, "grad_norm": 1.088852882385254, "learning_rate": 8.723517954950111e-05}
+{"ts": "2025-12-24T10:19:53", "event": "train_log", "step": 4092, "epoch": 1.765480043149946, "progress_pct": 29.42, "epoch_pct": 29.42, "eta": "47:50:14", "max_grad_norm": 0.8, "loss": 0.6185593605041504, "grad_norm": 1.0686509609222412, "learning_rate": 8.72187883130676e-05}
+{"ts": "2025-12-24T10:20:05", "event": "train_log", "step": 4094, "epoch": 1.7663430420711974, "progress_pct": 29.44, "epoch_pct": 29.44, "eta": "47:48:44", "max_grad_norm": 0.8, "loss": 0.7149755358695984, "grad_norm": 1.3372819423675537, "learning_rate": 8.720238810149109e-05}
+{"ts": "2025-12-24T10:20:18", "event": "train_log", "step": 4096, "epoch": 1.7672060409924488, "progress_pct": 29.45, "epoch_pct": 29.45, "eta": "47:47:14", "max_grad_norm": 0.8, "loss": 0.7330698370933533, "grad_norm": 1.2540380954742432, "learning_rate": 8.718597891872638e-05}
+{"ts": "2025-12-24T10:20:30", "event": "train_log", "step": 4098, "epoch": 1.7680690399137, "progress_pct": 29.47, "epoch_pct": 29.47, "eta": "47:45:44", "max_grad_norm": 0.8, "loss": 0.6567472815513611, "grad_norm": 1.1653515100479126, "learning_rate": 8.716956076873049e-05}
+{"ts": "2025-12-24T10:20:42", "event": "train_log", "step": 4100, "epoch": 1.7689320388349514, "progress_pct": 29.48, "epoch_pct": 29.48, "eta": "47:44:15", "max_grad_norm": 0.8, "loss": 0.612649142742157, "grad_norm": 1.0333259105682373, "learning_rate": 8.715313365546258e-05}
+{"ts": "2025-12-24T10:29:00", "event": "train_log", "step": 4100, "epoch": 1.7689320388349514, "progress_pct": 29.48, "epoch_pct": 29.48, "eta": "48:04:06", "max_grad_norm": 0.8, "eval_loss": 0.6929343938827515, "eval_runtime": 497.805, "eval_samples_per_second": 4.138, "eval_steps_per_second": 4.138}
+{"ts": "2025-12-24T10:29:13", "event": "train_log", "step": 4102, "epoch": 1.7697950377562028, "progress_pct": 29.49, "epoch_pct": 29.5, "eta": "48:02:37", "max_grad_norm": 0.8, "loss": 0.6242305636405945, "grad_norm": 1.1154836416244507, "learning_rate": 8.713669758288398e-05}
+{"ts": "2025-12-24T10:29:25", "event": "train_log", "step": 4104, "epoch": 1.770658036677454, "progress_pct": 29.51, "epoch_pct": 29.51, "eta": "48:01:06", "max_grad_norm": 0.8, "loss": 0.6430796384811401, "grad_norm": 1.2573596239089966, "learning_rate": 8.712025255495817e-05}
+{"ts": "2025-12-24T10:29:37", "event": "train_log", "step": 4106, "epoch": 1.7715210355987057, "progress_pct": 29.52, "epoch_pct": 29.53, "eta": "47:59:36", "max_grad_norm": 0.8, "loss": 0.6321606040000916, "grad_norm": 1.407214641571045, "learning_rate": 8.71037985756508e-05}
+{"ts": "2025-12-24T10:29:49", "event": "train_log", "step": 4108, "epoch": 1.7723840345199569, "progress_pct": 29.54, "epoch_pct": 29.54, "eta": "47:58:05", "max_grad_norm": 0.8, "loss": 0.6277352571487427, "grad_norm": 1.1721560955047607, "learning_rate": 8.708733564892969e-05}
+{"ts": "2025-12-24T10:30:01", "event": "train_log", "step": 4110, "epoch": 1.773247033441208, "progress_pct": 29.55, "epoch_pct": 29.55, "eta": "47:56:34", "max_grad_norm": 0.8, "loss": 0.6275392770767212, "grad_norm": 1.0967402458190918, "learning_rate": 8.707086377876475e-05}
+{"ts": "2025-12-24T10:30:14", "event": "train_log", "step": 4112, "epoch": 1.7741100323624597, "progress_pct": 29.57, "epoch_pct": 29.57, "eta": "47:55:06", "max_grad_norm": 0.8, "loss": 0.5985271334648132, "grad_norm": 1.0293948650360107, "learning_rate": 8.705438296912813e-05}
+{"ts": "2025-12-24T10:30:26", "event": "train_log", "step": 4114, "epoch": 1.7749730312837109, "progress_pct": 29.58, "epoch_pct": 29.58, "eta": "47:53:35", "max_grad_norm": 0.8, "loss": 0.6175792217254639, "grad_norm": 1.1681201457977295, "learning_rate": 8.703789322399409e-05}
+{"ts": "2025-12-24T10:30:38", "event": "train_log", "step": 4116, "epoch": 1.7758360302049623, "progress_pct": 29.59, "epoch_pct": 29.6, "eta": "47:52:05", "max_grad_norm": 0.8, "loss": 0.6355584859848022, "grad_norm": 1.3726415634155273, "learning_rate": 8.702139454733907e-05}
+{"ts": "2025-12-24T10:30:50", "event": "train_log", "step": 4118, "epoch": 1.7766990291262137, "progress_pct": 29.61, "epoch_pct": 29.61, "eta": "47:50:35", "max_grad_norm": 0.8, "loss": 0.5877590179443359, "grad_norm": 1.1292396783828735, "learning_rate": 8.700488694314166e-05}
+{"ts": "2025-12-24T10:31:03", "event": "train_log", "step": 4120, "epoch": 1.7775620280474649, "progress_pct": 29.62, "epoch_pct": 29.63, "eta": "47:49:06", "max_grad_norm": 0.8, "loss": 0.6119363307952881, "grad_norm": 1.086949348449707, "learning_rate": 8.698837041538259e-05}
+{"ts": "2025-12-24T10:31:16", "event": "train_log", "step": 4122, "epoch": 1.7784250269687163, "progress_pct": 29.64, "epoch_pct": 29.64, "eta": "47:47:37", "max_grad_norm": 0.8, "loss": 0.7157787084579468, "grad_norm": 1.0434467792510986, "learning_rate": 8.697184496804472e-05}
+{"ts": "2025-12-24T10:31:27", "event": "train_log", "step": 4124, "epoch": 1.7792880258899677, "progress_pct": 29.65, "epoch_pct": 29.65, "eta": "47:46:07", "max_grad_norm": 0.8, "loss": 0.6830587983131409, "grad_norm": 1.0321751832962036, "learning_rate": 8.69553106051131e-05}
+{"ts": "2025-12-24T10:31:39", "event": "train_log", "step": 4126, "epoch": 1.780151024811219, "progress_pct": 29.67, "epoch_pct": 29.67, "eta": "47:44:37", "max_grad_norm": 0.8, "loss": 0.7015431523323059, "grad_norm": 1.3480322360992432, "learning_rate": 8.693876733057493e-05}
+{"ts": "2025-12-24T10:31:53", "event": "train_log", "step": 4128, "epoch": 1.7810140237324703, "progress_pct": 29.68, "epoch_pct": 29.68, "eta": "47:43:10", "max_grad_norm": 0.8, "loss": 0.5946303606033325, "grad_norm": 0.9731265306472778, "learning_rate": 8.692221514841957e-05}
+{"ts": "2025-12-24T10:32:05", "event": "train_log", "step": 4130, "epoch": 1.7818770226537217, "progress_pct": 29.7, "epoch_pct": 29.7, "eta": "47:41:41", "max_grad_norm": 0.8, "loss": 0.687828779220581, "grad_norm": 1.1613185405731201, "learning_rate": 8.690565406263844e-05}
+{"ts": "2025-12-24T10:32:17", "event": "train_log", "step": 4132, "epoch": 1.782740021574973, "progress_pct": 29.71, "epoch_pct": 29.71, "eta": "47:40:11", "max_grad_norm": 0.8, "loss": 0.6315785646438599, "grad_norm": 1.159293532371521, "learning_rate": 8.68890840772252e-05}
+{"ts": "2025-12-24T10:32:29", "event": "train_log", "step": 4134, "epoch": 1.7836030204962245, "progress_pct": 29.72, "epoch_pct": 29.73, "eta": "47:38:42", "max_grad_norm": 0.8, "loss": 0.6761205196380615, "grad_norm": 1.2591674327850342, "learning_rate": 8.687250519617564e-05}
+{"ts": "2025-12-24T10:32:42", "event": "train_log", "step": 4136, "epoch": 1.7844660194174757, "progress_pct": 29.74, "epoch_pct": 29.74, "eta": "47:37:14", "max_grad_norm": 0.8, "loss": 0.6714316010475159, "grad_norm": 1.082155466079712, "learning_rate": 8.685591742348769e-05}
+{"ts": "2025-12-24T10:32:54", "event": "train_log", "step": 4138, "epoch": 1.785329018338727, "progress_pct": 29.75, "epoch_pct": 29.76, "eta": "47:35:43", "max_grad_norm": 0.8, "loss": 0.7500089406967163, "grad_norm": 1.3653912544250488, "learning_rate": 8.68393207631614e-05}
+{"ts": "2025-12-24T10:33:06", "event": "train_log", "step": 4140, "epoch": 1.7861920172599786, "progress_pct": 29.77, "epoch_pct": 29.77, "eta": "47:34:15", "max_grad_norm": 0.8, "loss": 0.6818236708641052, "grad_norm": 1.124690055847168, "learning_rate": 8.682271521919896e-05}
+{"ts": "2025-12-24T10:33:19", "event": "train_log", "step": 4142, "epoch": 1.7870550161812297, "progress_pct": 29.78, "epoch_pct": 29.78, "eta": "47:32:48", "max_grad_norm": 0.8, "loss": 0.5876757502555847, "grad_norm": 1.0788345336914062, "learning_rate": 8.680610079560474e-05}
+{"ts": "2025-12-24T10:33:31", "event": "train_log", "step": 4144, "epoch": 1.7879180151024812, "progress_pct": 29.8, "epoch_pct": 29.8, "eta": "47:31:18", "max_grad_norm": 0.8, "loss": 0.7281823754310608, "grad_norm": 1.0923465490341187, "learning_rate": 8.678947749638526e-05}
+{"ts": "2025-12-24T10:33:43", "event": "train_log", "step": 4146, "epoch": 1.7887810140237326, "progress_pct": 29.81, "epoch_pct": 29.81, "eta": "47:29:50", "max_grad_norm": 0.8, "loss": 0.70050048828125, "grad_norm": 1.0655275583267212, "learning_rate": 8.67728453255491e-05}
+{"ts": "2025-12-24T10:33:56", "event": "train_log", "step": 4148, "epoch": 1.7896440129449838, "progress_pct": 29.82, "epoch_pct": 29.83, "eta": "47:28:22", "max_grad_norm": 0.8, "loss": 0.6276566386222839, "grad_norm": 1.2138395309448242, "learning_rate": 8.67562042871071e-05}
+{"ts": "2025-12-24T10:34:09", "event": "train_log", "step": 4150, "epoch": 1.7905070118662352, "progress_pct": 29.84, "epoch_pct": 29.84, "eta": "47:26:55", "max_grad_norm": 0.8, "loss": 0.5959358811378479, "grad_norm": 1.1036893129348755, "learning_rate": 8.673955438507211e-05}
+{"ts": "2025-12-24T10:34:22", "event": "train_log", "step": 4152, "epoch": 1.7913700107874866, "progress_pct": 29.85, "epoch_pct": 29.86, "eta": "47:25:28", "max_grad_norm": 0.8, "loss": 0.6574703454971313, "grad_norm": 1.0354493856430054, "learning_rate": 8.67228956234592e-05}
+{"ts": "2025-12-24T10:34:35", "event": "train_log", "step": 4154, "epoch": 1.7922330097087378, "progress_pct": 29.87, "epoch_pct": 29.87, "eta": "47:24:01", "max_grad_norm": 0.8, "loss": 0.6122542023658752, "grad_norm": 0.985424280166626, "learning_rate": 8.670622800628556e-05}
+{"ts": "2025-12-24T10:34:48", "event": "train_log", "step": 4156, "epoch": 1.7930960086299892, "progress_pct": 29.88, "epoch_pct": 29.88, "eta": "47:22:35", "max_grad_norm": 0.8, "loss": 0.6271086931228638, "grad_norm": 1.1953589916229248, "learning_rate": 8.668955153757051e-05}
+{"ts": "2025-12-24T10:35:00", "event": "train_log", "step": 4158, "epoch": 1.7939590075512406, "progress_pct": 29.9, "epoch_pct": 29.9, "eta": "47:21:07", "max_grad_norm": 0.8, "loss": 0.6667966246604919, "grad_norm": 1.1208388805389404, "learning_rate": 8.66728662213355e-05}
+{"ts": "2025-12-24T10:35:12", "event": "train_log", "step": 4160, "epoch": 1.7948220064724918, "progress_pct": 29.91, "epoch_pct": 29.91, "eta": "47:19:37", "max_grad_norm": 0.8, "loss": 0.6663333177566528, "grad_norm": 1.1992945671081543, "learning_rate": 8.665617206160415e-05}
+{"ts": "2025-12-24T10:35:23", "event": "train_log", "step": 4162, "epoch": 1.7956850053937432, "progress_pct": 29.93, "epoch_pct": 29.93, "eta": "47:18:08", "max_grad_norm": 0.8, "loss": 0.7308130264282227, "grad_norm": 1.2541922330856323, "learning_rate": 8.663946906240213e-05}
+{"ts": "2025-12-24T10:35:36", "event": "train_log", "step": 4164, "epoch": 1.7965480043149946, "progress_pct": 29.94, "epoch_pct": 29.94, "eta": "47:16:42", "max_grad_norm": 0.8, "loss": 0.5975028276443481, "grad_norm": 1.1797630786895752, "learning_rate": 8.662275722775733e-05}
+{"ts": "2025-12-24T10:35:50", "event": "train_log", "step": 4166, "epoch": 1.7974110032362458, "progress_pct": 29.95, "epoch_pct": 29.96, "eta": "47:15:16", "max_grad_norm": 0.8, "loss": 0.6519654393196106, "grad_norm": 1.0160104036331177, "learning_rate": 8.66060365616997e-05}
+{"ts": "2025-12-24T10:36:03", "event": "train_log", "step": 4168, "epoch": 1.7982740021574974, "progress_pct": 29.97, "epoch_pct": 29.97, "eta": "47:13:51", "max_grad_norm": 0.8, "loss": 0.5663926601409912, "grad_norm": 0.9538162350654602, "learning_rate": 8.658930706826138e-05}
+{"ts": "2025-12-24T10:36:16", "event": "train_log", "step": 4170, "epoch": 1.7991370010787486, "progress_pct": 29.98, "epoch_pct": 29.99, "eta": "47:12:25", "max_grad_norm": 0.8, "loss": 0.6580140590667725, "grad_norm": 1.2381815910339355, "learning_rate": 8.657256875147662e-05}
+{"ts": "2025-12-24T10:36:30", "event": "train_log", "step": 4172, "epoch": 1.8, "progress_pct": 30.0, "epoch_pct": 30.0, "eta": "47:11:01", "max_grad_norm": 0.8, "loss": 0.5924336314201355, "grad_norm": 1.1187360286712646, "learning_rate": 8.655582161538177e-05}
+{"ts": "2025-12-24T10:36:42", "event": "train_log", "step": 4174, "epoch": 1.8008629989212515, "progress_pct": 30.01, "epoch_pct": 30.01, "eta": "47:09:32", "max_grad_norm": 0.8, "loss": 0.6451319456100464, "grad_norm": 1.1784483194351196, "learning_rate": 8.653906566401533e-05}
+{"ts": "2025-12-24T10:36:55", "event": "train_log", "step": 4176, "epoch": 1.8017259978425026, "progress_pct": 30.03, "epoch_pct": 30.03, "eta": "47:08:06", "max_grad_norm": 0.8, "loss": 0.6331294775009155, "grad_norm": 1.1566145420074463, "learning_rate": 8.652230090141791e-05}
+{"ts": "2025-12-24T10:37:07", "event": "train_log", "step": 4178, "epoch": 1.802588996763754, "progress_pct": 30.04, "epoch_pct": 30.04, "eta": "47:06:37", "max_grad_norm": 0.8, "loss": 0.6497499346733093, "grad_norm": 1.2551170587539673, "learning_rate": 8.65055273316323e-05}
+{"ts": "2025-12-24T10:37:19", "event": "train_log", "step": 4180, "epoch": 1.8034519956850055, "progress_pct": 30.05, "epoch_pct": 30.06, "eta": "47:05:10", "max_grad_norm": 0.8, "loss": 0.6330156326293945, "grad_norm": 1.0211492776870728, "learning_rate": 8.648874495870332e-05}
+{"ts": "2025-12-24T10:37:31", "event": "train_log", "step": 4182, "epoch": 1.8043149946062567, "progress_pct": 30.07, "epoch_pct": 30.07, "eta": "47:03:43", "max_grad_norm": 0.8, "loss": 0.6429944634437561, "grad_norm": 1.0939637422561646, "learning_rate": 8.6471953786678e-05}
+{"ts": "2025-12-24T10:37:43", "event": "train_log", "step": 4184, "epoch": 1.805177993527508, "progress_pct": 30.08, "epoch_pct": 30.09, "eta": "47:02:15", "max_grad_norm": 0.8, "loss": 0.7175302505493164, "grad_norm": 1.2167190313339233, "learning_rate": 8.645515381960543e-05}
+{"ts": "2025-12-24T10:37:55", "event": "train_log", "step": 4186, "epoch": 1.8060409924487595, "progress_pct": 30.1, "epoch_pct": 30.1, "eta": "47:00:46", "max_grad_norm": 0.8, "loss": 0.6182355284690857, "grad_norm": 1.2631546258926392, "learning_rate": 8.643834506153685e-05}
+{"ts": "2025-12-24T10:38:07", "event": "train_log", "step": 4188, "epoch": 1.8069039913700107, "progress_pct": 30.11, "epoch_pct": 30.12, "eta": "46:59:19", "max_grad_norm": 0.8, "loss": 0.6001572012901306, "grad_norm": 1.2085437774658203, "learning_rate": 8.642152751652564e-05}
+{"ts": "2025-12-24T10:38:20", "event": "train_log", "step": 4190, "epoch": 1.807766990291262, "progress_pct": 30.13, "epoch_pct": 30.13, "eta": "46:57:52", "max_grad_norm": 0.8, "loss": 0.6273662447929382, "grad_norm": 1.1557323932647705, "learning_rate": 8.640470118862726e-05}
+{"ts": "2025-12-24T10:38:32", "event": "train_log", "step": 4192, "epoch": 1.8086299892125135, "progress_pct": 30.14, "epoch_pct": 30.14, "eta": "46:56:26", "max_grad_norm": 0.8, "loss": 0.5653483867645264, "grad_norm": 1.0053116083145142, "learning_rate": 8.638786608189928e-05}
+{"ts": "2025-12-24T10:38:44", "event": "train_log", "step": 4194, "epoch": 1.8094929881337647, "progress_pct": 30.16, "epoch_pct": 30.16, "eta": "46:54:58", "max_grad_norm": 0.8, "loss": 0.7450010776519775, "grad_norm": 1.135331630706787, "learning_rate": 8.637102220040145e-05}
+{"ts": "2025-12-24T10:38:56", "event": "train_log", "step": 4196, "epoch": 1.8103559870550163, "progress_pct": 30.17, "epoch_pct": 30.17, "eta": "46:53:30", "max_grad_norm": 0.8, "loss": 0.6311311721801758, "grad_norm": 1.1989952325820923, "learning_rate": 8.635416954819557e-05}
+{"ts": "2025-12-24T10:39:08", "event": "train_log", "step": 4198, "epoch": 1.8112189859762675, "progress_pct": 30.18, "epoch_pct": 30.19, "eta": "46:52:02", "max_grad_norm": 0.8, "loss": 0.6941008567810059, "grad_norm": 1.3958007097244263, "learning_rate": 8.633730812934559e-05}
+{"ts": "2025-12-24T10:39:20", "event": "train_log", "step": 4200, "epoch": 1.812081984897519, "progress_pct": 30.2, "epoch_pct": 30.2, "eta": "46:50:35", "max_grad_norm": 0.8, "loss": 0.729922890663147, "grad_norm": 1.135804533958435, "learning_rate": 8.632043794791758e-05}
+{"ts": "2025-12-24T10:47:38", "event": "train_log", "step": 4200, "epoch": 1.812081984897519, "progress_pct": 30.2, "epoch_pct": 30.2, "eta": "47:09:48", "max_grad_norm": 0.8, "eval_loss": 0.6918642520904541, "eval_runtime": 498.4981, "eval_samples_per_second": 4.132, "eval_steps_per_second": 4.132}
+{"ts": "2025-12-24T10:47:51", "event": "train_log", "step": 4202, "epoch": 1.8129449838187703, "progress_pct": 30.21, "epoch_pct": 30.22, "eta": "47:08:21", "max_grad_norm": 0.8, "loss": 0.5869379043579102, "grad_norm": 1.082674503326416, "learning_rate": 8.630355900797967e-05}
+{"ts": "2025-12-24T10:48:04", "event": "train_log", "step": 4204, "epoch": 1.8138079827400215, "progress_pct": 30.23, "epoch_pct": 30.23, "eta": "47:06:57", "max_grad_norm": 0.8, "loss": 0.6634620428085327, "grad_norm": 1.290963888168335, "learning_rate": 8.628667131360218e-05}
+{"ts": "2025-12-24T10:48:16", "event": "train_log", "step": 4206, "epoch": 1.814670981661273, "progress_pct": 30.24, "epoch_pct": 30.24, "eta": "47:05:29", "max_grad_norm": 0.8, "loss": 0.6378920674324036, "grad_norm": 1.1635067462921143, "learning_rate": 8.626977486885747e-05}
+{"ts": "2025-12-24T10:48:29", "event": "train_log", "step": 4208, "epoch": 1.8155339805825244, "progress_pct": 30.26, "epoch_pct": 30.26, "eta": "47:04:02", "max_grad_norm": 0.8, "loss": 0.6387537121772766, "grad_norm": 1.5235686302185059, "learning_rate": 8.625286967782006e-05}
+{"ts": "2025-12-24T10:48:41", "event": "train_log", "step": 4210, "epoch": 1.8163969795037755, "progress_pct": 30.27, "epoch_pct": 30.27, "eta": "47:02:35", "max_grad_norm": 0.8, "loss": 0.7544008493423462, "grad_norm": 1.1585710048675537, "learning_rate": 8.623595574456655e-05}
+{"ts": "2025-12-24T10:48:53", "event": "train_log", "step": 4212, "epoch": 1.817259978425027, "progress_pct": 30.28, "epoch_pct": 30.29, "eta": "47:01:08", "max_grad_norm": 0.8, "loss": 0.6456180810928345, "grad_norm": 1.0860902070999146, "learning_rate": 8.621903307317566e-05}
+{"ts": "2025-12-24T10:49:06", "event": "train_log", "step": 4214, "epoch": 1.8181229773462784, "progress_pct": 30.3, "epoch_pct": 30.3, "eta": "46:59:41", "max_grad_norm": 0.8, "loss": 0.6784940361976624, "grad_norm": 1.0690677165985107, "learning_rate": 8.620210166772824e-05}
+{"ts": "2025-12-24T10:49:18", "event": "train_log", "step": 4216, "epoch": 1.8189859762675296, "progress_pct": 30.31, "epoch_pct": 30.32, "eta": "46:58:13", "max_grad_norm": 0.8, "loss": 0.725097119808197, "grad_norm": 1.3315730094909668, "learning_rate": 8.61851615323072e-05}
+{"ts": "2025-12-24T10:49:30", "event": "train_log", "step": 4218, "epoch": 1.819848975188781, "progress_pct": 30.33, "epoch_pct": 30.33, "eta": "46:56:46", "max_grad_norm": 0.8, "loss": 0.6593776345252991, "grad_norm": 1.3689441680908203, "learning_rate": 8.616821267099757e-05}
+{"ts": "2025-12-24T10:49:42", "event": "train_log", "step": 4220, "epoch": 1.8207119741100324, "progress_pct": 30.34, "epoch_pct": 30.35, "eta": "46:55:19", "max_grad_norm": 0.8, "loss": 0.6334517598152161, "grad_norm": 1.0977379083633423, "learning_rate": 8.61512550878865e-05}
+{"ts": "2025-12-24T10:49:54", "event": "train_log", "step": 4222, "epoch": 1.8215749730312836, "progress_pct": 30.36, "epoch_pct": 30.36, "eta": "46:53:51", "max_grad_norm": 0.8, "loss": 0.7029228210449219, "grad_norm": 1.2982810735702515, "learning_rate": 8.613428878706326e-05}
+{"ts": "2025-12-24T10:50:06", "event": "train_log", "step": 4224, "epoch": 1.8224379719525352, "progress_pct": 30.37, "epoch_pct": 30.37, "eta": "46:52:25", "max_grad_norm": 0.8, "loss": 0.6502435207366943, "grad_norm": 1.117529273033142, "learning_rate": 8.611731377261916e-05}
+{"ts": "2025-12-24T10:50:18", "event": "train_log", "step": 4226, "epoch": 1.8233009708737864, "progress_pct": 30.39, "epoch_pct": 30.39, "eta": "46:50:58", "max_grad_norm": 0.8, "loss": 0.6504095196723938, "grad_norm": 1.0791085958480835, "learning_rate": 8.610033004864765e-05}
+{"ts": "2025-12-24T10:50:31", "event": "train_log", "step": 4228, "epoch": 1.8241639697950378, "progress_pct": 30.4, "epoch_pct": 30.4, "eta": "46:49:32", "max_grad_norm": 0.8, "loss": 0.6614294052124023, "grad_norm": 1.017614483833313, "learning_rate": 8.60833376192443e-05}
+{"ts": "2025-12-24T10:50:43", "event": "train_log", "step": 4230, "epoch": 1.8250269687162892, "progress_pct": 30.41, "epoch_pct": 30.42, "eta": "46:48:06", "max_grad_norm": 0.8, "loss": 0.6149688959121704, "grad_norm": 1.0131317377090454, "learning_rate": 8.606633648850673e-05}
+{"ts": "2025-12-24T10:50:56", "event": "train_log", "step": 4232, "epoch": 1.8258899676375404, "progress_pct": 30.43, "epoch_pct": 30.43, "eta": "46:46:40", "max_grad_norm": 0.8, "loss": 0.5616279244422913, "grad_norm": 1.0489189624786377, "learning_rate": 8.604932666053474e-05}
+{"ts": "2025-12-24T10:51:07", "event": "train_log", "step": 4234, "epoch": 1.8267529665587918, "progress_pct": 30.44, "epoch_pct": 30.45, "eta": "46:45:12", "max_grad_norm": 0.8, "loss": 0.6322590112686157, "grad_norm": 1.210129976272583, "learning_rate": 8.60323081394301e-05}
+{"ts": "2025-12-24T10:51:20", "event": "train_log", "step": 4236, "epoch": 1.8276159654800432, "progress_pct": 30.46, "epoch_pct": 30.46, "eta": "46:43:48", "max_grad_norm": 0.8, "loss": 0.6540767550468445, "grad_norm": 1.2635412216186523, "learning_rate": 8.60152809292968e-05}
+{"ts": "2025-12-24T10:51:32", "event": "train_log", "step": 4238, "epoch": 1.8284789644012944, "progress_pct": 30.47, "epoch_pct": 30.47, "eta": "46:42:21", "max_grad_norm": 0.8, "loss": 0.6719615459442139, "grad_norm": 1.1350682973861694, "learning_rate": 8.599824503424086e-05}
+{"ts": "2025-12-24T10:51:44", "event": "train_log", "step": 4240, "epoch": 1.8293419633225458, "progress_pct": 30.49, "epoch_pct": 30.49, "eta": "46:40:54", "max_grad_norm": 0.8, "loss": 0.737215518951416, "grad_norm": 1.2475734949111938, "learning_rate": 8.598120045837043e-05}
+{"ts": "2025-12-24T10:51:57", "event": "train_log", "step": 4242, "epoch": 1.8302049622437973, "progress_pct": 30.5, "epoch_pct": 30.5, "eta": "46:39:28", "max_grad_norm": 0.8, "loss": 0.6711077094078064, "grad_norm": 1.109307885169983, "learning_rate": 8.596414720579568e-05}
+{"ts": "2025-12-24T10:52:09", "event": "train_log", "step": 4244, "epoch": 1.8310679611650484, "progress_pct": 30.51, "epoch_pct": 30.52, "eta": "46:38:02", "max_grad_norm": 0.8, "loss": 0.6282169222831726, "grad_norm": 1.3011112213134766, "learning_rate": 8.594708528062898e-05}
+{"ts": "2025-12-24T10:52:22", "event": "train_log", "step": 4246, "epoch": 1.8319309600862999, "progress_pct": 30.53, "epoch_pct": 30.53, "eta": "46:36:38", "max_grad_norm": 0.8, "loss": 0.6156499981880188, "grad_norm": 0.9670302271842957, "learning_rate": 8.593001468698471e-05}
+{"ts": "2025-12-24T10:52:35", "event": "train_log", "step": 4248, "epoch": 1.8327939590075513, "progress_pct": 30.54, "epoch_pct": 30.55, "eta": "46:35:13", "max_grad_norm": 0.8, "loss": 0.6415660977363586, "grad_norm": 1.1828471422195435, "learning_rate": 8.591293542897938e-05}
+{"ts": "2025-12-24T10:52:48", "event": "train_log", "step": 4250, "epoch": 1.8336569579288025, "progress_pct": 30.56, "epoch_pct": 30.56, "eta": "46:33:50", "max_grad_norm": 0.8, "loss": 0.551155149936676, "grad_norm": 0.9576874375343323, "learning_rate": 8.589584751073156e-05}
+{"ts": "2025-12-24T10:53:00", "event": "train_log", "step": 4252, "epoch": 1.834519956850054, "progress_pct": 30.57, "epoch_pct": 30.58, "eta": "46:32:25", "max_grad_norm": 0.8, "loss": 0.6177648305892944, "grad_norm": 1.163041591644287, "learning_rate": 8.587875093636193e-05}
+{"ts": "2025-12-24T10:53:14", "event": "train_log", "step": 4254, "epoch": 1.8353829557713053, "progress_pct": 30.59, "epoch_pct": 30.59, "eta": "46:31:01", "max_grad_norm": 0.8, "loss": 0.5701619386672974, "grad_norm": 0.9363775849342346, "learning_rate": 8.586164570999327e-05}
+{"ts": "2025-12-24T10:53:26", "event": "train_log", "step": 4256, "epoch": 1.8362459546925565, "progress_pct": 30.6, "epoch_pct": 30.6, "eta": "46:29:37", "max_grad_norm": 0.8, "loss": 0.6038158535957336, "grad_norm": 1.1433947086334229, "learning_rate": 8.584453183575038e-05}
+{"ts": "2025-12-24T10:53:38", "event": "train_log", "step": 4258, "epoch": 1.837108953613808, "progress_pct": 30.62, "epoch_pct": 30.62, "eta": "46:28:10", "max_grad_norm": 0.8, "loss": 0.6887767314910889, "grad_norm": 1.1908625364303589, "learning_rate": 8.582740931776026e-05}
+{"ts": "2025-12-24T10:53:50", "event": "train_log", "step": 4260, "epoch": 1.8379719525350593, "progress_pct": 30.63, "epoch_pct": 30.63, "eta": "46:26:44", "max_grad_norm": 0.8, "loss": 0.6212680339813232, "grad_norm": 1.0207780599594116, "learning_rate": 8.581027816015187e-05}
+{"ts": "2025-12-24T10:54:02", "event": "train_log", "step": 4262, "epoch": 1.8388349514563107, "progress_pct": 30.64, "epoch_pct": 30.65, "eta": "46:25:17", "max_grad_norm": 0.8, "loss": 0.6180729866027832, "grad_norm": 1.213992953300476, "learning_rate": 8.579313836705635e-05}
+{"ts": "2025-12-24T10:54:14", "event": "train_log", "step": 4264, "epoch": 1.8396979503775621, "progress_pct": 30.66, "epoch_pct": 30.66, "eta": "46:23:52", "max_grad_norm": 0.8, "loss": 0.6168033480644226, "grad_norm": 1.1175587177276611, "learning_rate": 8.577598994260688e-05}
+{"ts": "2025-12-24T10:54:27", "event": "train_log", "step": 4266, "epoch": 1.8405609492988133, "progress_pct": 30.67, "epoch_pct": 30.68, "eta": "46:22:29", "max_grad_norm": 0.8, "loss": 0.5986083149909973, "grad_norm": 0.9955618977546692, "learning_rate": 8.575883289093869e-05}
+{"ts": "2025-12-24T10:54:39", "event": "train_log", "step": 4268, "epoch": 1.8414239482200647, "progress_pct": 30.69, "epoch_pct": 30.69, "eta": "46:21:04", "max_grad_norm": 0.8, "loss": 0.6276304721832275, "grad_norm": 1.0735031366348267, "learning_rate": 8.574166721618915e-05}
+{"ts": "2025-12-24T10:54:52", "event": "train_log", "step": 4270, "epoch": 1.8422869471413161, "progress_pct": 30.7, "epoch_pct": 30.7, "eta": "46:19:39", "max_grad_norm": 0.8, "loss": 0.6733853220939636, "grad_norm": 1.1660799980163574, "learning_rate": 8.57244929224977e-05}
+{"ts": "2025-12-24T10:55:04", "event": "train_log", "step": 4272, "epoch": 1.8431499460625673, "progress_pct": 30.72, "epoch_pct": 30.72, "eta": "46:18:13", "max_grad_norm": 0.8, "loss": 0.6379335522651672, "grad_norm": 1.149767279624939, "learning_rate": 8.570731001400581e-05}
+{"ts": "2025-12-24T10:55:16", "event": "train_log", "step": 4274, "epoch": 1.8440129449838187, "progress_pct": 30.73, "epoch_pct": 30.73, "eta": "46:16:48", "max_grad_norm": 0.8, "loss": 0.6092600226402283, "grad_norm": 1.0226882696151733, "learning_rate": 8.56901184948571e-05}
+{"ts": "2025-12-24T10:55:28", "event": "train_log", "step": 4276, "epoch": 1.8448759439050701, "progress_pct": 30.74, "epoch_pct": 30.75, "eta": "46:15:23", "max_grad_norm": 0.8, "loss": 0.6186777949333191, "grad_norm": 1.116834282875061, "learning_rate": 8.567291836919719e-05}
+{"ts": "2025-12-24T10:55:40", "event": "train_log", "step": 4278, "epoch": 1.8457389428263213, "progress_pct": 30.76, "epoch_pct": 30.76, "eta": "46:13:57", "max_grad_norm": 0.8, "loss": 0.6713102459907532, "grad_norm": 1.219565987586975, "learning_rate": 8.565570964117385e-05}
+{"ts": "2025-12-24T10:55:51", "event": "train_log", "step": 4280, "epoch": 1.846601941747573, "progress_pct": 30.77, "epoch_pct": 30.78, "eta": "46:12:30", "max_grad_norm": 0.8, "loss": 0.6963690519332886, "grad_norm": 1.1785911321640015, "learning_rate": 8.563849231493683e-05}
+{"ts": "2025-12-24T10:56:05", "event": "train_log", "step": 4282, "epoch": 1.8474649406688242, "progress_pct": 30.79, "epoch_pct": 30.79, "eta": "46:11:09", "max_grad_norm": 0.8, "loss": 0.5925922989845276, "grad_norm": 0.9836017489433289, "learning_rate": 8.562126639463808e-05}
+{"ts": "2025-12-24T10:56:18", "event": "train_log", "step": 4284, "epoch": 1.8483279395900754, "progress_pct": 30.8, "epoch_pct": 30.81, "eta": "46:09:46", "max_grad_norm": 0.8, "loss": 0.6510034203529358, "grad_norm": 1.027338981628418, "learning_rate": 8.560403188443151e-05}
+{"ts": "2025-12-24T10:56:30", "event": "train_log", "step": 4286, "epoch": 1.849190938511327, "progress_pct": 30.82, "epoch_pct": 30.82, "eta": "46:08:21", "max_grad_norm": 0.8, "loss": 0.6744941473007202, "grad_norm": 1.0148735046386719, "learning_rate": 8.558678878847317e-05}
+{"ts": "2025-12-24T10:56:43", "event": "train_log", "step": 4288, "epoch": 1.8500539374325782, "progress_pct": 30.83, "epoch_pct": 30.83, "eta": "46:06:59", "max_grad_norm": 0.8, "loss": 0.5461649298667908, "grad_norm": 1.0071876049041748, "learning_rate": 8.556953711092112e-05}
+{"ts": "2025-12-24T10:56:56", "event": "train_log", "step": 4290, "epoch": 1.8509169363538296, "progress_pct": 30.85, "epoch_pct": 30.85, "eta": "46:05:35", "max_grad_norm": 0.8, "loss": 0.650245189666748, "grad_norm": 1.1736774444580078, "learning_rate": 8.555227685593555e-05}
+{"ts": "2025-12-24T10:57:08", "event": "train_log", "step": 4292, "epoch": 1.851779935275081, "progress_pct": 30.86, "epoch_pct": 30.86, "eta": "46:04:10", "max_grad_norm": 0.8, "loss": 0.6824072599411011, "grad_norm": 1.1919746398925781, "learning_rate": 8.553500802767869e-05}
+{"ts": "2025-12-24T10:57:20", "event": "train_log", "step": 4294, "epoch": 1.8526429341963322, "progress_pct": 30.87, "epoch_pct": 30.88, "eta": "46:02:45", "max_grad_norm": 0.8, "loss": 0.6167742609977722, "grad_norm": 1.1280466318130493, "learning_rate": 8.551773063031484e-05}
+{"ts": "2025-12-24T10:57:32", "event": "train_log", "step": 4296, "epoch": 1.8535059331175836, "progress_pct": 30.89, "epoch_pct": 30.89, "eta": "46:01:21", "max_grad_norm": 0.8, "loss": 0.6046481132507324, "grad_norm": 1.32740318775177, "learning_rate": 8.550044466801034e-05}
+{"ts": "2025-12-24T10:57:44", "event": "train_log", "step": 4298, "epoch": 1.854368932038835, "progress_pct": 30.9, "epoch_pct": 30.91, "eta": "45:59:57", "max_grad_norm": 0.8, "loss": 0.6526167988777161, "grad_norm": 1.1019376516342163, "learning_rate": 8.548315014493367e-05}
+{"ts": "2025-12-24T10:57:57", "event": "train_log", "step": 4300, "epoch": 1.8552319309600862, "progress_pct": 30.92, "epoch_pct": 30.92, "eta": "45:58:33", "max_grad_norm": 0.8, "loss": 0.6063379645347595, "grad_norm": 1.1523452997207642, "learning_rate": 8.54658470652553e-05}
+{"ts": "2025-12-24T11:06:16", "event": "train_log", "step": 4300, "epoch": 1.8552319309600862, "progress_pct": 30.92, "epoch_pct": 30.92, "eta": "46:17:07", "max_grad_norm": 0.8, "eval_loss": 0.6890321373939514, "eval_runtime": 498.5289, "eval_samples_per_second": 4.132, "eval_steps_per_second": 4.132}
+{"ts": "2025-12-24T11:06:28", "event": "train_log", "step": 4302, "epoch": 1.8560949298813376, "progress_pct": 30.93, "epoch_pct": 30.93, "eta": "46:15:44", "max_grad_norm": 0.8, "loss": 0.6220276951789856, "grad_norm": 1.1315364837646484, "learning_rate": 8.544853543314778e-05}
+{"ts": "2025-12-24T11:06:40", "event": "train_log", "step": 4304, "epoch": 1.856957928802589, "progress_pct": 30.95, "epoch_pct": 30.95, "eta": "46:14:18", "max_grad_norm": 0.8, "loss": 0.6813952326774597, "grad_norm": 1.2688732147216797, "learning_rate": 8.543121525278575e-05}
+{"ts": "2025-12-24T11:06:52", "event": "train_log", "step": 4306, "epoch": 1.8578209277238402, "progress_pct": 30.96, "epoch_pct": 30.96, "eta": "46:12:53", "max_grad_norm": 0.8, "loss": 0.7254486083984375, "grad_norm": 1.2689135074615479, "learning_rate": 8.541388652834588e-05}
+{"ts": "2025-12-24T11:07:04", "event": "train_log", "step": 4308, "epoch": 1.8586839266450919, "progress_pct": 30.97, "epoch_pct": 30.98, "eta": "46:11:29", "max_grad_norm": 0.8, "loss": 0.6455208659172058, "grad_norm": 1.198376178741455, "learning_rate": 8.539654926400692e-05}
+{"ts": "2025-12-24T11:07:18", "event": "train_log", "step": 4310, "epoch": 1.859546925566343, "progress_pct": 30.99, "epoch_pct": 30.99, "eta": "46:10:07", "max_grad_norm": 0.8, "loss": 0.6356139779090881, "grad_norm": 1.0240654945373535, "learning_rate": 8.537920346394969e-05}
+{"ts": "2025-12-24T11:07:30", "event": "train_log", "step": 4312, "epoch": 1.8604099244875942, "progress_pct": 31.0, "epoch_pct": 31.01, "eta": "46:08:42", "max_grad_norm": 0.8, "loss": 0.6543931365013123, "grad_norm": 1.1928554773330688, "learning_rate": 8.536184913235704e-05}
+{"ts": "2025-12-24T11:07:42", "event": "train_log", "step": 4314, "epoch": 1.8612729234088459, "progress_pct": 31.02, "epoch_pct": 31.02, "eta": "46:07:16", "max_grad_norm": 0.8, "loss": 0.7181617021560669, "grad_norm": 1.190274953842163, "learning_rate": 8.534448627341384e-05}
+{"ts": "2025-12-24T11:07:54", "event": "train_log", "step": 4316, "epoch": 1.862135922330097, "progress_pct": 31.03, "epoch_pct": 31.04, "eta": "46:05:52", "max_grad_norm": 0.8, "loss": 0.661147952079773, "grad_norm": 1.2967594861984253, "learning_rate": 8.532711489130715e-05}
+{"ts": "2025-12-24T11:08:06", "event": "train_log", "step": 4318, "epoch": 1.8629989212513485, "progress_pct": 31.05, "epoch_pct": 31.05, "eta": "46:04:27", "max_grad_norm": 0.8, "loss": 0.6840251684188843, "grad_norm": 1.226694941520691, "learning_rate": 8.530973499022594e-05}
+{"ts": "2025-12-24T11:08:18", "event": "train_log", "step": 4320, "epoch": 1.8638619201725999, "progress_pct": 31.06, "epoch_pct": 31.06, "eta": "46:03:02", "max_grad_norm": 0.8, "loss": 0.6289029121398926, "grad_norm": 1.098263144493103, "learning_rate": 8.529234657436133e-05}
+{"ts": "2025-12-24T11:08:30", "event": "train_log", "step": 4322, "epoch": 1.864724919093851, "progress_pct": 31.08, "epoch_pct": 31.08, "eta": "46:01:38", "max_grad_norm": 0.8, "loss": 0.6778033375740051, "grad_norm": 1.0724968910217285, "learning_rate": 8.527494964790644e-05}
+{"ts": "2025-12-24T11:08:42", "event": "train_log", "step": 4324, "epoch": 1.8655879180151025, "progress_pct": 31.09, "epoch_pct": 31.09, "eta": "46:00:14", "max_grad_norm": 0.8, "loss": 0.6152138710021973, "grad_norm": 1.0636100769042969, "learning_rate": 8.525754421505646e-05}
+{"ts": "2025-12-24T11:08:55", "event": "train_log", "step": 4326, "epoch": 1.866450916936354, "progress_pct": 31.1, "epoch_pct": 31.11, "eta": "45:58:52", "max_grad_norm": 0.8, "loss": 0.6195679306983948, "grad_norm": 1.1964589357376099, "learning_rate": 8.524013028000861e-05}
+{"ts": "2025-12-24T11:09:08", "event": "train_log", "step": 4328, "epoch": 1.867313915857605, "progress_pct": 31.12, "epoch_pct": 31.12, "eta": "45:57:30", "max_grad_norm": 0.8, "loss": 0.5703424215316772, "grad_norm": 0.9388785362243652, "learning_rate": 8.522270784696223e-05}
+{"ts": "2025-12-24T11:09:22", "event": "train_log", "step": 4330, "epoch": 1.8681769147788565, "progress_pct": 31.13, "epoch_pct": 31.14, "eta": "45:56:08", "max_grad_norm": 0.8, "loss": 0.6520850658416748, "grad_norm": 1.0190461874008179, "learning_rate": 8.520527692011864e-05}
+{"ts": "2025-12-24T11:09:34", "event": "train_log", "step": 4332, "epoch": 1.869039913700108, "progress_pct": 31.15, "epoch_pct": 31.15, "eta": "45:54:45", "max_grad_norm": 0.8, "loss": 0.6146131157875061, "grad_norm": 1.2388478517532349, "learning_rate": 8.518783750368123e-05}
+{"ts": "2025-12-24T11:09:46", "event": "train_log", "step": 4334, "epoch": 1.869902912621359, "progress_pct": 31.16, "epoch_pct": 31.17, "eta": "45:53:19", "max_grad_norm": 0.8, "loss": 0.663919985294342, "grad_norm": 1.5210362672805786, "learning_rate": 8.517038960185542e-05}
+{"ts": "2025-12-24T11:09:58", "event": "train_log", "step": 4336, "epoch": 1.8707659115426107, "progress_pct": 31.18, "epoch_pct": 31.18, "eta": "45:51:56", "max_grad_norm": 0.8, "loss": 0.6086801886558533, "grad_norm": 1.0638145208358765, "learning_rate": 8.51529332188487e-05}
+{"ts": "2025-12-24T11:10:10", "event": "train_log", "step": 4338, "epoch": 1.871628910463862, "progress_pct": 31.19, "epoch_pct": 31.19, "eta": "45:50:32", "max_grad_norm": 0.8, "loss": 0.7241880297660828, "grad_norm": 1.1796083450317383, "learning_rate": 8.513546835887061e-05}
+{"ts": "2025-12-24T11:10:22", "event": "train_log", "step": 4340, "epoch": 1.8724919093851131, "progress_pct": 31.21, "epoch_pct": 31.21, "eta": "45:49:08", "max_grad_norm": 0.8, "loss": 0.6798967123031616, "grad_norm": 1.2609213590621948, "learning_rate": 8.511799502613273e-05}
+{"ts": "2025-12-24T11:10:35", "event": "train_log", "step": 4342, "epoch": 1.8733549083063648, "progress_pct": 31.22, "epoch_pct": 31.22, "eta": "45:47:45", "max_grad_norm": 0.8, "loss": 0.6314971446990967, "grad_norm": 1.1063754558563232, "learning_rate": 8.510051322484864e-05}
+{"ts": "2025-12-24T11:10:47", "event": "train_log", "step": 4344, "epoch": 1.874217907227616, "progress_pct": 31.23, "epoch_pct": 31.24, "eta": "45:46:21", "max_grad_norm": 0.8, "loss": 0.6779396533966064, "grad_norm": 1.2268476486206055, "learning_rate": 8.508302295923405e-05}
+{"ts": "2025-12-24T11:10:59", "event": "train_log", "step": 4346, "epoch": 1.8750809061488674, "progress_pct": 31.25, "epoch_pct": 31.25, "eta": "45:44:58", "max_grad_norm": 0.8, "loss": 0.6303188800811768, "grad_norm": 1.254714846611023, "learning_rate": 8.50655242335066e-05}
+{"ts": "2025-12-24T11:11:11", "event": "train_log", "step": 4348, "epoch": 1.8759439050701188, "progress_pct": 31.26, "epoch_pct": 31.27, "eta": "45:43:35", "max_grad_norm": 0.8, "loss": 0.626371443271637, "grad_norm": 1.0341795682907104, "learning_rate": 8.504801705188607e-05}
+{"ts": "2025-12-24T11:11:24", "event": "train_log", "step": 4350, "epoch": 1.87680690399137, "progress_pct": 31.28, "epoch_pct": 31.28, "eta": "45:42:13", "max_grad_norm": 0.8, "loss": 0.6452032327651978, "grad_norm": 1.148277997970581, "learning_rate": 8.503050141859423e-05}
+{"ts": "2025-12-24T11:11:36", "event": "train_log", "step": 4352, "epoch": 1.8776699029126214, "progress_pct": 31.29, "epoch_pct": 31.29, "eta": "45:40:50", "max_grad_norm": 0.8, "loss": 0.7556618452072144, "grad_norm": 1.1761208772659302, "learning_rate": 8.50129773378549e-05}
+{"ts": "2025-12-24T11:11:49", "event": "train_log", "step": 4354, "epoch": 1.8785329018338728, "progress_pct": 31.31, "epoch_pct": 31.31, "eta": "45:39:28", "max_grad_norm": 0.8, "loss": 0.6422954201698303, "grad_norm": 1.2172788381576538, "learning_rate": 8.49954448138939e-05}
+{"ts": "2025-12-24T11:12:02", "event": "train_log", "step": 4356, "epoch": 1.879395900755124, "progress_pct": 31.32, "epoch_pct": 31.32, "eta": "45:38:06", "max_grad_norm": 0.8, "loss": 0.6151976585388184, "grad_norm": 1.11391019821167, "learning_rate": 8.497790385093916e-05}
+{"ts": "2025-12-24T11:12:15", "event": "train_log", "step": 4358, "epoch": 1.8802588996763754, "progress_pct": 31.33, "epoch_pct": 31.34, "eta": "45:36:45", "max_grad_norm": 0.8, "loss": 0.5745062232017517, "grad_norm": 1.060592532157898, "learning_rate": 8.496035445322057e-05}
+{"ts": "2025-12-24T11:12:27", "event": "train_log", "step": 4360, "epoch": 1.8811218985976268, "progress_pct": 31.35, "epoch_pct": 31.35, "eta": "45:35:22", "max_grad_norm": 0.8, "loss": 0.6658170223236084, "grad_norm": 1.2631596326828003, "learning_rate": 8.49427966249701e-05}
+{"ts": "2025-12-24T11:12:40", "event": "train_log", "step": 4362, "epoch": 1.881984897518878, "progress_pct": 31.36, "epoch_pct": 31.37, "eta": "45:34:00", "max_grad_norm": 0.8, "loss": 0.6444084644317627, "grad_norm": 1.096709966659546, "learning_rate": 8.492523037042179e-05}
+{"ts": "2025-12-24T11:12:52", "event": "train_log", "step": 4364, "epoch": 1.8828478964401294, "progress_pct": 31.38, "epoch_pct": 31.38, "eta": "45:32:37", "max_grad_norm": 0.8, "loss": 0.6504629850387573, "grad_norm": 1.1675381660461426, "learning_rate": 8.490765569381159e-05}
+{"ts": "2025-12-24T11:13:04", "event": "train_log", "step": 4366, "epoch": 1.8837108953613808, "progress_pct": 31.39, "epoch_pct": 31.4, "eta": "45:31:14", "max_grad_norm": 0.8, "loss": 0.6365054249763489, "grad_norm": 1.2067846059799194, "learning_rate": 8.489007259937758e-05}
+{"ts": "2025-12-24T11:13:17", "event": "train_log", "step": 4368, "epoch": 1.884573894282632, "progress_pct": 31.41, "epoch_pct": 31.41, "eta": "45:29:53", "max_grad_norm": 0.8, "loss": 0.5638332366943359, "grad_norm": 1.0922157764434814, "learning_rate": 8.487248109135986e-05}
+{"ts": "2025-12-24T11:13:30", "event": "train_log", "step": 4370, "epoch": 1.8854368932038836, "progress_pct": 31.42, "epoch_pct": 31.42, "eta": "45:28:31", "max_grad_norm": 0.8, "loss": 0.5690349340438843, "grad_norm": 1.1562784910202026, "learning_rate": 8.485488117400052e-05}
+{"ts": "2025-12-24T11:13:42", "event": "train_log", "step": 4372, "epoch": 1.8862998921251348, "progress_pct": 31.44, "epoch_pct": 31.44, "eta": "45:27:08", "max_grad_norm": 0.8, "loss": 0.6626796126365662, "grad_norm": 1.2277979850769043, "learning_rate": 8.483727285154373e-05}
+{"ts": "2025-12-24T11:13:54", "event": "train_log", "step": 4374, "epoch": 1.8871628910463862, "progress_pct": 31.45, "epoch_pct": 31.45, "eta": "45:25:45", "max_grad_norm": 0.8, "loss": 0.7027351260185242, "grad_norm": 1.1707464456558228, "learning_rate": 8.481965612823562e-05}
+{"ts": "2025-12-24T11:14:06", "event": "train_log", "step": 4376, "epoch": 1.8880258899676376, "progress_pct": 31.46, "epoch_pct": 31.47, "eta": "45:24:24", "max_grad_norm": 0.8, "loss": 0.6483538150787354, "grad_norm": 1.0433859825134277, "learning_rate": 8.480203100832441e-05}
+{"ts": "2025-12-24T11:14:18", "event": "train_log", "step": 4378, "epoch": 1.8888888888888888, "progress_pct": 31.48, "epoch_pct": 31.48, "eta": "45:23:00", "max_grad_norm": 0.8, "loss": 0.6338735818862915, "grad_norm": 1.226797342300415, "learning_rate": 8.478439749606031e-05}
+{"ts": "2025-12-24T11:14:32", "event": "train_log", "step": 4380, "epoch": 1.8897518878101403, "progress_pct": 31.49, "epoch_pct": 31.5, "eta": "45:21:41", "max_grad_norm": 0.8, "loss": 0.5816352367401123, "grad_norm": 1.0683629512786865, "learning_rate": 8.476675559569557e-05}
+{"ts": "2025-12-24T11:14:45", "event": "train_log", "step": 4382, "epoch": 1.8906148867313917, "progress_pct": 31.51, "epoch_pct": 31.51, "eta": "45:20:20", "max_grad_norm": 0.8, "loss": 0.6338825821876526, "grad_norm": 1.030871033668518, "learning_rate": 8.474910531148444e-05}
+{"ts": "2025-12-24T11:14:57", "event": "train_log", "step": 4384, "epoch": 1.8914778856526429, "progress_pct": 31.52, "epoch_pct": 31.52, "eta": "45:18:59", "max_grad_norm": 0.8, "loss": 0.6636899709701538, "grad_norm": 1.13720703125, "learning_rate": 8.473144664768321e-05}
+{"ts": "2025-12-24T11:15:09", "event": "train_log", "step": 4386, "epoch": 1.8923408845738943, "progress_pct": 31.54, "epoch_pct": 31.54, "eta": "45:17:36", "max_grad_norm": 0.8, "loss": 0.7455838322639465, "grad_norm": 1.1313227415084839, "learning_rate": 8.47137796085502e-05}
+{"ts": "2025-12-24T11:15:22", "event": "train_log", "step": 4388, "epoch": 1.8932038834951457, "progress_pct": 31.55, "epoch_pct": 31.55, "eta": "45:16:16", "max_grad_norm": 0.8, "loss": 0.6590731739997864, "grad_norm": 0.9600776433944702, "learning_rate": 8.469610419834572e-05}
+{"ts": "2025-12-24T11:15:35", "event": "train_log", "step": 4390, "epoch": 1.8940668824163969, "progress_pct": 31.56, "epoch_pct": 31.57, "eta": "45:14:55", "max_grad_norm": 0.8, "loss": 0.6724982261657715, "grad_norm": 1.1583033800125122, "learning_rate": 8.467842042133213e-05}
+{"ts": "2025-12-24T11:15:47", "event": "train_log", "step": 4392, "epoch": 1.8949298813376483, "progress_pct": 31.58, "epoch_pct": 31.58, "eta": "45:13:33", "max_grad_norm": 0.8, "loss": 0.6483606100082397, "grad_norm": 1.2100845575332642, "learning_rate": 8.466072828177379e-05}
+{"ts": "2025-12-24T11:15:59", "event": "train_log", "step": 4394, "epoch": 1.8957928802588997, "progress_pct": 31.59, "epoch_pct": 31.6, "eta": "45:12:10", "max_grad_norm": 0.8, "loss": 0.6018115282058716, "grad_norm": 1.1870096921920776, "learning_rate": 8.464302778393707e-05}
+{"ts": "2025-12-24T11:16:11", "event": "train_log", "step": 4396, "epoch": 1.8966558791801509, "progress_pct": 31.61, "epoch_pct": 31.61, "eta": "45:10:49", "max_grad_norm": 0.8, "loss": 0.621696412563324, "grad_norm": 1.2486129999160767, "learning_rate": 8.462531893209036e-05}
+{"ts": "2025-12-24T11:16:25", "event": "train_log", "step": 4398, "epoch": 1.8975188781014025, "progress_pct": 31.62, "epoch_pct": 31.63, "eta": "45:09:30", "max_grad_norm": 0.8, "loss": 0.6095483899116516, "grad_norm": 1.1300864219665527, "learning_rate": 8.46076017305041e-05}
+{"ts": "2025-12-24T11:16:37", "event": "train_log", "step": 4400, "epoch": 1.8983818770226537, "progress_pct": 31.64, "epoch_pct": 31.64, "eta": "45:08:09", "max_grad_norm": 0.8, "loss": 0.5899648666381836, "grad_norm": 1.0515706539154053, "learning_rate": 8.458987618345069e-05}
+{"ts": "2025-12-24T11:24:55", "event": "train_log", "step": 4400, "epoch": 1.8983818770226537, "progress_pct": 31.64, "epoch_pct": 31.64, "eta": "45:26:05", "max_grad_norm": 0.8, "eval_loss": 0.687366783618927, "eval_runtime": 497.5319, "eval_samples_per_second": 4.14, "eval_steps_per_second": 4.14}
+{"ts": "2025-12-24T11:25:07", "event": "train_log", "step": 4402, "epoch": 1.8992448759439051, "progress_pct": 31.65, "epoch_pct": 31.65, "eta": "45:24:42", "max_grad_norm": 0.8, "loss": 0.6216431856155396, "grad_norm": 1.194498896598816, "learning_rate": 8.457214229520455e-05}
+{"ts": "2025-12-24T11:25:19", "event": "train_log", "step": 4404, "epoch": 1.9001078748651565, "progress_pct": 31.67, "epoch_pct": 31.67, "eta": "45:23:20", "max_grad_norm": 0.8, "loss": 0.5427393317222595, "grad_norm": 1.193523645401001, "learning_rate": 8.455440007004219e-05}
+{"ts": "2025-12-24T11:25:31", "event": "train_log", "step": 4406, "epoch": 1.9009708737864077, "progress_pct": 31.68, "epoch_pct": 31.68, "eta": "45:21:57", "max_grad_norm": 0.8, "loss": 0.5973649024963379, "grad_norm": 1.299268126487732, "learning_rate": 8.453664951224198e-05}
+{"ts": "2025-12-24T11:25:44", "event": "train_log", "step": 4408, "epoch": 1.9018338727076591, "progress_pct": 31.69, "epoch_pct": 31.7, "eta": "45:20:35", "max_grad_norm": 0.8, "loss": 0.6070631742477417, "grad_norm": 1.0188872814178467, "learning_rate": 8.451889062608446e-05}
+{"ts": "2025-12-24T11:25:56", "event": "train_log", "step": 4410, "epoch": 1.9026968716289105, "progress_pct": 31.71, "epoch_pct": 31.71, "eta": "45:19:13", "max_grad_norm": 0.8, "loss": 0.6672911643981934, "grad_norm": 1.1419471502304077, "learning_rate": 8.450112341585206e-05}
+{"ts": "2025-12-24T11:26:09", "event": "train_log", "step": 4412, "epoch": 1.9035598705501617, "progress_pct": 31.72, "epoch_pct": 31.73, "eta": "45:17:52", "max_grad_norm": 0.8, "loss": 0.7246735095977783, "grad_norm": 1.0402454137802124, "learning_rate": 8.44833478858293e-05}
+{"ts": "2025-12-24T11:26:21", "event": "train_log", "step": 4414, "epoch": 1.9044228694714131, "progress_pct": 31.74, "epoch_pct": 31.74, "eta": "45:16:31", "max_grad_norm": 0.8, "loss": 0.6222015619277954, "grad_norm": 1.1744418144226074, "learning_rate": 8.446556404030264e-05}
+{"ts": "2025-12-24T11:26:34", "event": "train_log", "step": 4416, "epoch": 1.9052858683926646, "progress_pct": 31.75, "epoch_pct": 31.75, "eta": "45:15:11", "max_grad_norm": 0.8, "loss": 0.5821858048439026, "grad_norm": 0.9495980739593506, "learning_rate": 8.44477718835606e-05}
+{"ts": "2025-12-24T11:26:47", "event": "train_log", "step": 4418, "epoch": 1.9061488673139158, "progress_pct": 31.77, "epoch_pct": 31.77, "eta": "45:13:50", "max_grad_norm": 0.8, "loss": 0.6768892407417297, "grad_norm": 1.1915034055709839, "learning_rate": 8.442997141989362e-05}
+{"ts": "2025-12-24T11:27:00", "event": "train_log", "step": 4420, "epoch": 1.9070118662351672, "progress_pct": 31.78, "epoch_pct": 31.78, "eta": "45:12:30", "max_grad_norm": 0.8, "loss": 0.5882668495178223, "grad_norm": 1.083964228630066, "learning_rate": 8.441216265359428e-05}
+{"ts": "2025-12-24T11:27:12", "event": "train_log", "step": 4422, "epoch": 1.9078748651564186, "progress_pct": 31.79, "epoch_pct": 31.8, "eta": "45:11:09", "max_grad_norm": 0.8, "loss": 0.6347549557685852, "grad_norm": 1.1280747652053833, "learning_rate": 8.439434558895702e-05}
+{"ts": "2025-12-24T11:27:24", "event": "train_log", "step": 4424, "epoch": 1.9087378640776698, "progress_pct": 31.81, "epoch_pct": 31.81, "eta": "45:09:47", "max_grad_norm": 0.8, "loss": 0.6966704726219177, "grad_norm": 1.2000610828399658, "learning_rate": 8.43765202302784e-05}
+{"ts": "2025-12-24T11:27:37", "event": "train_log", "step": 4426, "epoch": 1.9096008629989214, "progress_pct": 31.82, "epoch_pct": 31.83, "eta": "45:08:26", "max_grad_norm": 0.8, "loss": 0.6997919082641602, "grad_norm": 1.0745738744735718, "learning_rate": 8.435868658185687e-05}
+{"ts": "2025-12-24T11:27:50", "event": "train_log", "step": 4428, "epoch": 1.9104638619201726, "progress_pct": 31.84, "epoch_pct": 31.84, "eta": "45:07:06", "max_grad_norm": 0.8, "loss": 0.6183820366859436, "grad_norm": 0.9859429597854614, "learning_rate": 8.434084464799298e-05}
+{"ts": "2025-12-24T11:28:02", "event": "train_log", "step": 4430, "epoch": 1.911326860841424, "progress_pct": 31.85, "epoch_pct": 31.86, "eta": "45:05:44", "max_grad_norm": 0.8, "loss": 0.6683753132820129, "grad_norm": 1.1770579814910889, "learning_rate": 8.43229944329892e-05}
+{"ts": "2025-12-24T11:28:14", "event": "train_log", "step": 4432, "epoch": 1.9121898597626754, "progress_pct": 31.87, "epoch_pct": 31.87, "eta": "45:04:22", "max_grad_norm": 0.8, "loss": 0.6190024614334106, "grad_norm": 1.180786371231079, "learning_rate": 8.430513594115006e-05}
+{"ts": "2025-12-24T11:28:26", "event": "train_log", "step": 4434, "epoch": 1.9130528586839266, "progress_pct": 31.88, "epoch_pct": 31.88, "eta": "45:03:00", "max_grad_norm": 0.8, "loss": 0.6235488653182983, "grad_norm": 1.098340630531311, "learning_rate": 8.428726917678201e-05}
+{"ts": "2025-12-24T11:28:38", "event": "train_log", "step": 4436, "epoch": 1.913915857605178, "progress_pct": 31.9, "epoch_pct": 31.9, "eta": "45:01:40", "max_grad_norm": 0.8, "loss": 0.6297350525856018, "grad_norm": 1.0071780681610107, "learning_rate": 8.426939414419359e-05}
+{"ts": "2025-12-24T11:28:50", "event": "train_log", "step": 4438, "epoch": 1.9147788565264294, "progress_pct": 31.91, "epoch_pct": 31.91, "eta": "45:00:18", "max_grad_norm": 0.8, "loss": 0.6700394749641418, "grad_norm": 1.2799793481826782, "learning_rate": 8.425151084769524e-05}
+{"ts": "2025-12-24T11:29:03", "event": "train_log", "step": 4440, "epoch": 1.9156418554476806, "progress_pct": 31.92, "epoch_pct": 31.93, "eta": "44:58:58", "max_grad_norm": 0.8, "loss": 0.5961334109306335, "grad_norm": 1.0455679893493652, "learning_rate": 8.423361929159948e-05}
+{"ts": "2025-12-24T11:29:15", "event": "train_log", "step": 4442, "epoch": 1.916504854368932, "progress_pct": 31.94, "epoch_pct": 31.94, "eta": "44:57:36", "max_grad_norm": 0.8, "loss": 0.7021171450614929, "grad_norm": 1.212123990058899, "learning_rate": 8.421571948022073e-05}
+{"ts": "2025-12-24T11:29:27", "event": "train_log", "step": 4444, "epoch": 1.9173678532901834, "progress_pct": 31.95, "epoch_pct": 31.96, "eta": "44:56:14", "max_grad_norm": 0.8, "loss": 0.6828257441520691, "grad_norm": 1.2384458780288696, "learning_rate": 8.419781141787548e-05}
+{"ts": "2025-12-24T11:29:39", "event": "train_log", "step": 4446, "epoch": 1.9182308522114346, "progress_pct": 31.97, "epoch_pct": 31.97, "eta": "44:54:54", "max_grad_norm": 0.8, "loss": 0.6749114394187927, "grad_norm": 1.100545048713684, "learning_rate": 8.417989510888217e-05}
+{"ts": "2025-12-24T11:29:52", "event": "train_log", "step": 4448, "epoch": 1.919093851132686, "progress_pct": 31.98, "epoch_pct": 31.98, "eta": "44:53:34", "max_grad_norm": 0.8, "loss": 0.6202837228775024, "grad_norm": 1.0241851806640625, "learning_rate": 8.416197055756124e-05}
+{"ts": "2025-12-24T11:30:05", "event": "train_log", "step": 4450, "epoch": 1.9199568500539375, "progress_pct": 32.0, "epoch_pct": 32.0, "eta": "44:52:14", "max_grad_norm": 0.8, "loss": 0.6312109231948853, "grad_norm": 1.1138529777526855, "learning_rate": 8.414403776823511e-05}
+{"ts": "2025-12-24T11:30:17", "event": "train_log", "step": 4452, "epoch": 1.9208198489751886, "progress_pct": 32.01, "epoch_pct": 32.01, "eta": "44:50:53", "max_grad_norm": 0.8, "loss": 0.677727997303009, "grad_norm": 1.1574071645736694, "learning_rate": 8.41260967452282e-05}
+{"ts": "2025-12-24T11:30:29", "event": "train_log", "step": 4454, "epoch": 1.9216828478964403, "progress_pct": 32.02, "epoch_pct": 32.03, "eta": "44:49:33", "max_grad_norm": 0.8, "loss": 0.625155508518219, "grad_norm": 1.1757844686508179, "learning_rate": 8.41081474928669e-05}
+{"ts": "2025-12-24T11:30:41", "event": "train_log", "step": 4456, "epoch": 1.9225458468176915, "progress_pct": 32.04, "epoch_pct": 32.04, "eta": "44:48:11", "max_grad_norm": 0.8, "loss": 0.6271302700042725, "grad_norm": 1.2577253580093384, "learning_rate": 8.409019001547959e-05}
+{"ts": "2025-12-24T11:30:53", "event": "train_log", "step": 4458, "epoch": 1.9234088457389427, "progress_pct": 32.05, "epoch_pct": 32.06, "eta": "44:46:52", "max_grad_norm": 0.8, "loss": 0.60232013463974, "grad_norm": 1.0348252058029175, "learning_rate": 8.407222431739663e-05}
+{"ts": "2025-12-24T11:31:05", "event": "train_log", "step": 4460, "epoch": 1.9242718446601943, "progress_pct": 32.07, "epoch_pct": 32.07, "eta": "44:45:30", "max_grad_norm": 0.8, "loss": 0.6764297485351562, "grad_norm": 1.2921109199523926, "learning_rate": 8.405425040295039e-05}
+{"ts": "2025-12-24T11:31:18", "event": "train_log", "step": 4462, "epoch": 1.9251348435814455, "progress_pct": 32.08, "epoch_pct": 32.09, "eta": "44:44:11", "max_grad_norm": 0.8, "loss": 0.5721253752708435, "grad_norm": 1.2089459896087646, "learning_rate": 8.403626827647517e-05}
+{"ts": "2025-12-24T11:31:31", "event": "train_log", "step": 4464, "epoch": 1.925997842502697, "progress_pct": 32.1, "epoch_pct": 32.1, "eta": "44:42:52", "max_grad_norm": 0.8, "loss": 0.6360597014427185, "grad_norm": 1.1275287866592407, "learning_rate": 8.401827794230729e-05}
+{"ts": "2025-12-24T11:31:43", "event": "train_log", "step": 4466, "epoch": 1.9268608414239483, "progress_pct": 32.11, "epoch_pct": 32.11, "eta": "44:41:32", "max_grad_norm": 0.8, "loss": 0.654008150100708, "grad_norm": 1.162341594696045, "learning_rate": 8.400027940478506e-05}
+{"ts": "2025-12-24T11:31:56", "event": "train_log", "step": 4468, "epoch": 1.9277238403451995, "progress_pct": 32.13, "epoch_pct": 32.13, "eta": "44:40:12", "max_grad_norm": 0.8, "loss": 0.6744330525398254, "grad_norm": 1.1538556814193726, "learning_rate": 8.398227266824871e-05}
+{"ts": "2025-12-24T11:32:08", "event": "train_log", "step": 4470, "epoch": 1.928586839266451, "progress_pct": 32.14, "epoch_pct": 32.14, "eta": "44:38:52", "max_grad_norm": 0.8, "loss": 0.6474409103393555, "grad_norm": 0.9633672833442688, "learning_rate": 8.39642577370405e-05}
+{"ts": "2025-12-24T11:32:20", "event": "train_log", "step": 4472, "epoch": 1.9294498381877023, "progress_pct": 32.15, "epoch_pct": 32.16, "eta": "44:37:32", "max_grad_norm": 0.8, "loss": 0.6781411170959473, "grad_norm": 1.1969858407974243, "learning_rate": 8.394623461550465e-05}
+{"ts": "2025-12-24T11:32:32", "event": "train_log", "step": 4474, "epoch": 1.9303128371089535, "progress_pct": 32.17, "epoch_pct": 32.17, "eta": "44:36:11", "max_grad_norm": 0.8, "loss": 0.6781613826751709, "grad_norm": 1.160700798034668, "learning_rate": 8.392820330798733e-05}
+{"ts": "2025-12-24T11:32:44", "event": "train_log", "step": 4476, "epoch": 1.931175836030205, "progress_pct": 32.18, "epoch_pct": 32.19, "eta": "44:34:50", "max_grad_norm": 0.8, "loss": 0.6875717043876648, "grad_norm": 1.201220154762268, "learning_rate": 8.391016381883678e-05}
+{"ts": "2025-12-24T11:32:57", "event": "train_log", "step": 4478, "epoch": 1.9320388349514563, "progress_pct": 32.2, "epoch_pct": 32.2, "eta": "44:33:31", "max_grad_norm": 0.8, "loss": 0.6446496248245239, "grad_norm": 1.090810775756836, "learning_rate": 8.389211615240306e-05}
+{"ts": "2025-12-24T11:33:09", "event": "train_log", "step": 4480, "epoch": 1.9329018338727075, "progress_pct": 32.21, "epoch_pct": 32.22, "eta": "44:32:13", "max_grad_norm": 0.8, "loss": 0.6835271716117859, "grad_norm": 1.0396226644515991, "learning_rate": 8.387406031303833e-05}
+{"ts": "2025-12-24T11:33:21", "event": "train_log", "step": 4482, "epoch": 1.9337648327939592, "progress_pct": 32.23, "epoch_pct": 32.23, "eta": "44:30:52", "max_grad_norm": 0.8, "loss": 0.587812066078186, "grad_norm": 1.1818757057189941, "learning_rate": 8.385599630509664e-05}
+{"ts": "2025-12-24T11:33:33", "event": "train_log", "step": 4484, "epoch": 1.9346278317152104, "progress_pct": 32.24, "epoch_pct": 32.24, "eta": "44:29:32", "max_grad_norm": 0.8, "loss": 0.6302334666252136, "grad_norm": 1.0651488304138184, "learning_rate": 8.383792413293408e-05}
+{"ts": "2025-12-24T11:33:46", "event": "train_log", "step": 4486, "epoch": 1.9354908306364615, "progress_pct": 32.25, "epoch_pct": 32.26, "eta": "44:28:13", "max_grad_norm": 0.8, "loss": 0.5748880505561829, "grad_norm": 1.0889242887496948, "learning_rate": 8.381984380090865e-05}
+{"ts": "2025-12-24T11:33:58", "event": "train_log", "step": 4488, "epoch": 1.9363538295577132, "progress_pct": 32.27, "epoch_pct": 32.27, "eta": "44:26:53", "max_grad_norm": 0.8, "loss": 0.642596960067749, "grad_norm": 1.0353775024414062, "learning_rate": 8.380175531338037e-05}
+{"ts": "2025-12-24T11:34:11", "event": "train_log", "step": 4490, "epoch": 1.9372168284789644, "progress_pct": 32.28, "epoch_pct": 32.29, "eta": "44:25:34", "max_grad_norm": 0.8, "loss": 0.6552023887634277, "grad_norm": 1.089475154876709, "learning_rate": 8.378365867471115e-05}
+{"ts": "2025-12-24T11:34:23", "event": "train_log", "step": 4492, "epoch": 1.9380798274002158, "progress_pct": 32.3, "epoch_pct": 32.3, "eta": "44:24:15", "max_grad_norm": 0.8, "loss": 0.6077851057052612, "grad_norm": 1.0442018508911133, "learning_rate": 8.376555388926495e-05}
+{"ts": "2025-12-24T11:34:36", "event": "train_log", "step": 4494, "epoch": 1.9389428263214672, "progress_pct": 32.31, "epoch_pct": 32.32, "eta": "44:22:57", "max_grad_norm": 0.8, "loss": 0.5917877554893494, "grad_norm": 1.0850406885147095, "learning_rate": 8.374744096140764e-05}
+{"ts": "2025-12-24T11:34:48", "event": "train_log", "step": 4496, "epoch": 1.9398058252427184, "progress_pct": 32.33, "epoch_pct": 32.33, "eta": "44:21:37", "max_grad_norm": 0.8, "loss": 0.6818947196006775, "grad_norm": 1.0924079418182373, "learning_rate": 8.372931989550708e-05}
+{"ts": "2025-12-24T11:34:59", "event": "train_log", "step": 4498, "epoch": 1.9406688241639698, "progress_pct": 32.34, "epoch_pct": 32.34, "eta": "44:20:16", "max_grad_norm": 0.8, "loss": 0.6909443736076355, "grad_norm": 1.2747828960418701, "learning_rate": 8.371119069593308e-05}
+{"ts": "2025-12-24T11:35:12", "event": "train_log", "step": 4500, "epoch": 1.9415318230852212, "progress_pct": 32.36, "epoch_pct": 32.36, "eta": "44:18:58", "max_grad_norm": 0.8, "loss": 0.6680766344070435, "grad_norm": 1.0612404346466064, "learning_rate": 8.369305336705742e-05}
+{"ts": "2025-12-24T11:43:29", "event": "train_log", "step": 4500, "epoch": 1.9415318230852212, "progress_pct": 32.36, "epoch_pct": 32.36, "eta": "44:36:17", "max_grad_norm": 0.8, "eval_loss": 0.6851214170455933, "eval_runtime": 497.1141, "eval_samples_per_second": 4.144, "eval_steps_per_second": 4.144}
+{"ts": "2025-12-24T11:43:42", "event": "train_log", "step": 4502, "epoch": 1.9423948220064724, "progress_pct": 32.37, "epoch_pct": 32.37, "eta": "44:34:59", "max_grad_norm": 0.8, "loss": 0.5932779312133789, "grad_norm": 1.013123869895935, "learning_rate": 8.367490791325381e-05}
+{"ts": "2025-12-24T11:43:55", "event": "train_log", "step": 4504, "epoch": 1.9432578209277238, "progress_pct": 32.38, "epoch_pct": 32.39, "eta": "44:33:40", "max_grad_norm": 0.8, "loss": 0.6291557550430298, "grad_norm": 1.185764193534851, "learning_rate": 8.365675433889799e-05}
+{"ts": "2025-12-24T11:44:07", "event": "train_log", "step": 4506, "epoch": 1.9441208198489752, "progress_pct": 32.4, "epoch_pct": 32.4, "eta": "44:32:20", "max_grad_norm": 0.8, "loss": 0.6341477036476135, "grad_norm": 1.0753179788589478, "learning_rate": 8.363859264836757e-05}
+{"ts": "2025-12-24T11:44:19", "event": "train_log", "step": 4508, "epoch": 1.9449838187702264, "progress_pct": 32.41, "epoch_pct": 32.42, "eta": "44:30:59", "max_grad_norm": 0.8, "loss": 0.7104454636573792, "grad_norm": 1.2008280754089355, "learning_rate": 8.362042284604217e-05}
+{"ts": "2025-12-24T11:44:32", "event": "train_log", "step": 4510, "epoch": 1.945846817691478, "progress_pct": 32.43, "epoch_pct": 32.43, "eta": "44:29:42", "max_grad_norm": 0.8, "loss": 0.5895495414733887, "grad_norm": 0.9920274615287781, "learning_rate": 8.360224493630338e-05}
+{"ts": "2025-12-24T11:44:44", "event": "train_log", "step": 4512, "epoch": 1.9467098166127292, "progress_pct": 32.44, "epoch_pct": 32.45, "eta": "44:28:22", "max_grad_norm": 0.8, "loss": 0.5886778831481934, "grad_norm": 1.0015859603881836, "learning_rate": 8.358405892353469e-05}
+{"ts": "2025-12-24T11:44:57", "event": "train_log", "step": 4514, "epoch": 1.9475728155339804, "progress_pct": 32.46, "epoch_pct": 32.46, "eta": "44:27:04", "max_grad_norm": 0.8, "loss": 0.7066217660903931, "grad_norm": 1.0359071493148804, "learning_rate": 8.356586481212159e-05}
+{"ts": "2025-12-24T11:45:09", "event": "train_log", "step": 4516, "epoch": 1.948435814455232, "progress_pct": 32.47, "epoch_pct": 32.47, "eta": "44:25:43", "max_grad_norm": 0.8, "loss": 0.6521034240722656, "grad_norm": 1.2075080871582031, "learning_rate": 8.354766260645151e-05}
+{"ts": "2025-12-24T11:45:21", "event": "train_log", "step": 4518, "epoch": 1.9492988133764833, "progress_pct": 32.48, "epoch_pct": 32.49, "eta": "44:24:24", "max_grad_norm": 0.8, "loss": 0.5828588008880615, "grad_norm": 1.0007784366607666, "learning_rate": 8.352945231091384e-05}
+{"ts": "2025-12-24T11:45:33", "event": "train_log", "step": 4520, "epoch": 1.9501618122977347, "progress_pct": 32.5, "epoch_pct": 32.5, "eta": "44:23:04", "max_grad_norm": 0.8, "loss": 0.6763349771499634, "grad_norm": 1.3472176790237427, "learning_rate": 8.351123392989987e-05}
+{"ts": "2025-12-24T11:45:46", "event": "train_log", "step": 4522, "epoch": 1.951024811218986, "progress_pct": 32.51, "epoch_pct": 32.52, "eta": "44:21:45", "max_grad_norm": 0.8, "loss": 0.6664134860038757, "grad_norm": 1.1955901384353638, "learning_rate": 8.349300746780293e-05}
+{"ts": "2025-12-24T11:45:58", "event": "train_log", "step": 4524, "epoch": 1.9518878101402373, "progress_pct": 32.53, "epoch_pct": 32.53, "eta": "44:20:26", "max_grad_norm": 0.8, "loss": 0.6009165644645691, "grad_norm": 1.0307800769805908, "learning_rate": 8.347477292901821e-05}
+{"ts": "2025-12-24T11:46:10", "event": "train_log", "step": 4526, "epoch": 1.9527508090614887, "progress_pct": 32.54, "epoch_pct": 32.55, "eta": "44:19:08", "max_grad_norm": 0.8, "loss": 0.594152569770813, "grad_norm": 1.1367076635360718, "learning_rate": 8.345653031794292e-05}
+{"ts": "2025-12-24T11:46:23", "event": "train_log", "step": 4528, "epoch": 1.95361380798274, "progress_pct": 32.56, "epoch_pct": 32.56, "eta": "44:17:49", "max_grad_norm": 0.8, "loss": 0.6537042260169983, "grad_norm": 1.18290376663208, "learning_rate": 8.343827963897614e-05}
+{"ts": "2025-12-24T11:46:35", "event": "train_log", "step": 4530, "epoch": 1.9544768069039913, "progress_pct": 32.57, "epoch_pct": 32.57, "eta": "44:16:30", "max_grad_norm": 0.8, "loss": 0.6444078683853149, "grad_norm": 1.1995106935501099, "learning_rate": 8.3420020896519e-05}
+{"ts": "2025-12-24T11:46:47", "event": "train_log", "step": 4532, "epoch": 1.9553398058252427, "progress_pct": 32.59, "epoch_pct": 32.59, "eta": "44:15:10", "max_grad_norm": 0.8, "loss": 0.6323422789573669, "grad_norm": 1.2222436666488647, "learning_rate": 8.340175409497446e-05}
+{"ts": "2025-12-24T11:46:59", "event": "train_log", "step": 4534, "epoch": 1.956202804746494, "progress_pct": 32.6, "epoch_pct": 32.6, "eta": "44:13:51", "max_grad_norm": 0.8, "loss": 0.5935752987861633, "grad_norm": 1.103392481803894, "learning_rate": 8.338347923874749e-05}
+{"ts": "2025-12-24T11:47:12", "event": "train_log", "step": 4536, "epoch": 1.9570658036677453, "progress_pct": 32.61, "epoch_pct": 32.62, "eta": "44:12:34", "max_grad_norm": 0.8, "loss": 0.5969859957695007, "grad_norm": 0.8767373561859131, "learning_rate": 8.3365196332245e-05}
+{"ts": "2025-12-24T11:47:25", "event": "train_log", "step": 4538, "epoch": 1.9579288025889967, "progress_pct": 32.63, "epoch_pct": 32.63, "eta": "44:11:16", "max_grad_norm": 0.8, "loss": 0.6458547711372375, "grad_norm": 1.0201600790023804, "learning_rate": 8.334690537987585e-05}
+{"ts": "2025-12-24T11:47:37", "event": "train_log", "step": 4540, "epoch": 1.9587918015102481, "progress_pct": 32.64, "epoch_pct": 32.65, "eta": "44:09:56", "max_grad_norm": 0.8, "loss": 0.7252980470657349, "grad_norm": 1.179448127746582, "learning_rate": 8.332860638605076e-05}
+{"ts": "2025-12-24T11:47:49", "event": "train_log", "step": 4542, "epoch": 1.9596548004314993, "progress_pct": 32.66, "epoch_pct": 32.66, "eta": "44:08:39", "max_grad_norm": 0.8, "loss": 0.5783288478851318, "grad_norm": 0.9380082488059998, "learning_rate": 8.33102993551825e-05}
+{"ts": "2025-12-24T11:48:02", "event": "train_log", "step": 4544, "epoch": 1.960517799352751, "progress_pct": 32.67, "epoch_pct": 32.68, "eta": "44:07:20", "max_grad_norm": 0.8, "loss": 0.6980715990066528, "grad_norm": 1.3073279857635498, "learning_rate": 8.329198429168573e-05}
+{"ts": "2025-12-24T11:48:14", "event": "train_log", "step": 4546, "epoch": 1.9613807982740021, "progress_pct": 32.69, "epoch_pct": 32.69, "eta": "44:06:02", "max_grad_norm": 0.8, "loss": 0.7058124542236328, "grad_norm": 1.057144045829773, "learning_rate": 8.3273661199977e-05}
+{"ts": "2025-12-24T11:48:26", "event": "train_log", "step": 4548, "epoch": 1.9622437971952535, "progress_pct": 32.7, "epoch_pct": 32.7, "eta": "44:04:42", "max_grad_norm": 0.8, "loss": 0.6461919546127319, "grad_norm": 1.1609148979187012, "learning_rate": 8.325533008447488e-05}
+{"ts": "2025-12-24T11:48:38", "event": "train_log", "step": 4550, "epoch": 1.963106796116505, "progress_pct": 32.71, "epoch_pct": 32.72, "eta": "44:03:25", "max_grad_norm": 0.8, "loss": 0.6129994988441467, "grad_norm": 1.0891038179397583, "learning_rate": 8.323699094959983e-05}
+{"ts": "2025-12-24T11:48:51", "event": "train_log", "step": 4552, "epoch": 1.9639697950377561, "progress_pct": 32.73, "epoch_pct": 32.73, "eta": "44:02:07", "max_grad_norm": 0.8, "loss": 0.5800982713699341, "grad_norm": 1.127821922302246, "learning_rate": 8.321864379977425e-05}
+{"ts": "2025-12-24T11:49:03", "event": "train_log", "step": 4554, "epoch": 1.9648327939590076, "progress_pct": 32.74, "epoch_pct": 32.75, "eta": "44:00:49", "max_grad_norm": 0.8, "loss": 0.5805943012237549, "grad_norm": 1.027156114578247, "learning_rate": 8.320028863942247e-05}
+{"ts": "2025-12-24T11:49:16", "event": "train_log", "step": 4556, "epoch": 1.965695792880259, "progress_pct": 32.76, "epoch_pct": 32.76, "eta": "43:59:31", "max_grad_norm": 0.8, "loss": 0.6185514330863953, "grad_norm": 0.9744936227798462, "learning_rate": 8.318192547297074e-05}
+{"ts": "2025-12-24T11:49:28", "event": "train_log", "step": 4558, "epoch": 1.9665587918015102, "progress_pct": 32.77, "epoch_pct": 32.78, "eta": "43:58:12", "max_grad_norm": 0.8, "loss": 0.6853391528129578, "grad_norm": 1.227415680885315, "learning_rate": 8.316355430484725e-05}
+{"ts": "2025-12-24T11:49:41", "event": "train_log", "step": 4560, "epoch": 1.9674217907227616, "progress_pct": 32.79, "epoch_pct": 32.79, "eta": "43:56:56", "max_grad_norm": 0.8, "loss": 0.5712144374847412, "grad_norm": 1.0274931192398071, "learning_rate": 8.314517513948219e-05}
+{"ts": "2025-12-24T11:49:54", "event": "train_log", "step": 4562, "epoch": 1.968284789644013, "progress_pct": 32.8, "epoch_pct": 32.8, "eta": "43:55:39", "max_grad_norm": 0.8, "loss": 0.6449635624885559, "grad_norm": 1.1257516145706177, "learning_rate": 8.312678798130753e-05}
+{"ts": "2025-12-24T11:50:06", "event": "train_log", "step": 4564, "epoch": 1.9691477885652642, "progress_pct": 32.82, "epoch_pct": 32.82, "eta": "43:54:20", "max_grad_norm": 0.8, "loss": 0.6193190217018127, "grad_norm": 1.0795423984527588, "learning_rate": 8.31083928347573e-05}
+{"ts": "2025-12-24T11:50:18", "event": "train_log", "step": 4566, "epoch": 1.9700107874865156, "progress_pct": 32.83, "epoch_pct": 32.83, "eta": "43:53:03", "max_grad_norm": 0.8, "loss": 0.6526190042495728, "grad_norm": 1.147713303565979, "learning_rate": 8.308998970426738e-05}
+{"ts": "2025-12-24T11:50:31", "event": "train_log", "step": 4568, "epoch": 1.970873786407767, "progress_pct": 32.84, "epoch_pct": 32.85, "eta": "43:51:46", "max_grad_norm": 0.8, "loss": 0.6300464272499084, "grad_norm": 1.061994194984436, "learning_rate": 8.307157859427562e-05}
+{"ts": "2025-12-24T11:50:44", "event": "train_log", "step": 4570, "epoch": 1.9717367853290182, "progress_pct": 32.86, "epoch_pct": 32.86, "eta": "43:50:29", "max_grad_norm": 0.8, "loss": 0.6029738783836365, "grad_norm": 1.062877893447876, "learning_rate": 8.305315950922177e-05}
+{"ts": "2025-12-24T11:50:56", "event": "train_log", "step": 4572, "epoch": 1.9725997842502698, "progress_pct": 32.87, "epoch_pct": 32.88, "eta": "43:49:12", "max_grad_norm": 0.8, "loss": 0.6268710494041443, "grad_norm": 1.1812671422958374, "learning_rate": 8.30347324535475e-05}
+{"ts": "2025-12-24T11:51:08", "event": "train_log", "step": 4574, "epoch": 1.973462783171521, "progress_pct": 32.89, "epoch_pct": 32.89, "eta": "43:47:53", "max_grad_norm": 0.8, "loss": 0.6771203279495239, "grad_norm": 1.4144628047943115, "learning_rate": 8.301629743169644e-05}
+{"ts": "2025-12-24T11:51:21", "event": "train_log", "step": 4576, "epoch": 1.9743257820927724, "progress_pct": 32.9, "epoch_pct": 32.91, "eta": "43:46:36", "max_grad_norm": 0.8, "loss": 0.6194970011711121, "grad_norm": 1.1511911153793335, "learning_rate": 8.299785444811406e-05}
+{"ts": "2025-12-24T11:51:33", "event": "train_log", "step": 4578, "epoch": 1.9751887810140238, "progress_pct": 32.92, "epoch_pct": 32.92, "eta": "43:45:20", "max_grad_norm": 0.8, "loss": 0.6515170335769653, "grad_norm": 0.93571937084198, "learning_rate": 8.297940350724787e-05}
+{"ts": "2025-12-24T11:51:46", "event": "train_log", "step": 4580, "epoch": 1.976051779935275, "progress_pct": 32.93, "epoch_pct": 32.93, "eta": "43:44:03", "max_grad_norm": 0.8, "loss": 0.588933527469635, "grad_norm": 1.0542898178100586, "learning_rate": 8.296094461354718e-05}
+{"ts": "2025-12-24T11:51:59", "event": "train_log", "step": 4582, "epoch": 1.9769147788565264, "progress_pct": 32.95, "epoch_pct": 32.95, "eta": "43:42:47", "max_grad_norm": 0.8, "loss": 0.6178847551345825, "grad_norm": 1.3952802419662476, "learning_rate": 8.294247777146328e-05}
+{"ts": "2025-12-24T11:52:12", "event": "train_log", "step": 4584, "epoch": 1.9777777777777779, "progress_pct": 32.96, "epoch_pct": 32.96, "eta": "43:41:31", "max_grad_norm": 0.8, "loss": 0.6646043062210083, "grad_norm": 1.0117886066436768, "learning_rate": 8.292400298544938e-05}
+{"ts": "2025-12-24T11:52:25", "event": "train_log", "step": 4586, "epoch": 1.978640776699029, "progress_pct": 32.97, "epoch_pct": 32.98, "eta": "43:40:14", "max_grad_norm": 0.8, "loss": 0.6752292513847351, "grad_norm": 1.1439177989959717, "learning_rate": 8.290552025996057e-05}
+{"ts": "2025-12-24T11:52:37", "event": "train_log", "step": 4588, "epoch": 1.9795037756202805, "progress_pct": 32.99, "epoch_pct": 32.99, "eta": "43:38:57", "max_grad_norm": 0.8, "loss": 0.6157812476158142, "grad_norm": 1.414772629737854, "learning_rate": 8.288702959945389e-05}
+{"ts": "2025-12-24T11:52:49", "event": "train_log", "step": 4590, "epoch": 1.9803667745415319, "progress_pct": 33.0, "epoch_pct": 33.01, "eta": "43:37:40", "max_grad_norm": 0.8, "loss": 0.7817996144294739, "grad_norm": 1.2501076459884644, "learning_rate": 8.28685310083883e-05}
+{"ts": "2025-12-24T11:53:02", "event": "train_log", "step": 4592, "epoch": 1.981229773462783, "progress_pct": 33.02, "epoch_pct": 33.02, "eta": "43:36:23", "max_grad_norm": 0.8, "loss": 0.6192898154258728, "grad_norm": 1.038573145866394, "learning_rate": 8.285002449122461e-05}
+{"ts": "2025-12-24T11:53:15", "event": "train_log", "step": 4594, "epoch": 1.9820927723840345, "progress_pct": 33.03, "epoch_pct": 33.03, "eta": "43:35:07", "max_grad_norm": 0.8, "loss": 0.6016383171081543, "grad_norm": 1.2579835653305054, "learning_rate": 8.28315100524256e-05}
+{"ts": "2025-12-24T11:53:27", "event": "train_log", "step": 4596, "epoch": 1.9829557713052859, "progress_pct": 33.05, "epoch_pct": 33.05, "eta": "43:33:50", "max_grad_norm": 0.8, "loss": 0.6279599666595459, "grad_norm": 1.1748261451721191, "learning_rate": 8.281298769645596e-05}
+{"ts": "2025-12-24T11:53:40", "event": "train_log", "step": 4598, "epoch": 1.983818770226537, "progress_pct": 33.06, "epoch_pct": 33.06, "eta": "43:32:34", "max_grad_norm": 0.8, "loss": 0.5715296864509583, "grad_norm": 1.119197130203247, "learning_rate": 8.279445742778225e-05}
+{"ts": "2025-12-24T11:53:52", "event": "train_log", "step": 4600, "epoch": 1.9846817691477887, "progress_pct": 33.07, "epoch_pct": 33.08, "eta": "43:31:18", "max_grad_norm": 0.8, "loss": 0.5900562405586243, "grad_norm": 1.200272798538208, "learning_rate": 8.277591925087296e-05}
+{"ts": "2025-12-24T12:02:11", "event": "train_log", "step": 4600, "epoch": 1.9846817691477887, "progress_pct": 33.07, "epoch_pct": 33.08, "eta": "43:48:06", "max_grad_norm": 0.8, "eval_loss": 0.6843361258506775, "eval_runtime": 498.3201, "eval_samples_per_second": 4.134, "eval_steps_per_second": 4.134}
+{"ts": "2025-12-24T12:02:24", "event": "train_log", "step": 4602, "epoch": 1.98554476806904, "progress_pct": 33.09, "epoch_pct": 33.09, "eta": "43:46:50", "max_grad_norm": 0.8, "loss": 0.5733479261398315, "grad_norm": 1.011797547340393, "learning_rate": 8.275737317019851e-05}
+{"ts": "2025-12-24T12:02:36", "event": "train_log", "step": 4604, "epoch": 1.9864077669902913, "progress_pct": 33.1, "epoch_pct": 33.11, "eta": "43:45:33", "max_grad_norm": 0.8, "loss": 0.7013291120529175, "grad_norm": 1.3356634378433228, "learning_rate": 8.273881919023119e-05}
+{"ts": "2025-12-24T12:02:49", "event": "train_log", "step": 4606, "epoch": 1.9872707659115427, "progress_pct": 33.12, "epoch_pct": 33.12, "eta": "43:44:17", "max_grad_norm": 0.8, "loss": 0.5518448948860168, "grad_norm": 1.2911512851715088, "learning_rate": 8.272025731544522e-05}
+{"ts": "2025-12-24T12:03:02", "event": "train_log", "step": 4608, "epoch": 1.988133764832794, "progress_pct": 33.13, "epoch_pct": 33.14, "eta": "43:43:01", "max_grad_norm": 0.8, "loss": 0.5563185214996338, "grad_norm": 0.9338319897651672, "learning_rate": 8.270168755031668e-05}
+{"ts": "2025-12-24T12:03:14", "event": "train_log", "step": 4610, "epoch": 1.9889967637540453, "progress_pct": 33.15, "epoch_pct": 33.15, "eta": "43:41:42", "max_grad_norm": 0.8, "loss": 0.6348791122436523, "grad_norm": 1.4735357761383057, "learning_rate": 8.268310989932362e-05}
+{"ts": "2025-12-24T12:03:26", "event": "train_log", "step": 4612, "epoch": 1.9898597626752967, "progress_pct": 33.16, "epoch_pct": 33.16, "eta": "43:40:25", "max_grad_norm": 0.8, "loss": 0.6168153285980225, "grad_norm": 1.2777546644210815, "learning_rate": 8.266452436694593e-05}
+{"ts": "2025-12-24T12:03:40", "event": "train_log", "step": 4614, "epoch": 1.990722761596548, "progress_pct": 33.18, "epoch_pct": 33.18, "eta": "43:39:11", "max_grad_norm": 0.8, "loss": 0.5402522683143616, "grad_norm": 1.087555170059204, "learning_rate": 8.264593095766547e-05}
+{"ts": "2025-12-24T12:03:52", "event": "train_log", "step": 4616, "epoch": 1.9915857605177993, "progress_pct": 33.19, "epoch_pct": 33.19, "eta": "43:37:53", "max_grad_norm": 0.8, "loss": 0.6326624155044556, "grad_norm": 1.130277395248413, "learning_rate": 8.262732967596591e-05}
+{"ts": "2025-12-24T12:04:04", "event": "train_log", "step": 4618, "epoch": 1.9924487594390508, "progress_pct": 33.2, "epoch_pct": 33.21, "eta": "43:36:36", "max_grad_norm": 0.8, "loss": 0.6706728935241699, "grad_norm": 1.200890064239502, "learning_rate": 8.260872052633289e-05}
+{"ts": "2025-12-24T12:04:16", "event": "train_log", "step": 4620, "epoch": 1.993311758360302, "progress_pct": 33.22, "epoch_pct": 33.22, "eta": "43:35:18", "max_grad_norm": 0.8, "loss": 0.6825381517410278, "grad_norm": 1.2254823446273804, "learning_rate": 8.25901035132539e-05}
+{"ts": "2025-12-24T12:04:28", "event": "train_log", "step": 4622, "epoch": 1.9941747572815534, "progress_pct": 33.23, "epoch_pct": 33.24, "eta": "43:34:01", "max_grad_norm": 0.8, "loss": 0.5958535671234131, "grad_norm": 1.1195508241653442, "learning_rate": 8.257147864121838e-05}
+{"ts": "2025-12-24T12:04:41", "event": "train_log", "step": 4624, "epoch": 1.9950377562028048, "progress_pct": 33.25, "epoch_pct": 33.25, "eta": "43:32:45", "max_grad_norm": 0.8, "loss": 0.5317433476448059, "grad_norm": 1.1744102239608765, "learning_rate": 8.255284591471762e-05}
+{"ts": "2025-12-24T12:04:53", "event": "train_log", "step": 4626, "epoch": 1.995900755124056, "progress_pct": 33.26, "epoch_pct": 33.27, "eta": "43:31:27", "max_grad_norm": 0.8, "loss": 0.6555085182189941, "grad_norm": 1.161426305770874, "learning_rate": 8.253420533824481e-05}
+{"ts": "2025-12-24T12:05:06", "event": "train_log", "step": 4628, "epoch": 1.9967637540453076, "progress_pct": 33.28, "epoch_pct": 33.28, "eta": "43:30:11", "max_grad_norm": 0.8, "loss": 0.6568807363510132, "grad_norm": 1.1517757177352905, "learning_rate": 8.251555691629504e-05}
+{"ts": "2025-12-24T12:05:18", "event": "train_log", "step": 4630, "epoch": 1.9976267529665588, "progress_pct": 33.29, "epoch_pct": 33.29, "eta": "43:28:54", "max_grad_norm": 0.8, "loss": 0.6311682462692261, "grad_norm": 1.0744560956954956, "learning_rate": 8.249690065336531e-05}
+{"ts": "2025-12-24T12:05:31", "event": "train_log", "step": 4632, "epoch": 1.9984897518878102, "progress_pct": 33.3, "epoch_pct": 33.31, "eta": "43:27:39", "max_grad_norm": 0.8, "loss": 0.6244885921478271, "grad_norm": 1.107391595840454, "learning_rate": 8.247823655395448e-05}
+{"ts": "2025-12-24T12:05:44", "event": "train_log", "step": 4634, "epoch": 1.9993527508090616, "progress_pct": 33.32, "epoch_pct": 33.32, "eta": "43:26:25", "max_grad_norm": 0.8, "loss": 0.5207338929176331, "grad_norm": 0.8700313568115234, "learning_rate": 8.245956462256332e-05}
+{"ts": "2025-12-24T12:05:54", "event": "train_log", "step": 4636, "epoch": 2.0, "progress_pct": 33.33, "epoch_pct": 33.33, "eta": "43:25:02", "max_grad_norm": 0.8, "loss": 0.588297963142395, "grad_norm": 1.6266279220581055, "learning_rate": 8.244088486369449e-05}
+{"ts": "2025-12-24T12:06:05", "event": "train_log", "step": 4638, "epoch": 2.000862998921251, "progress_pct": 33.35, "epoch_pct": 33.35, "eta": "43:23:44", "max_grad_norm": 0.8, "loss": 0.6075456142425537, "grad_norm": 1.3440468311309814, "learning_rate": 8.24221972818525e-05}
+{"ts": "2025-12-24T12:06:18", "event": "train_log", "step": 4640, "epoch": 2.001725997842503, "progress_pct": 33.36, "epoch_pct": 33.36, "eta": "43:22:29", "max_grad_norm": 0.8, "loss": 0.6029927134513855, "grad_norm": 1.1257013082504272, "learning_rate": 8.240350188154381e-05}
+{"ts": "2025-12-24T12:06:31", "event": "train_log", "step": 4642, "epoch": 2.002588996763754, "progress_pct": 33.38, "epoch_pct": 33.38, "eta": "43:21:14", "max_grad_norm": 0.8, "loss": 0.627281665802002, "grad_norm": 1.048938512802124, "learning_rate": 8.238479866727673e-05}
+{"ts": "2025-12-24T12:06:43", "event": "train_log", "step": 4644, "epoch": 2.003451995685005, "progress_pct": 33.39, "epoch_pct": 33.39, "eta": "43:19:57", "max_grad_norm": 0.8, "loss": 0.5027982592582703, "grad_norm": 1.095493197441101, "learning_rate": 8.236608764356141e-05}
+{"ts": "2025-12-24T12:06:55", "event": "train_log", "step": 4646, "epoch": 2.004314994606257, "progress_pct": 33.41, "epoch_pct": 33.41, "eta": "43:18:40", "max_grad_norm": 0.8, "loss": 0.5688901543617249, "grad_norm": 1.155165672302246, "learning_rate": 8.234736881490998e-05}
+{"ts": "2025-12-24T12:07:07", "event": "train_log", "step": 4648, "epoch": 2.005177993527508, "progress_pct": 33.42, "epoch_pct": 33.42, "eta": "43:17:24", "max_grad_norm": 0.8, "loss": 0.6120443344116211, "grad_norm": 1.1753524541854858, "learning_rate": 8.232864218583638e-05}
+{"ts": "2025-12-24T12:07:20", "event": "train_log", "step": 4650, "epoch": 2.006040992448759, "progress_pct": 33.43, "epoch_pct": 33.43, "eta": "43:16:08", "max_grad_norm": 0.8, "loss": 0.6186259388923645, "grad_norm": 1.1251633167266846, "learning_rate": 8.230990776085645e-05}
+{"ts": "2025-12-24T12:07:33", "event": "train_log", "step": 4652, "epoch": 2.006903991370011, "progress_pct": 33.45, "epoch_pct": 33.45, "eta": "43:14:54", "max_grad_norm": 0.8, "loss": 0.5442436933517456, "grad_norm": 1.1182184219360352, "learning_rate": 8.229116554448792e-05}
+{"ts": "2025-12-24T12:07:47", "event": "train_log", "step": 4654, "epoch": 2.007766990291262, "progress_pct": 33.46, "epoch_pct": 33.46, "eta": "43:13:40", "max_grad_norm": 0.8, "loss": 0.5792908668518066, "grad_norm": 1.2066760063171387, "learning_rate": 8.22724155412504e-05}
+{"ts": "2025-12-24T12:07:58", "event": "train_log", "step": 4656, "epoch": 2.0086299892125137, "progress_pct": 33.48, "epoch_pct": 33.48, "eta": "43:12:23", "max_grad_norm": 0.8, "loss": 0.6537746787071228, "grad_norm": 1.1738206148147583, "learning_rate": 8.225365775566535e-05}
+{"ts": "2025-12-24T12:08:10", "event": "train_log", "step": 4658, "epoch": 2.009492988133765, "progress_pct": 33.49, "epoch_pct": 33.49, "eta": "43:11:07", "max_grad_norm": 0.8, "loss": 0.6149779558181763, "grad_norm": 1.052475094795227, "learning_rate": 8.223489219225614e-05}
+{"ts": "2025-12-24T12:08:23", "event": "train_log", "step": 4660, "epoch": 2.010355987055016, "progress_pct": 33.51, "epoch_pct": 33.51, "eta": "43:09:51", "max_grad_norm": 0.8, "loss": 0.5928406119346619, "grad_norm": 1.063461184501648, "learning_rate": 8.221611885554798e-05}
+{"ts": "2025-12-24T12:08:35", "event": "train_log", "step": 4662, "epoch": 2.0112189859762677, "progress_pct": 33.52, "epoch_pct": 33.52, "eta": "43:08:35", "max_grad_norm": 0.8, "loss": 0.5731621980667114, "grad_norm": 1.1224561929702759, "learning_rate": 8.2197337750068e-05}
+{"ts": "2025-12-24T12:08:48", "event": "train_log", "step": 4664, "epoch": 2.012081984897519, "progress_pct": 33.53, "epoch_pct": 33.53, "eta": "43:07:20", "max_grad_norm": 0.8, "loss": 0.5881656408309937, "grad_norm": 0.999091386795044, "learning_rate": 8.217854888034516e-05}
+{"ts": "2025-12-24T12:09:00", "event": "train_log", "step": 4666, "epoch": 2.01294498381877, "progress_pct": 33.55, "epoch_pct": 33.55, "eta": "43:06:03", "max_grad_norm": 0.8, "loss": 0.5597090721130371, "grad_norm": 1.1999011039733887, "learning_rate": 8.215975225091031e-05}
+{"ts": "2025-12-24T12:09:12", "event": "train_log", "step": 4668, "epoch": 2.0138079827400217, "progress_pct": 33.56, "epoch_pct": 33.56, "eta": "43:04:48", "max_grad_norm": 0.8, "loss": 0.5848296284675598, "grad_norm": 1.1901999711990356, "learning_rate": 8.21409478662962e-05}
+{"ts": "2025-12-24T12:09:25", "event": "train_log", "step": 4670, "epoch": 2.014670981661273, "progress_pct": 33.58, "epoch_pct": 33.58, "eta": "43:03:34", "max_grad_norm": 0.8, "loss": 0.6188482642173767, "grad_norm": 1.1389679908752441, "learning_rate": 8.21221357310374e-05}
+{"ts": "2025-12-24T12:09:39", "event": "train_log", "step": 4672, "epoch": 2.015533980582524, "progress_pct": 33.59, "epoch_pct": 33.59, "eta": "43:02:20", "max_grad_norm": 0.8, "loss": 0.6273277401924133, "grad_norm": 1.0234452486038208, "learning_rate": 8.210331584967038e-05}
+{"ts": "2025-12-24T12:09:51", "event": "train_log", "step": 4674, "epoch": 2.0163969795037757, "progress_pct": 33.61, "epoch_pct": 33.61, "eta": "43:01:04", "max_grad_norm": 0.8, "loss": 0.6496222615242004, "grad_norm": 1.241886854171753, "learning_rate": 8.208448822673345e-05}
+{"ts": "2025-12-24T12:10:04", "event": "train_log", "step": 4676, "epoch": 2.017259978425027, "progress_pct": 33.62, "epoch_pct": 33.62, "eta": "42:59:51", "max_grad_norm": 0.8, "loss": 0.49053460359573364, "grad_norm": 0.8620371222496033, "learning_rate": 8.206565286676681e-05}
+{"ts": "2025-12-24T12:10:17", "event": "train_log", "step": 4678, "epoch": 2.018122977346278, "progress_pct": 33.64, "epoch_pct": 33.64, "eta": "42:58:36", "max_grad_norm": 0.8, "loss": 0.6046832799911499, "grad_norm": 1.2506135702133179, "learning_rate": 8.204680977431257e-05}
+{"ts": "2025-12-24T12:10:29", "event": "train_log", "step": 4680, "epoch": 2.0189859762675297, "progress_pct": 33.65, "epoch_pct": 33.65, "eta": "42:57:21", "max_grad_norm": 0.8, "loss": 0.6834520697593689, "grad_norm": 1.1049364805221558, "learning_rate": 8.20279589539146e-05}
+{"ts": "2025-12-24T12:10:42", "event": "train_log", "step": 4682, "epoch": 2.019848975188781, "progress_pct": 33.66, "epoch_pct": 33.66, "eta": "42:56:07", "max_grad_norm": 0.8, "loss": 0.5879999995231628, "grad_norm": 1.1099281311035156, "learning_rate": 8.200910041011872e-05}
+{"ts": "2025-12-24T12:10:54", "event": "train_log", "step": 4684, "epoch": 2.0207119741100326, "progress_pct": 33.68, "epoch_pct": 33.68, "eta": "42:54:52", "max_grad_norm": 0.8, "loss": 0.5747155547142029, "grad_norm": 1.237915277481079, "learning_rate": 8.199023414747256e-05}
+{"ts": "2025-12-24T12:11:07", "event": "train_log", "step": 4686, "epoch": 2.0215749730312838, "progress_pct": 33.69, "epoch_pct": 33.69, "eta": "42:53:38", "max_grad_norm": 0.8, "loss": 0.5814566016197205, "grad_norm": 0.9957588315010071, "learning_rate": 8.197136017052565e-05}
+{"ts": "2025-12-24T12:11:20", "event": "train_log", "step": 4688, "epoch": 2.022437971952535, "progress_pct": 33.71, "epoch_pct": 33.71, "eta": "42:52:24", "max_grad_norm": 0.8, "loss": 0.518153965473175, "grad_norm": 1.0488251447677612, "learning_rate": 8.195247848382937e-05}
+{"ts": "2025-12-24T12:11:33", "event": "train_log", "step": 4690, "epoch": 2.0233009708737866, "progress_pct": 33.72, "epoch_pct": 33.72, "eta": "42:51:09", "max_grad_norm": 0.8, "loss": 0.5115796327590942, "grad_norm": 1.132694959640503, "learning_rate": 8.193358909193696e-05}
+{"ts": "2025-12-24T12:11:44", "event": "train_log", "step": 4692, "epoch": 2.0241639697950378, "progress_pct": 33.74, "epoch_pct": 33.74, "eta": "42:49:53", "max_grad_norm": 0.8, "loss": 0.6047922968864441, "grad_norm": 1.2228015661239624, "learning_rate": 8.191469199940347e-05}
+{"ts": "2025-12-24T12:11:57", "event": "train_log", "step": 4694, "epoch": 2.025026968716289, "progress_pct": 33.75, "epoch_pct": 33.75, "eta": "42:48:38", "max_grad_norm": 0.8, "loss": 0.5399408936500549, "grad_norm": 1.1258149147033691, "learning_rate": 8.18957872107859e-05}
+{"ts": "2025-12-24T12:12:09", "event": "train_log", "step": 4696, "epoch": 2.0258899676375406, "progress_pct": 33.76, "epoch_pct": 33.76, "eta": "42:47:23", "max_grad_norm": 0.8, "loss": 0.5849165320396423, "grad_norm": 1.1760696172714233, "learning_rate": 8.187687473064302e-05}
+{"ts": "2025-12-24T12:12:21", "event": "train_log", "step": 4698, "epoch": 2.026752966558792, "progress_pct": 33.78, "epoch_pct": 33.78, "eta": "42:46:08", "max_grad_norm": 0.8, "loss": 0.6258512139320374, "grad_norm": 1.2452905178070068, "learning_rate": 8.18579545635355e-05}
+{"ts": "2025-12-24T12:12:33", "event": "train_log", "step": 4700, "epoch": 2.027615965480043, "progress_pct": 33.79, "epoch_pct": 33.79, "eta": "42:44:53", "max_grad_norm": 0.8, "loss": 0.5776714086532593, "grad_norm": 1.2427078485488892, "learning_rate": 8.183902671402586e-05}
+{"ts": "2025-12-24T12:20:51", "event": "train_log", "step": 4700, "epoch": 2.027615965480043, "progress_pct": 33.79, "epoch_pct": 33.79, "eta": "43:01:08", "max_grad_norm": 0.8, "eval_loss": 0.6861101984977722, "eval_runtime": 498.0682, "eval_samples_per_second": 4.136, "eval_steps_per_second": 4.136}
+{"ts": "2025-12-24T12:21:04", "event": "train_log", "step": 4702, "epoch": 2.0284789644012946, "progress_pct": 33.81, "epoch_pct": 33.81, "eta": "42:59:53", "max_grad_norm": 0.8, "loss": 0.5752840042114258, "grad_norm": 1.228092074394226, "learning_rate": 8.182009118667843e-05}
+{"ts": "2025-12-24T12:21:16", "event": "train_log", "step": 4704, "epoch": 2.029341963322546, "progress_pct": 33.82, "epoch_pct": 33.82, "eta": "42:58:38", "max_grad_norm": 0.8, "loss": 0.563294529914856, "grad_norm": 1.3129007816314697, "learning_rate": 8.18011479860595e-05}
+{"ts": "2025-12-24T12:21:29", "event": "train_log", "step": 4706, "epoch": 2.030204962243797, "progress_pct": 33.84, "epoch_pct": 33.84, "eta": "42:57:24", "max_grad_norm": 0.8, "loss": 0.5915329456329346, "grad_norm": 1.0661308765411377, "learning_rate": 8.178219711673707e-05}
+{"ts": "2025-12-24T12:21:42", "event": "train_log", "step": 4708, "epoch": 2.0310679611650486, "progress_pct": 33.85, "epoch_pct": 33.85, "eta": "42:56:09", "max_grad_norm": 0.8, "loss": 0.5696240663528442, "grad_norm": 1.3143688440322876, "learning_rate": 8.176323858328108e-05}
+{"ts": "2025-12-24T12:21:54", "event": "train_log", "step": 4710, "epoch": 2.0319309600863, "progress_pct": 33.87, "epoch_pct": 33.87, "eta": "42:54:54", "max_grad_norm": 0.8, "loss": 0.6608805656433105, "grad_norm": 1.1617306470870972, "learning_rate": 8.17442723902633e-05}
+{"ts": "2025-12-24T12:22:06", "event": "train_log", "step": 4712, "epoch": 2.0327939590075514, "progress_pct": 33.88, "epoch_pct": 33.88, "eta": "42:53:39", "max_grad_norm": 0.8, "loss": 0.629772961139679, "grad_norm": 1.5208172798156738, "learning_rate": 8.172529854225734e-05}
+{"ts": "2025-12-24T12:22:19", "event": "train_log", "step": 4714, "epoch": 2.0336569579288026, "progress_pct": 33.89, "epoch_pct": 33.89, "eta": "42:52:25", "max_grad_norm": 0.8, "loss": 0.5098171234130859, "grad_norm": 1.1141911745071411, "learning_rate": 8.170631704383866e-05}
+{"ts": "2025-12-24T12:22:31", "event": "train_log", "step": 4716, "epoch": 2.034519956850054, "progress_pct": 33.91, "epoch_pct": 33.91, "eta": "42:51:10", "max_grad_norm": 0.8, "loss": 0.6024818420410156, "grad_norm": 1.3250960111618042, "learning_rate": 8.168732789958452e-05}
+{"ts": "2025-12-24T12:22:43", "event": "train_log", "step": 4718, "epoch": 2.0353829557713055, "progress_pct": 33.92, "epoch_pct": 33.92, "eta": "42:49:54", "max_grad_norm": 0.8, "loss": 0.6644518375396729, "grad_norm": 1.248282551765442, "learning_rate": 8.166833111407415e-05}
+{"ts": "2025-12-24T12:22:56", "event": "train_log", "step": 4720, "epoch": 2.0362459546925566, "progress_pct": 33.94, "epoch_pct": 33.94, "eta": "42:48:39", "max_grad_norm": 0.8, "loss": 0.6094741821289062, "grad_norm": 1.1487561464309692, "learning_rate": 8.164932669188847e-05}
+{"ts": "2025-12-24T12:23:07", "event": "train_log", "step": 4722, "epoch": 2.037108953613808, "progress_pct": 33.95, "epoch_pct": 33.95, "eta": "42:47:23", "max_grad_norm": 0.8, "loss": 0.6366296410560608, "grad_norm": 1.1959017515182495, "learning_rate": 8.163031463761035e-05}
+{"ts": "2025-12-24T12:23:19", "event": "train_log", "step": 4724, "epoch": 2.0379719525350595, "progress_pct": 33.97, "epoch_pct": 33.97, "eta": "42:46:08", "max_grad_norm": 0.8, "loss": 0.6124740839004517, "grad_norm": 1.3806601762771606, "learning_rate": 8.161129495582444e-05}
+{"ts": "2025-12-24T12:23:32", "event": "train_log", "step": 4726, "epoch": 2.0388349514563107, "progress_pct": 33.98, "epoch_pct": 33.98, "eta": "42:44:54", "max_grad_norm": 0.8, "loss": 0.5960785150527954, "grad_norm": 1.1086201667785645, "learning_rate": 8.159226765111725e-05}
+{"ts": "2025-12-24T12:23:44", "event": "train_log", "step": 4728, "epoch": 2.039697950377562, "progress_pct": 33.99, "epoch_pct": 33.99, "eta": "42:43:38", "max_grad_norm": 0.8, "loss": 0.5667157173156738, "grad_norm": 1.193153977394104, "learning_rate": 8.157323272807713e-05}
+{"ts": "2025-12-24T12:23:55", "event": "train_log", "step": 4730, "epoch": 2.0405609492988135, "progress_pct": 34.01, "epoch_pct": 34.01, "eta": "42:42:22", "max_grad_norm": 0.8, "loss": 0.6202070116996765, "grad_norm": 1.3693162202835083, "learning_rate": 8.155419019129429e-05}
+{"ts": "2025-12-24T12:24:07", "event": "train_log", "step": 4732, "epoch": 2.0414239482200647, "progress_pct": 34.02, "epoch_pct": 34.02, "eta": "42:41:06", "max_grad_norm": 0.8, "loss": 0.6065918803215027, "grad_norm": 1.3801589012145996, "learning_rate": 8.153514004536071e-05}
+{"ts": "2025-12-24T12:24:19", "event": "train_log", "step": 4734, "epoch": 2.042286947141316, "progress_pct": 34.04, "epoch_pct": 34.04, "eta": "42:39:51", "max_grad_norm": 0.8, "loss": 0.5777455568313599, "grad_norm": 1.2313286066055298, "learning_rate": 8.151608229487028e-05}
+{"ts": "2025-12-24T12:24:32", "event": "train_log", "step": 4736, "epoch": 2.0431499460625675, "progress_pct": 34.05, "epoch_pct": 34.05, "eta": "42:38:38", "max_grad_norm": 0.8, "loss": 0.5413148403167725, "grad_norm": 1.244084119796753, "learning_rate": 8.149701694441867e-05}
+{"ts": "2025-12-24T12:24:44", "event": "train_log", "step": 4738, "epoch": 2.0440129449838187, "progress_pct": 34.07, "epoch_pct": 34.07, "eta": "42:37:24", "max_grad_norm": 0.8, "loss": 0.5795334577560425, "grad_norm": 1.3601038455963135, "learning_rate": 8.147794399860343e-05}
+{"ts": "2025-12-24T12:24:58", "event": "train_log", "step": 4740, "epoch": 2.0448759439050703, "progress_pct": 34.08, "epoch_pct": 34.08, "eta": "42:36:11", "max_grad_norm": 0.8, "loss": 0.6170360445976257, "grad_norm": 1.1994701623916626, "learning_rate": 8.145886346202389e-05}
+{"ts": "2025-12-24T12:25:10", "event": "train_log", "step": 4742, "epoch": 2.0457389428263215, "progress_pct": 34.1, "epoch_pct": 34.1, "eta": "42:34:56", "max_grad_norm": 0.8, "loss": 0.5899945497512817, "grad_norm": 1.1869628429412842, "learning_rate": 8.143977533928124e-05}
+{"ts": "2025-12-24T12:25:22", "event": "train_log", "step": 4744, "epoch": 2.0466019417475727, "progress_pct": 34.11, "epoch_pct": 34.11, "eta": "42:33:42", "max_grad_norm": 0.8, "loss": 0.602806806564331, "grad_norm": 1.1021242141723633, "learning_rate": 8.142067963497851e-05}
+{"ts": "2025-12-24T12:25:34", "event": "train_log", "step": 4746, "epoch": 2.0474649406688243, "progress_pct": 34.12, "epoch_pct": 34.12, "eta": "42:32:28", "max_grad_norm": 0.8, "loss": 0.6312713623046875, "grad_norm": 1.080838918685913, "learning_rate": 8.140157635372052e-05}
+{"ts": "2025-12-24T12:25:47", "event": "train_log", "step": 4748, "epoch": 2.0483279395900755, "progress_pct": 34.14, "epoch_pct": 34.14, "eta": "42:31:14", "max_grad_norm": 0.8, "loss": 0.6082773208618164, "grad_norm": 4.683399677276611, "learning_rate": 8.138246550011395e-05}
+{"ts": "2025-12-24T12:25:59", "event": "train_log", "step": 4750, "epoch": 2.0491909385113267, "progress_pct": 34.15, "epoch_pct": 34.15, "eta": "42:30:00", "max_grad_norm": 0.8, "loss": 0.633202314376831, "grad_norm": 1.4420764446258545, "learning_rate": 8.136334707876731e-05}
+{"ts": "2025-12-24T12:26:12", "event": "train_log", "step": 4752, "epoch": 2.0500539374325784, "progress_pct": 34.17, "epoch_pct": 34.17, "eta": "42:28:46", "max_grad_norm": 0.8, "loss": 0.5840990543365479, "grad_norm": 1.1558455228805542, "learning_rate": 8.134422109429089e-05}
+{"ts": "2025-12-24T12:26:26", "event": "train_log", "step": 4754, "epoch": 2.0509169363538295, "progress_pct": 34.18, "epoch_pct": 34.18, "eta": "42:27:35", "max_grad_norm": 0.8, "loss": 0.5268401503562927, "grad_norm": 1.0109466314315796, "learning_rate": 8.132508755129686e-05}
+{"ts": "2025-12-24T12:26:38", "event": "train_log", "step": 4756, "epoch": 2.0517799352750807, "progress_pct": 34.2, "epoch_pct": 34.2, "eta": "42:26:21", "max_grad_norm": 0.8, "loss": 0.6302611827850342, "grad_norm": 1.1790412664413452, "learning_rate": 8.13059464543992e-05}
+{"ts": "2025-12-24T12:26:51", "event": "train_log", "step": 4758, "epoch": 2.0526429341963324, "progress_pct": 34.21, "epoch_pct": 34.21, "eta": "42:25:09", "max_grad_norm": 0.8, "loss": 0.5571357607841492, "grad_norm": 1.114366054534912, "learning_rate": 8.128679780821364e-05}
+{"ts": "2025-12-24T12:27:04", "event": "train_log", "step": 4760, "epoch": 2.0535059331175836, "progress_pct": 34.22, "epoch_pct": 34.23, "eta": "42:23:56", "max_grad_norm": 0.8, "loss": 0.5323016047477722, "grad_norm": 1.044449806213379, "learning_rate": 8.126764161735787e-05}
+{"ts": "2025-12-24T12:27:17", "event": "train_log", "step": 4762, "epoch": 2.0543689320388347, "progress_pct": 34.24, "epoch_pct": 34.24, "eta": "42:22:43", "max_grad_norm": 0.8, "loss": 0.6113949418067932, "grad_norm": 1.2595316171646118, "learning_rate": 8.124847788645127e-05}
+{"ts": "2025-12-24T12:27:30", "event": "train_log", "step": 4764, "epoch": 2.0552319309600864, "progress_pct": 34.25, "epoch_pct": 34.25, "eta": "42:21:31", "max_grad_norm": 0.8, "loss": 0.5971490740776062, "grad_norm": 1.0767589807510376, "learning_rate": 8.122930662011507e-05}
+{"ts": "2025-12-24T12:27:42", "event": "train_log", "step": 4766, "epoch": 2.0560949298813376, "progress_pct": 34.27, "epoch_pct": 34.27, "eta": "42:20:18", "max_grad_norm": 0.8, "loss": 0.5750036239624023, "grad_norm": 1.112290859222412, "learning_rate": 8.12101278229724e-05}
+{"ts": "2025-12-24T12:27:55", "event": "train_log", "step": 4768, "epoch": 2.056957928802589, "progress_pct": 34.28, "epoch_pct": 34.28, "eta": "42:19:04", "max_grad_norm": 0.8, "loss": 0.6123761534690857, "grad_norm": 1.1439614295959473, "learning_rate": 8.119094149964807e-05}
+{"ts": "2025-12-24T12:28:07", "event": "train_log", "step": 4770, "epoch": 2.0578209277238404, "progress_pct": 34.3, "epoch_pct": 34.3, "eta": "42:17:51", "max_grad_norm": 0.8, "loss": 0.5592629313468933, "grad_norm": 1.257593035697937, "learning_rate": 8.117174765476882e-05}
+{"ts": "2025-12-24T12:28:20", "event": "train_log", "step": 4772, "epoch": 2.0586839266450916, "progress_pct": 34.31, "epoch_pct": 34.31, "eta": "42:16:38", "max_grad_norm": 0.8, "loss": 0.5902644991874695, "grad_norm": 1.1807165145874023, "learning_rate": 8.115254629296316e-05}
+{"ts": "2025-12-24T12:28:32", "event": "train_log", "step": 4774, "epoch": 2.0595469255663432, "progress_pct": 34.33, "epoch_pct": 34.33, "eta": "42:15:24", "max_grad_norm": 0.8, "loss": 0.5889407396316528, "grad_norm": 1.5350799560546875, "learning_rate": 8.113333741886138e-05}
+{"ts": "2025-12-24T12:28:45", "event": "train_log", "step": 4776, "epoch": 2.0604099244875944, "progress_pct": 34.34, "epoch_pct": 34.34, "eta": "42:14:11", "max_grad_norm": 0.8, "loss": 0.5814542770385742, "grad_norm": 1.0872184038162231, "learning_rate": 8.111412103709564e-05}
+{"ts": "2025-12-24T12:28:57", "event": "train_log", "step": 4778, "epoch": 2.0612729234088456, "progress_pct": 34.35, "epoch_pct": 34.35, "eta": "42:12:57", "max_grad_norm": 0.8, "loss": 0.5762463808059692, "grad_norm": 1.2108840942382812, "learning_rate": 8.109489715229986e-05}
+{"ts": "2025-12-24T12:29:08", "event": "train_log", "step": 4780, "epoch": 2.0621359223300972, "progress_pct": 34.37, "epoch_pct": 34.37, "eta": "42:11:42", "max_grad_norm": 0.8, "loss": 0.6050645112991333, "grad_norm": 1.3237541913986206, "learning_rate": 8.107566576910984e-05}
+{"ts": "2025-12-24T12:29:21", "event": "train_log", "step": 4782, "epoch": 2.0629989212513484, "progress_pct": 34.38, "epoch_pct": 34.38, "eta": "42:10:29", "max_grad_norm": 0.8, "loss": 0.6265539526939392, "grad_norm": 1.1384598016738892, "learning_rate": 8.105642689216308e-05}
+{"ts": "2025-12-24T12:29:33", "event": "train_log", "step": 4784, "epoch": 2.0638619201725996, "progress_pct": 34.4, "epoch_pct": 34.4, "eta": "42:09:17", "max_grad_norm": 0.8, "loss": 0.6093327403068542, "grad_norm": 1.3722445964813232, "learning_rate": 8.103718052609898e-05}
+{"ts": "2025-12-24T12:29:46", "event": "train_log", "step": 4786, "epoch": 2.0647249190938513, "progress_pct": 34.41, "epoch_pct": 34.41, "eta": "42:08:05", "max_grad_norm": 0.8, "loss": 0.5730020403862, "grad_norm": 1.0969057083129883, "learning_rate": 8.101792667555872e-05}
+{"ts": "2025-12-24T12:29:58", "event": "train_log", "step": 4788, "epoch": 2.0655879180151024, "progress_pct": 34.43, "epoch_pct": 34.43, "eta": "42:06:51", "max_grad_norm": 0.8, "loss": 0.5690904259681702, "grad_norm": 1.2427756786346436, "learning_rate": 8.099866534518528e-05}
+{"ts": "2025-12-24T12:30:10", "event": "train_log", "step": 4790, "epoch": 2.0664509169363536, "progress_pct": 34.44, "epoch_pct": 34.44, "eta": "42:05:37", "max_grad_norm": 0.8, "loss": 0.5794310569763184, "grad_norm": 1.3292549848556519, "learning_rate": 8.097939653962341e-05}
+{"ts": "2025-12-24T12:30:23", "event": "train_log", "step": 4792, "epoch": 2.0673139158576053, "progress_pct": 34.45, "epoch_pct": 34.46, "eta": "42:04:25", "max_grad_norm": 0.8, "loss": 0.5536732077598572, "grad_norm": 1.1240622997283936, "learning_rate": 8.096012026351975e-05}
+{"ts": "2025-12-24T12:30:35", "event": "train_log", "step": 4794, "epoch": 2.0681769147788565, "progress_pct": 34.47, "epoch_pct": 34.47, "eta": "42:03:11", "max_grad_norm": 0.8, "loss": 0.608305811882019, "grad_norm": 1.2449970245361328, "learning_rate": 8.094083652152265e-05}
+{"ts": "2025-12-24T12:30:47", "event": "train_log", "step": 4796, "epoch": 2.0690399137001076, "progress_pct": 34.48, "epoch_pct": 34.48, "eta": "42:01:57", "max_grad_norm": 0.8, "loss": 0.5528193116188049, "grad_norm": 1.3017584085464478, "learning_rate": 8.092154531828229e-05}
+{"ts": "2025-12-24T12:31:00", "event": "train_log", "step": 4798, "epoch": 2.0699029126213593, "progress_pct": 34.5, "epoch_pct": 34.5, "eta": "42:00:46", "max_grad_norm": 0.8, "loss": 0.5502005219459534, "grad_norm": 1.1794867515563965, "learning_rate": 8.090224665845067e-05}
+{"ts": "2025-12-24T12:31:13", "event": "train_log", "step": 4800, "epoch": 2.0707659115426105, "progress_pct": 34.51, "epoch_pct": 34.51, "eta": "41:59:34", "max_grad_norm": 0.8, "loss": 0.5831656455993652, "grad_norm": 1.1529462337493896, "learning_rate": 8.08829405466816e-05}
+{"ts": "2025-12-24T12:39:30", "event": "train_log", "step": 4800, "epoch": 2.0707659115426105, "progress_pct": 34.51, "epoch_pct": 34.51, "eta": "42:15:18", "max_grad_norm": 0.8, "eval_loss": 0.6874316930770874, "eval_runtime": 497.7977, "eval_samples_per_second": 4.138, "eval_steps_per_second": 4.138}
+{"ts": "2025-12-24T12:39:31", "event": "train_log", "step": 4800, "epoch": 2.0707659115426105, "progress_pct": 34.51, "epoch_pct": 34.51, "eta": "42:15:19", "max_grad_norm": 0.8, "train_runtime": 80168.2654, "train_samples_per_second": 1.388, "train_steps_per_second": 0.173, "total_flos": 5.005098268090491e+18, "train_loss": 0.733072043210268}
+{"ts": "2025-12-24T12:47:48", "event": "train_log", "step": 4800, "epoch": 2.0707659115426105, "progress_pct": 34.51, "epoch_pct": 34.51, "eta": "42:31:02", "max_grad_norm": 0.8, "eval_loss": 0.6851214170455933, "eval_runtime": 497.0682, "eval_samples_per_second": 4.144, "eval_steps_per_second": 4.144}
diff --git a/sft_devstral_24B/wandb/debug-internal.log b/sft_devstral_24B/wandb/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..4a3e538e7581e1086374e9b27a45569ef0521436
--- /dev/null
+++ b/sft_devstral_24B/wandb/debug-internal.log
@@ -0,0 +1,11 @@
+{"time":"2025-12-23T14:22:36.190806696Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T14:22:36.346431407Z","level":"INFO","msg":"stream: created new stream","id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346514258Z","level":"INFO","msg":"handler: started","stream_id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346675637Z","level":"INFO","msg":"stream: started","id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346701299Z","level":"INFO","msg":"writer: started","stream_id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346725215Z","level":"INFO","msg":"sender: started","stream_id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.47834979Z","level":"INFO","msg":"stream: closing","id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.717701027Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2025-12-24T12:47:48.809988924Z","level":"INFO","msg":"handler: closed","stream_id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.810086512Z","level":"INFO","msg":"sender: closed","stream_id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.810099591Z","level":"INFO","msg":"stream: closed","id":"cktxoubm"}
diff --git a/sft_devstral_24B/wandb/debug.log b/sft_devstral_24B/wandb/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..38dc783a2520c4ef56c29b4b4862fe4ba48f4669
--- /dev/null
+++ b/sft_devstral_24B/wandb/debug.log
@@ -0,0 +1,26 @@
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Configure stats pid to 924716
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/run-20251223_142235-cktxoubm/logs/debug.log
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/run-20251223_142235-cktxoubm/logs/debug-internal.log
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:init():841] calling init triggers
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 8, 'lora_alpha': 16, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '1e-4', 'weight_decay': 0.0, 'warmup_ratio': 0.08, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 0.8, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 14:22:35,733 INFO    MainThread:924716 [wandb_init.py:init():889] starting backend
+2025-12-23 14:22:36,183 INFO    MainThread:924716 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 14:22:36,187 INFO    MainThread:924716 [wandb_init.py:init():900] backend started and connected
+2025-12-23 14:22:36,189 INFO    MainThread:924716 [wandb_init.py:init():970] updated telemetry
+2025-12-23 14:22:36,193 INFO    MainThread:924716 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 14:22:36,531 INFO    MainThread:924716 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 14:22:36,639 INFO    MainThread:924716 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 14:23:22,767 INFO    MainThread:924716 [wandb_run.py:_config_callback():1396] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.0', 'base_model_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'inference_mode': False, 'r': 8, 'target_modules': ['v_proj', 'k_proj', 'q_proj', 'o_proj'], 'exclude_modules': None, 'lora_alpha': 16, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': None, 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False}}, 'image_token_index': 10, 'projector_hidden_act': 'gelu', 'vision_feature_layer': -1, 'vision_config': {'hidden_size': 1024, 'intermediate_size': 4096, 'num_hidden_layers': 24, 'num_attention_heads': 16, 'num_channels': 3, 'patch_size': 14, 'image_size': 1540, 'attention_dropout': 0.0, 'hidden_act': 'silu', 'head_dim': 64, 'initializer_range': 0.02, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'pixtral', 'output_attentions': False}, 'text_config': {'vocab_size': 131072, 'max_position_embeddings': 393216, 'hidden_size': 5120, 'intermediate_size': 32768, 'num_hidden_layers': 40, 'num_attention_heads': 32, 'sliding_window': None, 'head_dim': 128, 'num_key_value_heads': 8, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'use_cache': True, 'attention_dropout': 0.0, 'rope_parameters': {'beta_fast': 32.0, 'beta_slow': 1.0, 'factor': 48.0, 'llama_4_scaling_beta': 0.1, 'mscale': 1.0, 'mscale_all_dim': 1.0, 'original_max_position_embeddings': 8192, 'rope_theta': 100000000.0, 'rope_type': 'yarn', 'type': 'yarn'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 11, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'ministral3', 'output_attentions': False}, 'multimodal_projector_bias': False, 'spatial_merge_size': 2, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': ['Mistral3ForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'transformers_version': '5.0.0.dev0', 'model_type': 'mistral3', 'use_cache': False, 'output_attentions': False, 'output_dir': 'runs/instruct_run_24b/checkpoints', 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 0.0001, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 0.8, 'num_train_epochs': 6.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.08, 'warmup_steps': 0.08, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': None, 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 2, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'enable_jit_checkpoint': False, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'use_cpu': False, 'seed': 42, 'data_seed': None, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': -1, 'ddp_backend': None, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': 100, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'run_name': None, 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': True, 'metric_for_best_model': 'eval_loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'auto_find_batch_size': False, 'full_determinism': False, 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True}
+2025-12-23 14:23:22,778 INFO    MainThread:924716 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 24022764544 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x7f588665f580>>
+2025-12-23 14:23:22,779 INFO    MainThread:924716 [wandb_run.py:_config_callback():1396] config_cb model/num_parameters 24022764544 None
+2025-12-24 12:47:48,477 INFO    wandb-AsyncioManager-main:924716 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-24 12:47:48,478 INFO    wandb-AsyncioManager-main:924716 [mailbox.py:close():137] Closing mailbox, abandoning 1 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/files/requirements.txt b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug-core.log b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..046e9f1f8a18c1dd5d75eea515e02b4628c0106c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:32:55.161732994Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp92ud704q/port-897869.txt","pid":897869,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:32:55.162358364Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":897869}
+{"time":"2025-12-23T13:32:55.162360862Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-897869-898014-983993294/socket","Net":"unix"}}
+{"time":"2025-12-23T13:32:55.348549556Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:32:55.356229398Z","level":"INFO","msg":"handleInformInit: received","streamId":"t3oy806o","id":"1(@)"}
+{"time":"2025-12-23T13:32:55.465012306Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"t3oy806o","id":"1(@)"}
+{"time":"2025-12-23T13:32:58.32267842Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:32:58.322743392Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:32:58.322834552Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:32:58.322760427Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:32:58.322929711Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-897869-898014-983993294/socket","Net":"unix"}}
+{"time":"2025-12-23T13:32:58.324051402Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:32:58.324084641Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:32:58.324095991Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug-internal.log b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..6b3cdb8b3baa6bbb94cb7e0ce070e9b4b367531f
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug-internal.log
@@ -0,0 +1,12 @@
+{"time":"2025-12-23T13:32:55.359351974Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:32:55.464754045Z","level":"WARN","msg":"featurechecker: GraphQL client is nil, skipping feature loading"}
+{"time":"2025-12-23T13:32:55.464837482Z","level":"INFO","msg":"stream: created new stream","id":"t3oy806o"}
+{"time":"2025-12-23T13:32:55.464921711Z","level":"INFO","msg":"handler: started","stream_id":"t3oy806o"}
+{"time":"2025-12-23T13:32:55.465001765Z","level":"INFO","msg":"stream: started","id":"t3oy806o"}
+{"time":"2025-12-23T13:32:55.465004481Z","level":"INFO","msg":"writer: started","stream_id":"t3oy806o"}
+{"time":"2025-12-23T13:32:55.46502481Z","level":"INFO","msg":"sender: started","stream_id":"t3oy806o"}
+{"time":"2025-12-23T13:32:55.465740966Z","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-12-23T13:32:58.322761817Z","level":"INFO","msg":"stream: closing","id":"t3oy806o"}
+{"time":"2025-12-23T13:32:58.323064036Z","level":"INFO","msg":"handler: closed","stream_id":"t3oy806o"}
+{"time":"2025-12-23T13:32:58.323175551Z","level":"INFO","msg":"sender: closed","stream_id":"t3oy806o"}
+{"time":"2025-12-23T13:32:58.323194397Z","level":"INFO","msg":"stream: closed","id":"t3oy806o"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug.log b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..736e989cbf67b8bf11bbddfb8b6e9d6d4cc7860b
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/logs/debug.log
@@ -0,0 +1,23 @@
+2025-12-23 13:32:54,873 INFO    MainThread:897869 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_setup.py:_flush():80] Configure stats pid to 897869
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/offline-run-20251223_133254-t3oy806o/logs/debug.log
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/offline-run-20251223_133254-t3oy806o/logs/debug-internal.log
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:32:54,874 INFO    MainThread:897869 [wandb_init.py:init():889] starting backend
+2025-12-23 13:32:55,348 INFO    MainThread:897869 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:32:55,353 INFO    MainThread:897869 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:32:55,355 INFO    MainThread:897869 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:32:55,359 INFO    MainThread:897869 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:32:55,467 INFO    MainThread:897869 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:32:55,569 INFO    MainThread:897869 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:32:55,569 INFO    MainThread:897869 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:32:55,569 INFO    MainThread:897869 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:32:55,569 INFO    MainThread:897869 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:32:55,571 INFO    MainThread:897869 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:32:58,322 INFO    wandb-AsyncioManager-main:897869 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:32:58,322 INFO    wandb-AsyncioManager-main:897869 [mailbox.py:close():137] Closing mailbox, abandoning 0 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/run-t3oy806o.wandb b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/run-t3oy806o.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..0a5ce9f245c10a389492ec35dee351db276f11f9
Binary files /dev/null and b/sft_devstral_24B/wandb/offline-run-20251223_133254-t3oy806o/run-t3oy806o.wandb differ
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/files/requirements.txt b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug-core.log b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..48c07545325f41f54db75cad3eb5e37d6d9c7109
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:34:38.012255529Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpf04va_7i/port-898799.txt","pid":898799,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:34:38.013028641Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":898799}
+{"time":"2025-12-23T13:34:38.013018853Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-898799-898898-1696999894/socket","Net":"unix"}}
+{"time":"2025-12-23T13:34:38.196540655Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:34:38.203115384Z","level":"INFO","msg":"handleInformInit: received","streamId":"tqihmbvy","id":"1(@)"}
+{"time":"2025-12-23T13:34:38.311320904Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"tqihmbvy","id":"1(@)"}
+{"time":"2025-12-23T13:34:58.4039307Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:34:58.404019535Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:34:58.404068635Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:34:58.404116577Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:34:58.404253229Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-898799-898898-1696999894/socket","Net":"unix"}}
+{"time":"2025-12-23T13:34:58.40533486Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:34:58.405368266Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:34:58.405384372Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug-internal.log b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..79f67839422b43528da8339c491fc622b2d344bb
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug-internal.log
@@ -0,0 +1,12 @@
+{"time":"2025-12-23T13:34:38.205928509Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:34:38.311006531Z","level":"WARN","msg":"featurechecker: GraphQL client is nil, skipping feature loading"}
+{"time":"2025-12-23T13:34:38.311087297Z","level":"INFO","msg":"stream: created new stream","id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:38.311146087Z","level":"INFO","msg":"handler: started","stream_id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:38.311310287Z","level":"INFO","msg":"stream: started","id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:38.311334253Z","level":"INFO","msg":"writer: started","stream_id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:38.311354678Z","level":"INFO","msg":"sender: started","stream_id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:38.312086644Z","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-12-23T13:34:58.404003803Z","level":"INFO","msg":"stream: closing","id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:58.40438756Z","level":"INFO","msg":"handler: closed","stream_id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:58.404474419Z","level":"INFO","msg":"sender: closed","stream_id":"tqihmbvy"}
+{"time":"2025-12-23T13:34:58.404482953Z","level":"INFO","msg":"stream: closed","id":"tqihmbvy"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug.log b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..163888aa4eae0b05bb4374bfdbc393cbf3da85be
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug.log
@@ -0,0 +1,23 @@
+2025-12-23 13:34:37,735 INFO    MainThread:898799 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:34:37,735 INFO    MainThread:898799 [wandb_setup.py:_flush():80] Configure stats pid to 898799
+2025-12-23 13:34:37,735 INFO    MainThread:898799 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:34:37,735 INFO    MainThread:898799 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:34:37,735 INFO    MainThread:898799 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:34:37,735 INFO    MainThread:898799 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug.log
+2025-12-23 13:34:37,736 INFO    MainThread:898799 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/offline-run-20251223_133437-tqihmbvy/logs/debug-internal.log
+2025-12-23 13:34:37,736 INFO    MainThread:898799 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:34:37,736 INFO    MainThread:898799 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:34:37,736 INFO    MainThread:898799 [wandb_init.py:init():889] starting backend
+2025-12-23 13:34:38,196 INFO    MainThread:898799 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:34:38,201 INFO    MainThread:898799 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:34:38,203 INFO    MainThread:898799 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:34:38,208 INFO    MainThread:898799 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:34:38,315 INFO    MainThread:898799 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:34:38,416 INFO    MainThread:898799 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:34:38,416 INFO    MainThread:898799 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:34:38,416 INFO    MainThread:898799 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:34:38,416 INFO    MainThread:898799 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:34:38,418 INFO    MainThread:898799 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:34:58,404 INFO    wandb-AsyncioManager-main:898799 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:34:58,404 INFO    wandb-AsyncioManager-main:898799 [mailbox.py:close():137] Closing mailbox, abandoning 0 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/run-tqihmbvy.wandb b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/run-tqihmbvy.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..1c4bb2c28efe530a486d4ce1e8251773e33c45a2
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133437-tqihmbvy/run-tqihmbvy.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f43c36fbcb087ff7f6ddc483568b74ed25efbef305e73cabd5e8dd7f6d732f1
+size 213051
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/files/requirements.txt b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug-core.log b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..b581edd9e1b0c2124dba4a1051cca273f5c59ed8
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:36:54.018748834Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp0_lvntyf/port-899968.txt","pid":899968,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:36:54.019348285Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":899968}
+{"time":"2025-12-23T13:36:54.019346995Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-899968-900075-3515173994/socket","Net":"unix"}}
+{"time":"2025-12-23T13:36:54.205502994Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:36:54.212354049Z","level":"INFO","msg":"handleInformInit: received","streamId":"t8yclfex","id":"1(@)"}
+{"time":"2025-12-23T13:36:54.320175047Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"t8yclfex","id":"1(@)"}
+{"time":"2025-12-23T13:36:57.105524406Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:36:57.1055872Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:36:57.105617487Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:36:57.105654666Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:36:57.105732012Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-899968-900075-3515173994/socket","Net":"unix"}}
+{"time":"2025-12-23T13:36:57.106970666Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:36:57.106994091Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:36:57.107004794Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug-internal.log b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..4b1a5d8a0feec01445e545fd9edd7b6fbbc05b66
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug-internal.log
@@ -0,0 +1,12 @@
+{"time":"2025-12-23T13:36:54.215099603Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:36:54.31988329Z","level":"WARN","msg":"featurechecker: GraphQL client is nil, skipping feature loading"}
+{"time":"2025-12-23T13:36:54.319973518Z","level":"INFO","msg":"stream: created new stream","id":"t8yclfex"}
+{"time":"2025-12-23T13:36:54.320031747Z","level":"INFO","msg":"handler: started","stream_id":"t8yclfex"}
+{"time":"2025-12-23T13:36:54.320165549Z","level":"INFO","msg":"stream: started","id":"t8yclfex"}
+{"time":"2025-12-23T13:36:54.320193383Z","level":"INFO","msg":"writer: started","stream_id":"t8yclfex"}
+{"time":"2025-12-23T13:36:54.32021019Z","level":"INFO","msg":"sender: started","stream_id":"t8yclfex"}
+{"time":"2025-12-23T13:36:54.320993812Z","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-12-23T13:36:57.10560038Z","level":"INFO","msg":"stream: closing","id":"t8yclfex"}
+{"time":"2025-12-23T13:36:57.105835641Z","level":"INFO","msg":"handler: closed","stream_id":"t8yclfex"}
+{"time":"2025-12-23T13:36:57.105902965Z","level":"INFO","msg":"sender: closed","stream_id":"t8yclfex"}
+{"time":"2025-12-23T13:36:57.105915546Z","level":"INFO","msg":"stream: closed","id":"t8yclfex"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug.log b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..20af53b31de6aa5d7ec850d4d0611f640ad5c0dd
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/logs/debug.log
@@ -0,0 +1,23 @@
+2025-12-23 13:36:53,755 INFO    MainThread:899968 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:36:53,755 INFO    MainThread:899968 [wandb_setup.py:_flush():80] Configure stats pid to 899968
+2025-12-23 13:36:53,755 INFO    MainThread:899968 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:36:53,755 INFO    MainThread:899968 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:36:53,755 INFO    MainThread:899968 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:36:53,755 INFO    MainThread:899968 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/offline-run-20251223_133653-t8yclfex/logs/debug.log
+2025-12-23 13:36:53,755 INFO    MainThread:899968 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/offline-run-20251223_133653-t8yclfex/logs/debug-internal.log
+2025-12-23 13:36:53,756 INFO    MainThread:899968 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:36:53,756 INFO    MainThread:899968 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:36:53,756 INFO    MainThread:899968 [wandb_init.py:init():889] starting backend
+2025-12-23 13:36:54,205 INFO    MainThread:899968 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:36:54,210 INFO    MainThread:899968 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:36:54,213 INFO    MainThread:899968 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:36:54,217 INFO    MainThread:899968 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:36:54,323 INFO    MainThread:899968 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:36:54,424 INFO    MainThread:899968 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:36:54,424 INFO    MainThread:899968 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:36:54,424 INFO    MainThread:899968 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:36:54,424 INFO    MainThread:899968 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:36:54,426 INFO    MainThread:899968 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:36:57,105 INFO    wandb-AsyncioManager-main:899968 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:36:57,105 INFO    wandb-AsyncioManager-main:899968 [mailbox.py:close():137] Closing mailbox, abandoning 0 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/run-t8yclfex.wandb b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/run-t8yclfex.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..4a341cfed6df64f15ce564e6de0bf84e4475e3e5
Binary files /dev/null and b/sft_devstral_24B/wandb/offline-run-20251223_133653-t8yclfex/run-t8yclfex.wandb differ
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/files/requirements.txt b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug-core.log b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..a9b86e36c741da500aad410157b45a5a5ee649e0
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:37:50.440843766Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpmvl7h__3/port-900531.txt","pid":900531,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:37:50.441463065Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":900531}
+{"time":"2025-12-23T13:37:50.441448517Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-900531-900651-1897005423/socket","Net":"unix"}}
+{"time":"2025-12-23T13:37:50.627404782Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:37:50.634546017Z","level":"INFO","msg":"handleInformInit: received","streamId":"62isnb1e","id":"1(@)"}
+{"time":"2025-12-23T13:37:50.74193993Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"62isnb1e","id":"1(@)"}
+{"time":"2025-12-23T13:38:42.920178925Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:38:42.92025952Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:38:42.920253097Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:38:42.920324042Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:38:42.920372622Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-900531-900651-1897005423/socket","Net":"unix"}}
+{"time":"2025-12-23T13:38:42.923766245Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:38:42.923787648Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:38:42.923802311Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug-internal.log b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..d924c4416d550c36549c20a7b19721f7039ae5a2
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug-internal.log
@@ -0,0 +1,12 @@
+{"time":"2025-12-23T13:37:50.637326929Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:37:50.741664111Z","level":"WARN","msg":"featurechecker: GraphQL client is nil, skipping feature loading"}
+{"time":"2025-12-23T13:37:50.741732688Z","level":"INFO","msg":"stream: created new stream","id":"62isnb1e"}
+{"time":"2025-12-23T13:37:50.741794406Z","level":"INFO","msg":"handler: started","stream_id":"62isnb1e"}
+{"time":"2025-12-23T13:37:50.741926211Z","level":"INFO","msg":"stream: started","id":"62isnb1e"}
+{"time":"2025-12-23T13:37:50.741952982Z","level":"INFO","msg":"writer: started","stream_id":"62isnb1e"}
+{"time":"2025-12-23T13:37:50.741985097Z","level":"INFO","msg":"sender: started","stream_id":"62isnb1e"}
+{"time":"2025-12-23T13:37:50.742745975Z","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-12-23T13:38:42.920297297Z","level":"INFO","msg":"stream: closing","id":"62isnb1e"}
+{"time":"2025-12-23T13:38:42.920548453Z","level":"INFO","msg":"handler: closed","stream_id":"62isnb1e"}
+{"time":"2025-12-23T13:38:42.92060309Z","level":"INFO","msg":"sender: closed","stream_id":"62isnb1e"}
+{"time":"2025-12-23T13:38:42.92061497Z","level":"INFO","msg":"stream: closed","id":"62isnb1e"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug.log b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..4e43710c3c6a24e3a41f86b3b0b1cc1a00d15528
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/logs/debug.log
@@ -0,0 +1,26 @@
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_setup.py:_flush():80] Configure stats pid to 900531
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/offline-run-20251223_133750-62isnb1e/logs/debug.log
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/offline-run-20251223_133750-62isnb1e/logs/debug-internal.log
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:37:50,158 INFO    MainThread:900531 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:37:50,159 INFO    MainThread:900531 [wandb_init.py:init():889] starting backend
+2025-12-23 13:37:50,627 INFO    MainThread:900531 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:37:50,632 INFO    MainThread:900531 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:37:50,634 INFO    MainThread:900531 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:37:50,639 INFO    MainThread:900531 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:37:50,744 INFO    MainThread:900531 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:37:50,853 INFO    MainThread:900531 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:37:50,853 INFO    MainThread:900531 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:37:50,853 INFO    MainThread:900531 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:37:50,853 INFO    MainThread:900531 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:37:50,855 INFO    MainThread:900531 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:38:39,176 INFO    MainThread:900531 [wandb_run.py:_config_callback():1396] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.0', 'base_model_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'inference_mode': False, 'r': 64, 'target_modules': ['q_proj', 'v_proj', 'o_proj', 'k_proj'], 'exclude_modules': None, 'lora_alpha': 128, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': None, 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False}}, 'image_token_index': 10, 'projector_hidden_act': 'gelu', 'vision_feature_layer': -1, 'vision_config': {'hidden_size': 1024, 'intermediate_size': 4096, 'num_hidden_layers': 24, 'num_attention_heads': 16, 'num_channels': 3, 'patch_size': 14, 'image_size': 1540, 'attention_dropout': 0.0, 'hidden_act': 'silu', 'head_dim': 64, 'initializer_range': 0.02, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'pixtral', 'output_attentions': False}, 'text_config': {'vocab_size': 131072, 'max_position_embeddings': 393216, 'hidden_size': 5120, 'intermediate_size': 32768, 'num_hidden_layers': 40, 'num_attention_heads': 32, 'sliding_window': None, 'head_dim': 128, 'num_key_value_heads': 8, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'use_cache': True, 'attention_dropout': 0.0, 'rope_parameters': {'beta_fast': 32.0, 'beta_slow': 1.0, 'factor': 48.0, 'llama_4_scaling_beta': 0.1, 'mscale': 1.0, 'mscale_all_dim': 1.0, 'original_max_position_embeddings': 8192, 'rope_theta': 100000000.0, 'rope_type': 'yarn', 'type': 'yarn'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 11, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'ministral3', 'output_attentions': False}, 'multimodal_projector_bias': False, 'spatial_merge_size': 2, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': ['Mistral3ForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'transformers_version': '5.0.0.dev0', 'model_type': 'mistral3', 'use_cache': False, 'output_attentions': False, 'output_dir': 'runs/instruct_run_24b/checkpoints', 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 5e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 6.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.01, 'warmup_steps': 0.01, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': None, 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 2, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'enable_jit_checkpoint': False, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'use_cpu': False, 'seed': 42, 'data_seed': None, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': -1, 'ddp_backend': None, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': 100, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'run_name': None, 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': True, 'metric_for_best_model': 'eval_loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'auto_find_batch_size': False, 'full_determinism': False, 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True}
+2025-12-23 13:38:39,186 INFO    MainThread:900531 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 24102587392 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x7adf6f345150>>
+2025-12-23 13:38:39,186 INFO    MainThread:900531 [wandb_run.py:_config_callback():1396] config_cb model/num_parameters 24102587392 None
+2025-12-23 13:38:42,920 INFO    wandb-AsyncioManager-main:900531 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:38:42,920 INFO    wandb-AsyncioManager-main:900531 [mailbox.py:close():137] Closing mailbox, abandoning 0 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/run-62isnb1e.wandb b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/run-62isnb1e.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..8f1ec07e97946afa3ee36b9bc346ad6839d9586a
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_133750-62isnb1e/run-62isnb1e.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11456a28369ff08d724d0175625bb6af9f927fe32f0cd99157b201c23b74a1d4
+size 233270
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/files/requirements.txt b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug-core.log b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..fac066c16c13af87ad595dbda1c795441e20302f
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:40:33.684544708Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpb1565w4z/port-901979.txt","pid":901979,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:40:33.68516663Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":901979}
+{"time":"2025-12-23T13:40:33.685157587Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-901979-902257-1112620819/socket","Net":"unix"}}
+{"time":"2025-12-23T13:40:33.871178595Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:40:33.877832014Z","level":"INFO","msg":"handleInformInit: received","streamId":"mifdnkpc","id":"1(@)"}
+{"time":"2025-12-23T13:40:33.984977409Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"mifdnkpc","id":"1(@)"}
+{"time":"2025-12-23T13:40:59.232123491Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:40:59.232254745Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:40:59.232236338Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:40:59.232372368Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:40:59.232374173Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-901979-902257-1112620819/socket","Net":"unix"}}
+{"time":"2025-12-23T13:40:59.233360565Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:40:59.233394522Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:40:59.233409074Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug-internal.log b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..7a1388d2677260396437cc62ce0cb3c3b1d0de2b
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug-internal.log
@@ -0,0 +1,12 @@
+{"time":"2025-12-23T13:40:33.880649779Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:40:33.984721957Z","level":"WARN","msg":"featurechecker: GraphQL client is nil, skipping feature loading"}
+{"time":"2025-12-23T13:40:33.98480273Z","level":"INFO","msg":"stream: created new stream","id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:33.984866006Z","level":"INFO","msg":"handler: started","stream_id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:33.984968055Z","level":"INFO","msg":"stream: started","id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:33.984991606Z","level":"INFO","msg":"writer: started","stream_id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:33.985006821Z","level":"INFO","msg":"sender: started","stream_id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:33.985666188Z","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-12-23T13:40:59.232210988Z","level":"INFO","msg":"stream: closing","id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:59.232477353Z","level":"INFO","msg":"handler: closed","stream_id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:59.232587191Z","level":"INFO","msg":"sender: closed","stream_id":"mifdnkpc"}
+{"time":"2025-12-23T13:40:59.232606717Z","level":"INFO","msg":"stream: closed","id":"mifdnkpc"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug.log b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..544fe34cbfe065c333965c8801bcbd250accc393
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug.log
@@ -0,0 +1,26 @@
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_setup.py:_flush():80] Configure stats pid to 901979
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug.log
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/offline-run-20251223_134033-mifdnkpc/logs/debug-internal.log
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:40:33,420 INFO    MainThread:901979 [wandb_init.py:init():889] starting backend
+2025-12-23 13:40:33,871 INFO    MainThread:901979 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:40:33,876 INFO    MainThread:901979 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:40:33,877 INFO    MainThread:901979 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:40:33,882 INFO    MainThread:901979 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:40:33,987 INFO    MainThread:901979 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:40:34,090 INFO    MainThread:901979 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:40:34,090 INFO    MainThread:901979 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:40:34,090 INFO    MainThread:901979 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:40:34,090 INFO    MainThread:901979 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:40:34,092 INFO    MainThread:901979 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:40:54,425 INFO    MainThread:901979 [wandb_run.py:_config_callback():1396] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.0', 'base_model_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'inference_mode': False, 'r': 64, 'target_modules': ['k_proj', 'q_proj', 'o_proj', 'v_proj'], 'exclude_modules': None, 'lora_alpha': 128, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': None, 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False}}, 'image_token_index': 10, 'projector_hidden_act': 'gelu', 'vision_feature_layer': -1, 'vision_config': {'hidden_size': 1024, 'intermediate_size': 4096, 'num_hidden_layers': 24, 'num_attention_heads': 16, 'num_channels': 3, 'patch_size': 14, 'image_size': 1540, 'attention_dropout': 0.0, 'hidden_act': 'silu', 'head_dim': 64, 'initializer_range': 0.02, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'pixtral', 'output_attentions': False}, 'text_config': {'vocab_size': 131072, 'max_position_embeddings': 393216, 'hidden_size': 5120, 'intermediate_size': 32768, 'num_hidden_layers': 40, 'num_attention_heads': 32, 'sliding_window': None, 'head_dim': 128, 'num_key_value_heads': 8, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'use_cache': True, 'attention_dropout': 0.0, 'rope_parameters': {'beta_fast': 32.0, 'beta_slow': 1.0, 'factor': 48.0, 'llama_4_scaling_beta': 0.1, 'mscale': 1.0, 'mscale_all_dim': 1.0, 'original_max_position_embeddings': 8192, 'rope_theta': 100000000.0, 'rope_type': 'yarn', 'type': 'yarn'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 11, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'ministral3', 'output_attentions': False}, 'multimodal_projector_bias': False, 'spatial_merge_size': 2, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': ['Mistral3ForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'transformers_version': '5.0.0.dev0', 'model_type': 'mistral3', 'use_cache': False, 'output_attentions': False, 'output_dir': 'runs/instruct_run_24b/checkpoints', 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 5e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 6.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.01, 'warmup_steps': 0.01, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': None, 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 2, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'enable_jit_checkpoint': False, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'use_cpu': False, 'seed': 42, 'data_seed': None, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': -1, 'ddp_backend': None, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': 100, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'run_name': None, 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': True, 'metric_for_best_model': 'eval_loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'auto_find_batch_size': False, 'full_determinism': False, 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True}
+2025-12-23 13:40:54,436 INFO    MainThread:901979 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 24102587392 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x7b405e1fe080>>
+2025-12-23 13:40:54,436 INFO    MainThread:901979 [wandb_run.py:_config_callback():1396] config_cb model/num_parameters 24102587392 None
+2025-12-23 13:40:59,232 INFO    wandb-AsyncioManager-main:901979 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:40:59,232 INFO    wandb-AsyncioManager-main:901979 [mailbox.py:close():137] Closing mailbox, abandoning 0 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/run-mifdnkpc.wandb b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/run-mifdnkpc.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..fc9333845b86f3084b04689827e81862a8b618df
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134033-mifdnkpc/run-mifdnkpc.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7054b141d3f5b52a16fe0bca3b815352934edd9cc70961cd77fd626368360ad7
+size 222458
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/files/requirements.txt b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug-core.log b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..65285adfe108b6bbc69484f2e27eed7d3758b4a2
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:42:37.095468401Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpovlys2zm/port-903261.txt","pid":903261,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:42:37.096085989Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":903261}
+{"time":"2025-12-23T13:42:37.096073341Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-903261-903398-3954200984/socket","Net":"unix"}}
+{"time":"2025-12-23T13:42:37.281397511Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:42:37.288883109Z","level":"INFO","msg":"handleInformInit: received","streamId":"sicxj35d","id":"1(@)"}
+{"time":"2025-12-23T13:42:37.396858593Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"sicxj35d","id":"1(@)"}
+{"time":"2025-12-23T13:43:23.5354143Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:43:23.535493351Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:43:23.535526488Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:43:23.53560987Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:43:23.53569769Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-903261-903398-3954200984/socket","Net":"unix"}}
+{"time":"2025-12-23T13:43:23.536647501Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:43:23.536691414Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:43:23.536710026Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug-internal.log b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..5a90e073acb5de29f13075fb39f6c9884899381e
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug-internal.log
@@ -0,0 +1,12 @@
+{"time":"2025-12-23T13:42:37.291828328Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:42:37.396548602Z","level":"WARN","msg":"featurechecker: GraphQL client is nil, skipping feature loading"}
+{"time":"2025-12-23T13:42:37.396641226Z","level":"INFO","msg":"stream: created new stream","id":"sicxj35d"}
+{"time":"2025-12-23T13:42:37.396706199Z","level":"INFO","msg":"handler: started","stream_id":"sicxj35d"}
+{"time":"2025-12-23T13:42:37.396848979Z","level":"INFO","msg":"stream: started","id":"sicxj35d"}
+{"time":"2025-12-23T13:42:37.396874845Z","level":"INFO","msg":"writer: started","stream_id":"sicxj35d"}
+{"time":"2025-12-23T13:42:37.396897157Z","level":"INFO","msg":"sender: started","stream_id":"sicxj35d"}
+{"time":"2025-12-23T13:42:37.397870839Z","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-12-23T13:43:23.535509152Z","level":"INFO","msg":"stream: closing","id":"sicxj35d"}
+{"time":"2025-12-23T13:43:23.535772681Z","level":"INFO","msg":"handler: closed","stream_id":"sicxj35d"}
+{"time":"2025-12-23T13:43:23.535835667Z","level":"INFO","msg":"sender: closed","stream_id":"sicxj35d"}
+{"time":"2025-12-23T13:43:23.535847775Z","level":"INFO","msg":"stream: closed","id":"sicxj35d"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug.log b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..74f10e264f8010e4b36846036a0eeed3dea7e85c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/logs/debug.log
@@ -0,0 +1,26 @@
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_setup.py:_flush():80] Configure stats pid to 903261
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/offline-run-20251223_134236-sicxj35d/logs/debug.log
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/offline-run-20251223_134236-sicxj35d/logs/debug-internal.log
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': False, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:42:36,827 INFO    MainThread:903261 [wandb_init.py:init():889] starting backend
+2025-12-23 13:42:37,281 INFO    MainThread:903261 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:42:37,286 INFO    MainThread:903261 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:42:37,289 INFO    MainThread:903261 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:42:37,294 INFO    MainThread:903261 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:42:37,399 INFO    MainThread:903261 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:42:37,501 INFO    MainThread:903261 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:42:37,501 INFO    MainThread:903261 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:42:37,501 INFO    MainThread:903261 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:42:37,501 INFO    MainThread:903261 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:42:37,503 INFO    MainThread:903261 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:43:22,647 INFO    MainThread:903261 [wandb_run.py:_config_callback():1396] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.0', 'base_model_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'inference_mode': False, 'r': 64, 'target_modules': ['o_proj', 'q_proj', 'v_proj', 'k_proj'], 'exclude_modules': None, 'lora_alpha': 128, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': None, 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False}}, 'image_token_index': 10, 'projector_hidden_act': 'gelu', 'vision_feature_layer': -1, 'vision_config': {'hidden_size': 1024, 'intermediate_size': 4096, 'num_hidden_layers': 24, 'num_attention_heads': 16, 'num_channels': 3, 'patch_size': 14, 'image_size': 1540, 'attention_dropout': 0.0, 'hidden_act': 'silu', 'head_dim': 64, 'initializer_range': 0.02, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'pixtral', 'output_attentions': False}, 'text_config': {'vocab_size': 131072, 'max_position_embeddings': 393216, 'hidden_size': 5120, 'intermediate_size': 32768, 'num_hidden_layers': 40, 'num_attention_heads': 32, 'sliding_window': None, 'head_dim': 128, 'num_key_value_heads': 8, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'use_cache': True, 'attention_dropout': 0.0, 'rope_parameters': {'beta_fast': 32.0, 'beta_slow': 1.0, 'factor': 48.0, 'llama_4_scaling_beta': 0.1, 'mscale': 1.0, 'mscale_all_dim': 1.0, 'original_max_position_embeddings': 8192, 'rope_theta': 100000000.0, 'rope_type': 'yarn', 'type': 'yarn'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 11, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'ministral3', 'output_attentions': False}, 'multimodal_projector_bias': False, 'spatial_merge_size': 2, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': ['Mistral3ForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'transformers_version': '5.0.0.dev0', 'model_type': 'mistral3', 'use_cache': False, 'output_attentions': False, 'output_dir': 'runs/instruct_run_24b/checkpoints', 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 5e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 6.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.01, 'warmup_steps': 0.01, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': None, 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 2, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'enable_jit_checkpoint': False, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'use_cpu': False, 'seed': 42, 'data_seed': None, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': -1, 'ddp_backend': None, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': 100, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'run_name': None, 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': True, 'metric_for_best_model': 'eval_loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'auto_find_batch_size': False, 'full_determinism': False, 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True}
+2025-12-23 13:43:22,659 INFO    MainThread:903261 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 24102587392 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x794f55382050>>
+2025-12-23 13:43:22,659 INFO    MainThread:903261 [wandb_run.py:_config_callback():1396] config_cb model/num_parameters 24102587392 None
+2025-12-23 13:43:23,535 INFO    wandb-AsyncioManager-main:903261 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:43:23,535 INFO    wandb-AsyncioManager-main:903261 [mailbox.py:close():137] Closing mailbox, abandoning 0 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/run-sicxj35d.wandb b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/run-sicxj35d.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..f851c27c384a81b4ccb65a4c4d6cd2ca72ea32b5
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134236-sicxj35d/run-sicxj35d.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7980e05db931e8638a2c246c503820204d247fe9fc8d7f61ca594dd4c3cffb7
+size 246049
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/files/requirements.txt b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug-core.log b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..042544124782a0f38786f8423f268030b27e55ee
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:44:33.059029647Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp1wy5r270/port-904661.txt","pid":904661,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:44:33.059712446Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":904661}
+{"time":"2025-12-23T13:44:33.059687834Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-904661-904776-1525747917/socket","Net":"unix"}}
+{"time":"2025-12-23T13:44:33.245136893Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:44:33.252845836Z","level":"INFO","msg":"handleInformInit: received","streamId":"y3kepwdy","id":"1(@)"}
+{"time":"2025-12-23T13:44:33.360486149Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"y3kepwdy","id":"1(@)"}
+{"time":"2025-12-23T13:45:49.779741883Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:45:49.779822859Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:45:49.779815242Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:45:49.779939269Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-904661-904776-1525747917/socket","Net":"unix"}}
+{"time":"2025-12-23T13:45:49.779970923Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:45:49.780969653Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:45:49.780994962Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:45:49.781007111Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug-internal.log b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..8ce543c8f7e9b99bbf26e6838ac70fbac6b8071e
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug-internal.log
@@ -0,0 +1,12 @@
+{"time":"2025-12-23T13:44:33.255858317Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:44:33.360166544Z","level":"WARN","msg":"featurechecker: GraphQL client is nil, skipping feature loading"}
+{"time":"2025-12-23T13:44:33.360271068Z","level":"INFO","msg":"stream: created new stream","id":"y3kepwdy"}
+{"time":"2025-12-23T13:44:33.360371301Z","level":"INFO","msg":"handler: started","stream_id":"y3kepwdy"}
+{"time":"2025-12-23T13:44:33.360475996Z","level":"INFO","msg":"stream: started","id":"y3kepwdy"}
+{"time":"2025-12-23T13:44:33.360494414Z","level":"INFO","msg":"writer: started","stream_id":"y3kepwdy"}
+{"time":"2025-12-23T13:44:33.360499005Z","level":"INFO","msg":"sender: started","stream_id":"y3kepwdy"}
+{"time":"2025-12-23T13:44:33.362348014Z","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-12-23T13:45:49.779834201Z","level":"INFO","msg":"stream: closing","id":"y3kepwdy"}
+{"time":"2025-12-23T13:45:49.780068484Z","level":"INFO","msg":"handler: closed","stream_id":"y3kepwdy"}
+{"time":"2025-12-23T13:45:49.780114136Z","level":"INFO","msg":"sender: closed","stream_id":"y3kepwdy"}
+{"time":"2025-12-23T13:45:49.780128579Z","level":"INFO","msg":"stream: closed","id":"y3kepwdy"}
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug.log b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..b1f633922d3655c8136fa8dbd49dc93e99a64f3a
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug.log
@@ -0,0 +1,26 @@
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_setup.py:_flush():80] Configure stats pid to 904661
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug.log
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/offline-run-20251223_134432-y3kepwdy/logs/debug-internal.log
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:44:32,783 INFO    MainThread:904661 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:44:32,784 INFO    MainThread:904661 [wandb_init.py:init():889] starting backend
+2025-12-23 13:44:33,245 INFO    MainThread:904661 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:44:33,250 INFO    MainThread:904661 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:44:33,251 INFO    MainThread:904661 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:44:33,256 INFO    MainThread:904661 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:44:33,364 INFO    MainThread:904661 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:44:33,464 INFO    MainThread:904661 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:44:33,464 INFO    MainThread:904661 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:44:33,464 INFO    MainThread:904661 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:44:33,464 INFO    MainThread:904661 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:44:33,466 INFO    MainThread:904661 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:44:56,307 INFO    MainThread:904661 [wandb_run.py:_config_callback():1396] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.0', 'base_model_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'inference_mode': False, 'r': 64, 'target_modules': ['k_proj', 'q_proj', 'o_proj', 'v_proj'], 'exclude_modules': None, 'lora_alpha': 128, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': None, 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False}}, 'image_token_index': 10, 'projector_hidden_act': 'gelu', 'vision_feature_layer': -1, 'vision_config': {'hidden_size': 1024, 'intermediate_size': 4096, 'num_hidden_layers': 24, 'num_attention_heads': 16, 'num_channels': 3, 'patch_size': 14, 'image_size': 1540, 'attention_dropout': 0.0, 'hidden_act': 'silu', 'head_dim': 64, 'initializer_range': 0.02, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'pixtral', 'output_attentions': False}, 'text_config': {'vocab_size': 131072, 'max_position_embeddings': 393216, 'hidden_size': 5120, 'intermediate_size': 32768, 'num_hidden_layers': 40, 'num_attention_heads': 32, 'sliding_window': None, 'head_dim': 128, 'num_key_value_heads': 8, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'use_cache': True, 'attention_dropout': 0.0, 'rope_parameters': {'beta_fast': 32.0, 'beta_slow': 1.0, 'factor': 48.0, 'llama_4_scaling_beta': 0.1, 'mscale': 1.0, 'mscale_all_dim': 1.0, 'original_max_position_embeddings': 8192, 'rope_theta': 100000000.0, 'rope_type': 'yarn', 'type': 'yarn'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 11, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'ministral3', 'output_attentions': False}, 'multimodal_projector_bias': False, 'spatial_merge_size': 2, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': ['Mistral3ForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'transformers_version': '5.0.0.dev0', 'model_type': 'mistral3', 'use_cache': False, 'output_attentions': False, 'output_dir': 'runs/instruct_run_24b/checkpoints', 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 5e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 6.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.01, 'warmup_steps': 0.01, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': None, 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 2, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'enable_jit_checkpoint': False, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'use_cpu': False, 'seed': 42, 'data_seed': None, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': -1, 'ddp_backend': None, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': 100, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'run_name': None, 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': True, 'metric_for_best_model': 'eval_loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'auto_find_batch_size': False, 'full_determinism': False, 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True}
+2025-12-23 13:44:56,317 INFO    MainThread:904661 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 24102587392 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x78a0bc6ce020>>
+2025-12-23 13:44:56,317 INFO    MainThread:904661 [wandb_run.py:_config_callback():1396] config_cb model/num_parameters 24102587392 None
+2025-12-23 13:45:49,779 INFO    wandb-AsyncioManager-main:904661 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:45:49,780 INFO    wandb-AsyncioManager-main:904661 [mailbox.py:close():137] Closing mailbox, abandoning 0 handles.
diff --git a/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/run-y3kepwdy.wandb b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/run-y3kepwdy.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..1637286d49d66c6be13bbe631f2f8596bb003890
--- /dev/null
+++ b/sft_devstral_24B/wandb/offline-run-20251223_134432-y3kepwdy/run-y3kepwdy.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b8dab90261604d87bc592555c752e23d6046c9d0efa219cdc15beec7c676f32
+size 231104
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/config.yaml b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/config.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..2deb202a064b74f91e3febe6d2f97bb4c4dccf8c
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/config.yaml
@@ -0,0 +1,176 @@
+_wandb:
+    value:
+        cli_version: 0.23.1
+        e:
+            c4qlfqsdsdn87bdtewyqd1101bckxv0h:
+                args:
+                    - --config
+                    - ./SFT/config_instruct.yaml
+                codePath: SFT/run_instruct.py
+                codePathLocal: SFT/run_instruct.py
+                cpu_count: 12
+                cpu_count_logical: 24
+                cudaVersion: "13.0"
+                disk:
+                    /:
+                        total: "791251738624"
+                        used: "328812568576"
+                email: shaiksirajuddin9949@gmail.com
+                executable: /workspace/CPT-env/bin/python
+                git:
+                    commit: 92382202abe01938bfeff39a3bf7b82e6e786325
+                    remote: https://github.com/Ashx098/Trainer-kit.git
+                gpu: NVIDIA A100-SXM4-80GB
+                gpu_count: 2
+                gpu_nvidia:
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40
+                host: a100-2gpu-shell-session-9c959f5f-mx8rm
+                memory:
+                    total: "359047892992"
+                os: Linux-6.12.46+-x86_64-with-glibc2.35
+                program: /workspace/Trainer-kit/SFT/run_instruct.py
+                python: CPython 3.10.12
+                root: runs/instruct_run_24b
+                startedAt: "2025-12-23T13:31:36.896493Z"
+                writerId: c4qlfqsdsdn87bdtewyqd1101bckxv0h
+        m: []
+        python_version: 3.10.12
+        t:
+            "1":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "2":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "3":
+                - 15
+                - 16
+            "4": 3.10.12
+            "5": 0.23.1
+            "6": 5.0.0.dev0
+            "12": 0.23.1
+            "13": linux-x86_64
+data:
+    value:
+        custom_template: |-
+            ##INSTRUCTION
+            {instruction}<|im_end|>
+            ##TASK
+            {input}<|im_end|>
+            ##OUTPUT
+            {output}<|im_end|>
+        eval_jsonl: null
+        eval_split_ratio: 0.1
+        format_type: custom
+        input_field: input
+        instruction_field: instruction
+        max_length: 2048
+        num_proc: 4
+        output_field: output
+        shuffle: true
+        system_prompt: |
+            You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.
+
+            ## Output Format
+
+            ##OUTPUT
+            Explain the data flow and why each component must change:
+            - Flow: [Input → Processing → Output with arrows]
+            - For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"
+            - Explain coupling between components
+
+            ##SELECT
+            modify::crates/path/to/file.rs::impl::ComponentName
+            add::crates/another/file.rs::function::AnotherComponent
+            <EOS>
+
+            ## Rules
+
+            1. Use full paths: `remove::crates/folder/file.rs::Type::Name`
+            2. Use `::` for nested items: `status::StructName::Type::Name`
+            3. Always explain "must change because" and "without this"
+            3. Types of components: function, struct, enum, impl, trait
+            4. If there is extra information (e.g., enum variants), include that too.
+            5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>
+
+            ## Example
+
+            ##TASK
+            Add webhook subscription support
+
+            ##OUTPUT
+            The webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don't trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.
+
+            ##SELECT
+            crates/common_enums/src/enums.rs::EventClass
+            crates/common_enums/src/transformers.rs::SubscriptionStatus
+            <EOS>
+        train_jsonl: ../sft_dataset.jsonl
+model:
+    value:
+        attn_implementation: null
+        base_local_dir: base_model
+        bnb_4bit_compute_dtype: bfloat16
+        bnb_4bit_quant_type: nf4
+        bnb_4bit_use_double_quant: false
+        device_map: auto
+        repo_id: ./CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+        revision: null
+        tokenizer_use_fast: true
+        torch_dtype: bfloat16
+        trust_remote_code: true
+        use_4bit: false
+peft:
+    value:
+        bias: none
+        enabled: true
+        lora_alpha: 128
+        lora_dropout: 0.05
+        r: 64
+        target_modules: auto
+run_dir:
+    value: runs/instruct_run_24b
+train:
+    value:
+        early_stopping:
+            enabled: true
+            metric: eval_loss
+            min_delta: 0.001
+            mode: min
+            patience: 3
+        eval_steps: 100
+        evaluation_strategy: steps
+        gradient_accumulation_steps: 8
+        gradient_checkpointing: true
+        learning_rate: "5e-5"
+        load_best_model_at_end: true
+        logging_steps: 2
+        lr_scheduler_type: cosine
+        max_grad_norm: 1
+        num_train_epochs: 6
+        optim: adamw_torch
+        per_device_eval_batch_size: 1
+        per_device_train_batch_size: 1
+        resume_from_checkpoint: auto
+        save_steps: 500
+        save_strategy: steps
+        save_total_limit: 20
+        warmup_ratio: 0.01
+        weight_decay: 0
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/output.log b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/output.log
new file mode 100644
index 0000000000000000000000000000000000000000..72aac50700198ec7cf75c34e87e65c9691381c02
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/output.log
@@ -0,0 +1,44 @@
+Wandb initialized: project='sft-training', name='auto-generated'
+`torch_dtype` is deprecated! Use `dtype` instead!
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 535, in load_base_model_and_tokenizer
+    model = AutoModelForCausalLM.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 376, in from_pretrained
+    raise ValueError(
+ValueError: Unrecognized configuration class <class 'transformers.models.mistral3.configuration_mistral3.Mistral3Config'> for this kind of AutoModel: AutoModelForCausalLM.
+Model type should be one of AfmoeConfig, ApertusConfig, ArceeConfig, AriaTextConfig, BambaConfig, BartConfig, BertConfig, BertGenerationConfig, BigBirdConfig, BigBirdPegasusConfig, BioGptConfig, BitNetConfig, BlenderbotConfig, BlenderbotSmallConfig, BloomConfig, BltConfig, CamembertConfig, LlamaConfig, CodeGenConfig, CohereConfig, Cohere2Config, CpmAntConfig, CTRLConfig, CwmConfig, Data2VecTextConfig, DbrxConfig, DeepseekV2Config, DeepseekV3Config, DiffLlamaConfig, DogeConfig, Dots1Config, ElectraConfig, Emu3Config, ErnieConfig, Ernie4_5Config, Ernie4_5_MoeConfig, Exaone4Config, FalconConfig, FalconH1Config, FalconMambaConfig, FlexOlmoConfig, FuyuConfig, GemmaConfig, Gemma2Config, Gemma3Config, Gemma3TextConfig, Gemma3nConfig, Gemma3nTextConfig, GitConfig, GlmConfig, Glm4Config, Glm4MoeConfig, GotOcr2Config, GPT2Config, GPT2Config, GPTBigCodeConfig, GPTNeoConfig, GPTNeoXConfig, GPTNeoXJapaneseConfig, GptOssConfig, GPTJConfig, GraniteConfig, GraniteMoeConfig, GraniteMoeHybridConfig, GraniteMoeSharedConfig, HeliumConfig, HunYuanDenseV1Config, HunYuanMoEV1Config, Jais2Config, JambaConfig, JetMoeConfig, Lfm2Config, Lfm2MoeConfig, LlamaConfig, Llama4Config, Llama4TextConfig, LongcatFlashConfig, MambaConfig, Mamba2Config, MarianConfig, MBartConfig, MegatronBertConfig, MiniMaxConfig, MinistralConfig, Ministral3Config, MistralConfig, MixtralConfig, MllamaConfig, ModernBertDecoderConfig, MoshiConfig, MptConfig, MusicgenConfig, MusicgenMelodyConfig, MvpConfig, NanoChatConfig, NemotronConfig, OlmoConfig, Olmo2Config, Olmo3Config, OlmoeConfig, OpenAIGPTConfig, OPTConfig, PegasusConfig, PersimmonConfig, PhiConfig, Phi3Config, Phi4MultimodalConfig, PhimoeConfig, PLBartConfig, ProphetNetConfig, Qwen2Config, Qwen2MoeConfig, Qwen3Config, Qwen3MoeConfig, Qwen3NextConfig, RecurrentGemmaConfig, ReformerConfig, RemBertConfig, RobertaConfig, RobertaPreLayerNormConfig, RoCBertConfig, RoFormerConfig, RwkvConfig, SeedOssConfig, SmolLM3Config, StableLmConfig, Starcoder2Config, TrOCRConfig, VaultGemmaConfig, WhisperConfig, XGLMConfig, XLMConfig, XLMRobertaConfig, XLMRobertaXLConfig, XLNetConfig, xLSTMConfig, XmodConfig, ZambaConfig, Zamba2Config.
+
+During handling of the above exception, another exception occurred:
+
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 844, in <module>
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 710, in main
+    model, tokenizer = load_base_model_and_tokenizer(cfg, base_dir)
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 548, in load_base_model_and_tokenizer
+    model = AutoModelForCausalLM.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 376, in from_pretrained
+    raise ValueError(
+ValueError: Unrecognized configuration class <class 'transformers.models.mistral3.configuration_mistral3.Mistral3Config'> for this kind of AutoModel: AutoModelForCausalLM.
+Model type should be one of AfmoeConfig, ApertusConfig, ArceeConfig, AriaTextConfig, BambaConfig, BartConfig, BertConfig, BertGenerationConfig, BigBirdConfig, BigBirdPegasusConfig, BioGptConfig, BitNetConfig, BlenderbotConfig, BlenderbotSmallConfig, BloomConfig, BltConfig, CamembertConfig, LlamaConfig, CodeGenConfig, CohereConfig, Cohere2Config, CpmAntConfig, CTRLConfig, CwmConfig, Data2VecTextConfig, DbrxConfig, DeepseekV2Config, DeepseekV3Config, DiffLlamaConfig, DogeConfig, Dots1Config, ElectraConfig, Emu3Config, ErnieConfig, Ernie4_5Config, Ernie4_5_MoeConfig, Exaone4Config, FalconConfig, FalconH1Config, FalconMambaConfig, FlexOlmoConfig, FuyuConfig, GemmaConfig, Gemma2Config, Gemma3Config, Gemma3TextConfig, Gemma3nConfig, Gemma3nTextConfig, GitConfig, GlmConfig, Glm4Config, Glm4MoeConfig, GotOcr2Config, GPT2Config, GPT2Config, GPTBigCodeConfig, GPTNeoConfig, GPTNeoXConfig, GPTNeoXJapaneseConfig, GptOssConfig, GPTJConfig, GraniteConfig, GraniteMoeConfig, GraniteMoeHybridConfig, GraniteMoeSharedConfig, HeliumConfig, HunYuanDenseV1Config, HunYuanMoEV1Config, Jais2Config, JambaConfig, JetMoeConfig, Lfm2Config, Lfm2MoeConfig, LlamaConfig, Llama4Config, Llama4TextConfig, LongcatFlashConfig, MambaConfig, Mamba2Config, MarianConfig, MBartConfig, MegatronBertConfig, MiniMaxConfig, MinistralConfig, Ministral3Config, MistralConfig, MixtralConfig, MllamaConfig, ModernBertDecoderConfig, MoshiConfig, MptConfig, MusicgenConfig, MusicgenMelodyConfig, MvpConfig, NanoChatConfig, NemotronConfig, OlmoConfig, Olmo2Config, Olmo3Config, OlmoeConfig, OpenAIGPTConfig, OPTConfig, PegasusConfig, PersimmonConfig, PhiConfig, Phi3Config, Phi4MultimodalConfig, PhimoeConfig, PLBartConfig, ProphetNetConfig, Qwen2Config, Qwen2MoeConfig, Qwen3Config, Qwen3MoeConfig, Qwen3NextConfig, RecurrentGemmaConfig, ReformerConfig, RemBertConfig, RobertaConfig, RobertaPreLayerNormConfig, RoCBertConfig, RoFormerConfig, RwkvConfig, SeedOssConfig, SmolLM3Config, StableLmConfig, Starcoder2Config, TrOCRConfig, VaultGemmaConfig, WhisperConfig, XGLMConfig, XLMConfig, XLMRobertaConfig, XLMRobertaXLConfig, XLNetConfig, xLSTMConfig, XmodConfig, ZambaConfig, Zamba2Config.
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 535, in load_base_model_and_tokenizer
+    model = AutoModelForCausalLM.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 376, in from_pretrained
+    raise ValueError(
+ValueError: Unrecognized configuration class <class 'transformers.models.mistral3.configuration_mistral3.Mistral3Config'> for this kind of AutoModel: AutoModelForCausalLM.
+Model type should be one of AfmoeConfig, ApertusConfig, ArceeConfig, AriaTextConfig, BambaConfig, BartConfig, BertConfig, BertGenerationConfig, BigBirdConfig, BigBirdPegasusConfig, BioGptConfig, BitNetConfig, BlenderbotConfig, BlenderbotSmallConfig, BloomConfig, BltConfig, CamembertConfig, LlamaConfig, CodeGenConfig, CohereConfig, Cohere2Config, CpmAntConfig, CTRLConfig, CwmConfig, Data2VecTextConfig, DbrxConfig, DeepseekV2Config, DeepseekV3Config, DiffLlamaConfig, DogeConfig, Dots1Config, ElectraConfig, Emu3Config, ErnieConfig, Ernie4_5Config, Ernie4_5_MoeConfig, Exaone4Config, FalconConfig, FalconH1Config, FalconMambaConfig, FlexOlmoConfig, FuyuConfig, GemmaConfig, Gemma2Config, Gemma3Config, Gemma3TextConfig, Gemma3nConfig, Gemma3nTextConfig, GitConfig, GlmConfig, Glm4Config, Glm4MoeConfig, GotOcr2Config, GPT2Config, GPT2Config, GPTBigCodeConfig, GPTNeoConfig, GPTNeoXConfig, GPTNeoXJapaneseConfig, GptOssConfig, GPTJConfig, GraniteConfig, GraniteMoeConfig, GraniteMoeHybridConfig, GraniteMoeSharedConfig, HeliumConfig, HunYuanDenseV1Config, HunYuanMoEV1Config, Jais2Config, JambaConfig, JetMoeConfig, Lfm2Config, Lfm2MoeConfig, LlamaConfig, Llama4Config, Llama4TextConfig, LongcatFlashConfig, MambaConfig, Mamba2Config, MarianConfig, MBartConfig, MegatronBertConfig, MiniMaxConfig, MinistralConfig, Ministral3Config, MistralConfig, MixtralConfig, MllamaConfig, ModernBertDecoderConfig, MoshiConfig, MptConfig, MusicgenConfig, MusicgenMelodyConfig, MvpConfig, NanoChatConfig, NemotronConfig, OlmoConfig, Olmo2Config, Olmo3Config, OlmoeConfig, OpenAIGPTConfig, OPTConfig, PegasusConfig, PersimmonConfig, PhiConfig, Phi3Config, Phi4MultimodalConfig, PhimoeConfig, PLBartConfig, ProphetNetConfig, Qwen2Config, Qwen2MoeConfig, Qwen3Config, Qwen3MoeConfig, Qwen3NextConfig, RecurrentGemmaConfig, ReformerConfig, RemBertConfig, RobertaConfig, RobertaPreLayerNormConfig, RoCBertConfig, RoFormerConfig, RwkvConfig, SeedOssConfig, SmolLM3Config, StableLmConfig, Starcoder2Config, TrOCRConfig, VaultGemmaConfig, WhisperConfig, XGLMConfig, XLMConfig, XLMRobertaConfig, XLMRobertaXLConfig, XLNetConfig, xLSTMConfig, XmodConfig, ZambaConfig, Zamba2Config.
+
+During handling of the above exception, another exception occurred:
+
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 844, in <module>
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 710, in main
+    model, tokenizer = load_base_model_and_tokenizer(cfg, base_dir)
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 548, in load_base_model_and_tokenizer
+    model = AutoModelForCausalLM.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 376, in from_pretrained
+    raise ValueError(
+ValueError: Unrecognized configuration class <class 'transformers.models.mistral3.configuration_mistral3.Mistral3Config'> for this kind of AutoModel: AutoModelForCausalLM.
+Model type should be one of AfmoeConfig, ApertusConfig, ArceeConfig, AriaTextConfig, BambaConfig, BartConfig, BertConfig, BertGenerationConfig, BigBirdConfig, BigBirdPegasusConfig, BioGptConfig, BitNetConfig, BlenderbotConfig, BlenderbotSmallConfig, BloomConfig, BltConfig, CamembertConfig, LlamaConfig, CodeGenConfig, CohereConfig, Cohere2Config, CpmAntConfig, CTRLConfig, CwmConfig, Data2VecTextConfig, DbrxConfig, DeepseekV2Config, DeepseekV3Config, DiffLlamaConfig, DogeConfig, Dots1Config, ElectraConfig, Emu3Config, ErnieConfig, Ernie4_5Config, Ernie4_5_MoeConfig, Exaone4Config, FalconConfig, FalconH1Config, FalconMambaConfig, FlexOlmoConfig, FuyuConfig, GemmaConfig, Gemma2Config, Gemma3Config, Gemma3TextConfig, Gemma3nConfig, Gemma3nTextConfig, GitConfig, GlmConfig, Glm4Config, Glm4MoeConfig, GotOcr2Config, GPT2Config, GPT2Config, GPTBigCodeConfig, GPTNeoConfig, GPTNeoXConfig, GPTNeoXJapaneseConfig, GptOssConfig, GPTJConfig, GraniteConfig, GraniteMoeConfig, GraniteMoeHybridConfig, GraniteMoeSharedConfig, HeliumConfig, HunYuanDenseV1Config, HunYuanMoEV1Config, Jais2Config, JambaConfig, JetMoeConfig, Lfm2Config, Lfm2MoeConfig, LlamaConfig, Llama4Config, Llama4TextConfig, LongcatFlashConfig, MambaConfig, Mamba2Config, MarianConfig, MBartConfig, MegatronBertConfig, MiniMaxConfig, MinistralConfig, Ministral3Config, MistralConfig, MixtralConfig, MllamaConfig, ModernBertDecoderConfig, MoshiConfig, MptConfig, MusicgenConfig, MusicgenMelodyConfig, MvpConfig, NanoChatConfig, NemotronConfig, OlmoConfig, Olmo2Config, Olmo3Config, OlmoeConfig, OpenAIGPTConfig, OPTConfig, PegasusConfig, PersimmonConfig, PhiConfig, Phi3Config, Phi4MultimodalConfig, PhimoeConfig, PLBartConfig, ProphetNetConfig, Qwen2Config, Qwen2MoeConfig, Qwen3Config, Qwen3MoeConfig, Qwen3NextConfig, RecurrentGemmaConfig, ReformerConfig, RemBertConfig, RobertaConfig, RobertaPreLayerNormConfig, RoCBertConfig, RoFormerConfig, RwkvConfig, SeedOssConfig, SmolLM3Config, StableLmConfig, Starcoder2Config, TrOCRConfig, VaultGemmaConfig, WhisperConfig, XGLMConfig, XLMConfig, XLMRobertaConfig, XLMRobertaXLConfig, XLNetConfig, xLSTMConfig, XmodConfig, ZambaConfig, Zamba2Config.
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/requirements.txt b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/wandb-metadata.json b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/wandb-metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..718f26e44892e91ac722c74863183804241e6ed4
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/wandb-metadata.json
@@ -0,0 +1,51 @@
+{
+  "os":  "Linux-6.12.46+-x86_64-with-glibc2.35",
+  "python":  "CPython 3.10.12",
+  "startedAt":  "2025-12-23T13:31:36.896493Z",
+  "args":  [
+    "--config",
+    "./SFT/config_instruct.yaml"
+  ],
+  "program":  "/workspace/Trainer-kit/SFT/run_instruct.py",
+  "codePath":  "SFT/run_instruct.py",
+  "codePathLocal":  "SFT/run_instruct.py",
+  "git":  {
+    "remote":  "https://github.com/Ashx098/Trainer-kit.git",
+    "commit":  "92382202abe01938bfeff39a3bf7b82e6e786325"
+  },
+  "email":  "shaiksirajuddin9949@gmail.com",
+  "root":  "runs/instruct_run_24b",
+  "host":  "a100-2gpu-shell-session-9c959f5f-mx8rm",
+  "executable":  "/workspace/CPT-env/bin/python",
+  "cpu_count":  12,
+  "cpu_count_logical":  24,
+  "gpu":  "NVIDIA A100-SXM4-80GB",
+  "gpu_count":  2,
+  "disk":  {
+    "/":  {
+      "total":  "791251738624",
+      "used":  "328812568576"
+    }
+  },
+  "memory":  {
+    "total":  "359047892992"
+  },
+  "gpu_nvidia":  [
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba"
+    },
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40"
+    }
+  ],
+  "cudaVersion":  "13.0",
+  "writerId":  "c4qlfqsdsdn87bdtewyqd1101bckxv0h"
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/wandb-summary.json b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/wandb-summary.json
new file mode 100644
index 0000000000000000000000000000000000000000..8afb95f49483c85658a334253ad61c5e4b5851ef
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/files/wandb-summary.json
@@ -0,0 +1 @@
+{"_wandb":{"runtime":2},"_runtime":2}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug-core.log b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..34f547e462037a3df485fc5dce0b42bff8b7b8c1
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:31:37.168073493Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp0huvbl6z/port-897046.txt","pid":897046,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:31:37.168685303Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":897046}
+{"time":"2025-12-23T13:31:37.168652863Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-897046-897142-488394314/socket","Net":"unix"}}
+{"time":"2025-12-23T13:31:37.35327197Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:31:37.359949568Z","level":"INFO","msg":"handleInformInit: received","streamId":"ppb51ho8","id":"1(@)"}
+{"time":"2025-12-23T13:31:37.52174679Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"ppb51ho8","id":"1(@)"}
+{"time":"2025-12-23T13:31:40.674826026Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T13:31:40.674915128Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T13:31:40.674903068Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T13:31:40.675004141Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T13:31:40.67509001Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-897046-897142-488394314/socket","Net":"unix"}}
+{"time":"2025-12-23T13:31:41.098133307Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T13:31:41.098174296Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T13:31:41.098191949Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug-internal.log b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..20ac3d28869d64cdb3c4b2013f2d490334861370
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug-internal.log
@@ -0,0 +1,11 @@
+{"time":"2025-12-23T13:31:37.360069148Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:31:37.52154358Z","level":"INFO","msg":"stream: created new stream","id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:37.52164403Z","level":"INFO","msg":"handler: started","stream_id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:37.52173764Z","level":"INFO","msg":"stream: started","id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:37.521762025Z","level":"INFO","msg":"writer: started","stream_id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:37.521793953Z","level":"INFO","msg":"sender: started","stream_id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:40.674955453Z","level":"INFO","msg":"stream: closing","id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:40.940521882Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2025-12-23T13:31:41.096688165Z","level":"INFO","msg":"handler: closed","stream_id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:41.096803406Z","level":"INFO","msg":"sender: closed","stream_id":"ppb51ho8"}
+{"time":"2025-12-23T13:31:41.096811307Z","level":"INFO","msg":"stream: closed","id":"ppb51ho8"}
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug.log b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..e67bdc21689a201a2a9a62d4526a4f6d07e8df5e
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/logs/debug.log
@@ -0,0 +1,23 @@
+2025-12-23 13:31:36,898 INFO    MainThread:897046 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:31:36,898 INFO    MainThread:897046 [wandb_setup.py:_flush():80] Configure stats pid to 897046
+2025-12-23 13:31:36,898 INFO    MainThread:897046 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:31:36,898 INFO    MainThread:897046 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:31:36,898 INFO    MainThread:897046 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:31:36,898 INFO    MainThread:897046 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/run-20251223_133136-ppb51ho8/logs/debug.log
+2025-12-23 13:31:36,899 INFO    MainThread:897046 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/run-20251223_133136-ppb51ho8/logs/debug-internal.log
+2025-12-23 13:31:36,899 INFO    MainThread:897046 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:31:36,899 INFO    MainThread:897046 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:31:36,899 INFO    MainThread:897046 [wandb_init.py:init():889] starting backend
+2025-12-23 13:31:37,353 INFO    MainThread:897046 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:31:37,358 INFO    MainThread:897046 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:31:37,360 INFO    MainThread:897046 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:31:37,364 INFO    MainThread:897046 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:31:37,722 INFO    MainThread:897046 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:31:37,858 INFO    MainThread:897046 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:31:37,858 INFO    MainThread:897046 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:31:37,858 INFO    MainThread:897046 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:31:37,858 INFO    MainThread:897046 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:31:37,863 INFO    MainThread:897046 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:31:40,674 INFO    wandb-AsyncioManager-main:897046 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 13:31:40,675 INFO    wandb-AsyncioManager-main:897046 [mailbox.py:close():137] Closing mailbox, abandoning 1 handles.
diff --git a/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/run-ppb51ho8.wandb b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/run-ppb51ho8.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..03c2317a19dfbb1235548b10799f3c08fceab83b
Binary files /dev/null and b/sft_devstral_24B/wandb/run-20251223_133136-ppb51ho8/run-ppb51ho8.wandb differ
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/config.yaml b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/config.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..fb9dc1b3ac9624b251998665b6289f686b3b9787
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/config.yaml
@@ -0,0 +1,631 @@
+_name_or_path:
+    value: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+_wandb:
+    value:
+        cli_version: 0.23.1
+        e:
+            s1ye6j5wjt18kccvxoax2tmmzadckaes:
+                args:
+                    - --config
+                    - ./SFT/config_instruct.yaml
+                codePath: SFT/run_instruct.py
+                codePathLocal: SFT/run_instruct.py
+                cpu_count: 12
+                cpu_count_logical: 24
+                cudaVersion: "13.0"
+                disk:
+                    /:
+                        total: "791251738624"
+                        used: "329776107520"
+                email: shaiksirajuddin9949@gmail.com
+                executable: /workspace/CPT-env/bin/python
+                git:
+                    commit: 92382202abe01938bfeff39a3bf7b82e6e786325
+                    remote: https://github.com/Ashx098/Trainer-kit.git
+                gpu: NVIDIA A100-SXM4-80GB
+                gpu_count: 2
+                gpu_nvidia:
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40
+                host: a100-2gpu-shell-session-9c959f5f-mx8rm
+                memory:
+                    total: "359047892992"
+                os: Linux-6.12.46+-x86_64-with-glibc2.35
+                program: /workspace/Trainer-kit/SFT/run_instruct.py
+                python: CPython 3.10.12
+                root: runs/instruct_run_24b
+                startedAt: "2025-12-23T13:46:18.997931Z"
+                writerId: s1ye6j5wjt18kccvxoax2tmmzadckaes
+        m:
+            - "1": train/global_step
+              "6":
+                - 3
+              "7": []
+            - "2": '*'
+              "5": 1
+              "6":
+                - 1
+              "7": []
+        python_version: 3.10.12
+        t:
+            "1":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "2":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "3":
+                - 7
+                - 15
+                - 16
+                - 19
+                - 66
+            "4": 3.10.12
+            "5": 0.23.1
+            "6": 5.0.0.dev0
+            "9":
+                "1": transformers_trainer
+            "12": 0.23.1
+            "13": linux-x86_64
+accelerator_config:
+    value:
+        dispatch_batches: null
+        even_batches: true
+        gradient_accumulation_kwargs: null
+        non_blocking: false
+        split_batches: false
+        use_seedable_sampler: true
+adam_beta1:
+    value: 0.9
+adam_beta2:
+    value: 0.999
+adam_epsilon:
+    value: 1e-08
+add_cross_attention:
+    value: false
+architectures:
+    value:
+        - Mistral3ForConditionalGeneration
+auto_find_batch_size:
+    value: false
+average_tokens_across_devices:
+    value: true
+batch_eval_metrics:
+    value: false
+bf16:
+    value: true
+bf16_full_eval:
+    value: false
+bos_token_id:
+    value: null
+chunk_size_feed_forward:
+    value: 0
+cross_attention_hidden_size:
+    value: null
+data:
+    value:
+        custom_template: |-
+            ##INSTRUCTION
+            {instruction}<|im_end|>
+            ##TASK
+            {input}<|im_end|>
+            ##OUTPUT
+            {output}<|im_end|>
+        eval_jsonl: null
+        eval_split_ratio: 0.1
+        format_type: custom
+        input_field: input
+        instruction_field: instruction
+        max_length: 2048
+        num_proc: 4
+        output_field: output
+        shuffle: true
+        system_prompt: |
+            You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.
+
+            ## Output Format
+
+            ##OUTPUT
+            Explain the data flow and why each component must change:
+            - Flow: [Input → Processing → Output with arrows]
+            - For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"
+            - Explain coupling between components
+
+            ##SELECT
+            modify::crates/path/to/file.rs::impl::ComponentName
+            add::crates/another/file.rs::function::AnotherComponent
+            <EOS>
+
+            ## Rules
+
+            1. Use full paths: `remove::crates/folder/file.rs::Type::Name`
+            2. Use `::` for nested items: `status::StructName::Type::Name`
+            3. Always explain "must change because" and "without this"
+            3. Types of components: function, struct, enum, impl, trait
+            4. If there is extra information (e.g., enum variants), include that too.
+            5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>
+
+            ## Example
+
+            ##TASK
+            Add webhook subscription support
+
+            ##OUTPUT
+            The webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don't trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.
+
+            ##SELECT
+            crates/common_enums/src/enums.rs::EventClass
+            crates/common_enums/src/transformers.rs::SubscriptionStatus
+            <EOS>
+        train_jsonl: ../sft_dataset.jsonl
+data_seed:
+    value: null
+dataloader_drop_last:
+    value: false
+dataloader_num_workers:
+    value: 0
+dataloader_persistent_workers:
+    value: false
+dataloader_pin_memory:
+    value: true
+dataloader_prefetch_factor:
+    value: null
+ddp_backend:
+    value: null
+ddp_broadcast_buffers:
+    value: null
+ddp_bucket_cap_mb:
+    value: null
+ddp_find_unused_parameters:
+    value: null
+ddp_timeout:
+    value: 1800
+debug:
+    value: []
+decoder_start_token_id:
+    value: null
+deepspeed:
+    value: null
+disable_tqdm:
+    value: false
+do_eval:
+    value: true
+do_predict:
+    value: false
+do_train:
+    value: false
+dtype:
+    value: bfloat16
+enable_jit_checkpoint:
+    value: false
+eos_token_id:
+    value: null
+eval_accumulation_steps:
+    value: null
+eval_delay:
+    value: 0
+eval_do_concat_batches:
+    value: true
+eval_on_start:
+    value: false
+eval_steps:
+    value: 100
+eval_strategy:
+    value: steps
+eval_use_gather_object:
+    value: false
+finetuning_task:
+    value: null
+fp16:
+    value: false
+fp16_full_eval:
+    value: false
+fsdp:
+    value: []
+fsdp_config:
+    value:
+        min_num_params: 0
+        xla: false
+        xla_fsdp_grad_ckpt: false
+        xla_fsdp_v2: false
+full_determinism:
+    value: false
+gradient_accumulation_steps:
+    value: 8
+gradient_checkpointing:
+    value: false
+gradient_checkpointing_kwargs:
+    value: null
+greater_is_better:
+    value: false
+group_by_length:
+    value: false
+hub_always_push:
+    value: false
+hub_model_id:
+    value: null
+hub_private_repo:
+    value: null
+hub_revision:
+    value: null
+hub_strategy:
+    value: every_save
+hub_token:
+    value: <HUB_TOKEN>
+id2label:
+    value:
+        "0": LABEL_0
+        "1": LABEL_1
+ignore_data_skip:
+    value: false
+image_token_index:
+    value: 10
+include_for_metrics:
+    value: []
+include_num_input_tokens_seen:
+    value: "no"
+is_decoder:
+    value: false
+is_encoder_decoder:
+    value: false
+label_names:
+    value: null
+label_smoothing_factor:
+    value: 0
+label2id:
+    value:
+        LABEL_0: 0
+        LABEL_1: 1
+learning_rate:
+    value: 5e-05
+length_column_name:
+    value: length
+liger_kernel_config:
+    value: null
+load_best_model_at_end:
+    value: true
+local_rank:
+    value: -1
+log_level:
+    value: passive
+log_level_replica:
+    value: warning
+log_on_each_node:
+    value: true
+logging_dir:
+    value: null
+logging_first_step:
+    value: false
+logging_nan_inf_filter:
+    value: true
+logging_steps:
+    value: 2
+logging_strategy:
+    value: steps
+lr_scheduler_kwargs:
+    value: null
+lr_scheduler_type:
+    value: cosine
+max_grad_norm:
+    value: 1
+max_steps:
+    value: -1
+metric_for_best_model:
+    value: eval_loss
+model:
+    value:
+        attn_implementation: null
+        base_local_dir: base_model
+        bnb_4bit_compute_dtype: bfloat16
+        bnb_4bit_quant_type: nf4
+        bnb_4bit_use_double_quant: false
+        device_map: auto
+        repo_id: ./CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+        revision: null
+        tokenizer_use_fast: true
+        torch_dtype: bfloat16
+        trust_remote_code: true
+        use_4bit: false
+model/num_parameters:
+    value: 24102587392
+model_type:
+    value: mistral3
+multimodal_projector_bias:
+    value: false
+neftune_noise_alpha:
+    value: null
+num_train_epochs:
+    value: 6
+optim:
+    value: adamw_torch
+optim_args:
+    value: null
+optim_target_modules:
+    value: null
+output_attentions:
+    value: false
+output_dir:
+    value: runs/instruct_run_24b/checkpoints
+output_hidden_states:
+    value: false
+pad_token_id:
+    value: null
+parallelism_config:
+    value: null
+peft:
+    value:
+        bias: none
+        enabled: true
+        lora_alpha: 128
+        lora_dropout: 0.05
+        r: 64
+        target_modules: auto
+peft_config:
+    value:
+        default:
+            alora_invocation_tokens: null
+            arrow_config: null
+            auto_mapping: null
+            base_model_name_or_path: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+            bias: none
+            corda_config: null
+            ensure_weight_tying: false
+            eva_config: null
+            exclude_modules: null
+            fan_in_fan_out: false
+            inference_mode: false
+            init_lora_weights: true
+            layer_replication: null
+            layers_pattern: null
+            layers_to_transform: null
+            lora_alpha: 128
+            lora_bias: false
+            lora_dropout: 0.05
+            megatron_config: null
+            megatron_core: megatron.core
+            modules_to_save: null
+            peft_type: LORA
+            peft_version: 0.18.0
+            qalora_group_size: 16
+            r: 64
+            revision: null
+            runtime_config:
+                ephemeral_gpu_offload: false
+            target_modules:
+                - v_proj
+                - k_proj
+                - o_proj
+                - q_proj
+            target_parameters: null
+            task_type: CAUSAL_LM
+            trainable_token_indices: null
+            use_dora: false
+            use_qalora: false
+            use_rslora: false
+per_device_eval_batch_size:
+    value: 1
+per_device_train_batch_size:
+    value: 1
+prediction_loss_only:
+    value: false
+prefix:
+    value: null
+problem_type:
+    value: null
+project:
+    value: huggingface
+projector_hidden_act:
+    value: gelu
+push_to_hub:
+    value: false
+remove_unused_columns:
+    value: false
+report_to:
+    value:
+        - wandb
+restore_callback_states_from_checkpoint:
+    value: false
+resume_from_checkpoint:
+    value: null
+return_dict:
+    value: true
+run_dir:
+    value: runs/instruct_run_24b
+run_name:
+    value: null
+save_on_each_node:
+    value: false
+save_only_model:
+    value: false
+save_steps:
+    value: 500
+save_strategy:
+    value: steps
+save_total_limit:
+    value: 20
+seed:
+    value: 42
+sep_token_id:
+    value: null
+skip_memory_metrics:
+    value: true
+spatial_merge_size:
+    value: 2
+task_specific_params:
+    value: null
+text_config:
+    value:
+        _name_or_path: ""
+        add_cross_attention: false
+        architectures: null
+        attention_dropout: 0
+        bos_token_id: 1
+        chunk_size_feed_forward: 0
+        cross_attention_hidden_size: null
+        decoder_start_token_id: null
+        dtype: bfloat16
+        eos_token_id: 2
+        finetuning_task: null
+        head_dim: 128
+        hidden_act: silu
+        hidden_size: 5120
+        id2label:
+            "0": LABEL_0
+            "1": LABEL_1
+        initializer_range: 0.02
+        intermediate_size: 32768
+        is_decoder: false
+        is_encoder_decoder: false
+        label2id:
+            LABEL_0: 0
+            LABEL_1: 1
+        max_position_embeddings: 393216
+        model_type: ministral3
+        num_attention_heads: 32
+        num_hidden_layers: 40
+        num_key_value_heads: 8
+        output_attentions: false
+        output_hidden_states: false
+        pad_token_id: 11
+        prefix: null
+        problem_type: null
+        return_dict: true
+        rms_norm_eps: 1e-05
+        rope_parameters:
+            beta_fast: 32
+            beta_slow: 1
+            factor: 48
+            llama_4_scaling_beta: 0.1
+            mscale: 1
+            mscale_all_dim: 1
+            original_max_position_embeddings: 8192
+            rope_theta: 1e+08
+            rope_type: yarn
+            type: yarn
+        sep_token_id: null
+        sliding_window: null
+        task_specific_params: null
+        tie_word_embeddings: false
+        tokenizer_class: null
+        use_cache: true
+        vocab_size: 131072
+tf32:
+    value: null
+tie_word_embeddings:
+    value: false
+tokenizer_class:
+    value: null
+torch_compile:
+    value: false
+torch_compile_backend:
+    value: null
+torch_compile_mode:
+    value: null
+torch_empty_cache_steps:
+    value: null
+trackio_space_id:
+    value: trackio
+train:
+    value:
+        early_stopping:
+            enabled: true
+            metric: eval_loss
+            min_delta: 0.001
+            mode: min
+            patience: 3
+        eval_steps: 100
+        evaluation_strategy: steps
+        gradient_accumulation_steps: 8
+        gradient_checkpointing: true
+        learning_rate: "5e-5"
+        load_best_model_at_end: true
+        logging_steps: 2
+        lr_scheduler_type: cosine
+        max_grad_norm: 1
+        num_train_epochs: 6
+        optim: adamw_torch
+        per_device_eval_batch_size: 1
+        per_device_train_batch_size: 1
+        resume_from_checkpoint: auto
+        save_steps: 500
+        save_strategy: steps
+        save_total_limit: 20
+        warmup_ratio: 0.01
+        weight_decay: 0
+transformers_version:
+    value: 5.0.0.dev0
+use_cache:
+    value: false
+use_cpu:
+    value: false
+use_liger_kernel:
+    value: false
+vision_config:
+    value:
+        _name_or_path: ""
+        add_cross_attention: false
+        architectures: null
+        attention_dropout: 0
+        bos_token_id: null
+        chunk_size_feed_forward: 0
+        cross_attention_hidden_size: null
+        decoder_start_token_id: null
+        dtype: bfloat16
+        eos_token_id: null
+        finetuning_task: null
+        head_dim: 64
+        hidden_act: silu
+        hidden_size: 1024
+        id2label:
+            "0": LABEL_0
+            "1": LABEL_1
+        image_size: 1540
+        initializer_range: 0.02
+        intermediate_size: 4096
+        is_decoder: false
+        is_encoder_decoder: false
+        label2id:
+            LABEL_0: 0
+            LABEL_1: 1
+        model_type: pixtral
+        num_attention_heads: 16
+        num_channels: 3
+        num_hidden_layers: 24
+        output_attentions: false
+        output_hidden_states: false
+        pad_token_id: null
+        patch_size: 14
+        prefix: null
+        problem_type: null
+        return_dict: true
+        rope_parameters:
+            rope_theta: 10000
+            rope_type: default
+        sep_token_id: null
+        task_specific_params: null
+        tie_word_embeddings: true
+        tokenizer_class: null
+vision_feature_layer:
+    value: -1
+warmup_ratio:
+    value: 0.01
+warmup_steps:
+    value: 0.01
+weight_decay:
+    value: 0
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/output.log b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/output.log
new file mode 100644
index 0000000000000000000000000000000000000000..fea2d5213b31fc8b023247e713610f7c87243075
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/output.log
@@ -0,0 +1,244 @@
+Wandb initialized: project='sft-training', name='auto-generated'
+[info] Detected Mistral3 model architecture, loading with specific class
+Loading weights: 100%|█| 585/585 [00:12<00:00, 46.14it/s, Materializing param=model.vision_tower.transformer.
+[info] Ensuring all parameters are materialized...
+warmup_ratio is deprecated and will be removed in v5.2. Use `warmup_steps` instead.
+Early stopping enabled: patience=3, min_delta=0.001
+Starting instruction fine-tuning...
+  0%|                                                                              | 0/13908 [00:00<?, ?it/s]`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
+                                                                                                             
+{'loss': '1.399', 'grad_norm': '5.482', 'learning_rate': '3.571e-07', 'epoch': '0.000863'}
+{'loss': '1.369', 'grad_norm': '5.094', 'learning_rate': '1.071e-06', 'epoch': '0.001726'}
+{'loss': '1.443', 'grad_norm': '4.64', 'learning_rate': '1.786e-06', 'epoch': '0.002589'}
+{'loss': '1.318', 'grad_norm': '4.059', 'learning_rate': '2.5e-06', 'epoch': '0.003452'}
+{'loss': '1.389', 'grad_norm': '4.043', 'learning_rate': '3.214e-06', 'epoch': '0.004315'}
+{'loss': '1.329', 'grad_norm': '3.246', 'learning_rate': '3.929e-06', 'epoch': '0.005178'}
+{'loss': '1.318', 'grad_norm': '2.781', 'learning_rate': '4.643e-06', 'epoch': '0.006041'}
+{'loss': '1.252', 'grad_norm': '3.072', 'learning_rate': '5.357e-06', 'epoch': '0.006904'}
+{'loss': '1.29', 'grad_norm': '3.047', 'learning_rate': '6.071e-06', 'epoch': '0.007767'}
+{'loss': '1.237', 'grad_norm': '2.781', 'learning_rate': '6.786e-06', 'epoch': '0.00863'}
+{'loss': '1.168', 'grad_norm': '2.391', 'learning_rate': '7.5e-06', 'epoch': '0.009493'}
+{'loss': '1.13', 'grad_norm': '3.611', 'learning_rate': '8.214e-06', 'epoch': '0.01036'}
+{'loss': '1.205', 'grad_norm': '2.862', 'learning_rate': '8.929e-06', 'epoch': '0.01122'}
+{'loss': '1.069', 'grad_norm': '2.454', 'learning_rate': '9.643e-06', 'epoch': '0.01208'}
+{'loss': '1.161', 'grad_norm': '3.124', 'learning_rate': '1.036e-05', 'epoch': '0.01294'}
+{'loss': '1.11', 'grad_norm': '2.62', 'learning_rate': '1.107e-05', 'epoch': '0.01381'}
+{'loss': '1.107', 'grad_norm': '2.838', 'learning_rate': '1.179e-05', 'epoch': '0.01467'}
+{'loss': '1.004', 'grad_norm': '2.099', 'learning_rate': '1.25e-05', 'epoch': '0.01553'}
+{'loss': '1.071', 'grad_norm': '2.639', 'learning_rate': '1.321e-05', 'epoch': '0.0164'}
+{'loss': '1.065', 'grad_norm': '2.677', 'learning_rate': '1.393e-05', 'epoch': '0.01726'}
+{'loss': '1.016', 'grad_norm': '3.025', 'learning_rate': '1.464e-05', 'epoch': '0.01812'}
+{'loss': '1.091', 'grad_norm': '2.372', 'learning_rate': '1.536e-05', 'epoch': '0.01899'}
+{'loss': '1.03', 'grad_norm': '5.351', 'learning_rate': '1.607e-05', 'epoch': '0.01985'}
+{'loss': '1.008', 'grad_norm': '2.345', 'learning_rate': '1.679e-05', 'epoch': '0.02071'}
+{'loss': '0.9324', 'grad_norm': '3.227', 'learning_rate': '1.75e-05', 'epoch': '0.02157'}
+{'loss': '0.9128', 'grad_norm': '2.304', 'learning_rate': '1.821e-05', 'epoch': '0.02244'}
+{'loss': '0.9794', 'grad_norm': '2.35', 'learning_rate': '1.893e-05', 'epoch': '0.0233'}
+{'loss': '0.9212', 'grad_norm': '2.25', 'learning_rate': '1.964e-05', 'epoch': '0.02416'}
+{'loss': '0.9297', 'grad_norm': '2.167', 'learning_rate': '2.036e-05', 'epoch': '0.02503'}
+{'loss': '0.9785', 'grad_norm': '2.568', 'learning_rate': '2.107e-05', 'epoch': '0.02589'}
+{'loss': '0.8338', 'grad_norm': '2.042', 'learning_rate': '2.179e-05', 'epoch': '0.02675'}
+{'loss': '0.9178', 'grad_norm': '2.076', 'learning_rate': '2.25e-05', 'epoch': '0.02762'}
+{'loss': '1.038', 'grad_norm': '2.458', 'learning_rate': '2.321e-05', 'epoch': '0.02848'}
+{'loss': '1', 'grad_norm': '2.569', 'learning_rate': '2.393e-05', 'epoch': '0.02934'}
+{'loss': '0.9486', 'grad_norm': '2.312', 'learning_rate': '2.464e-05', 'epoch': '0.0302'}
+{'loss': '0.9413', 'grad_norm': '2.129', 'learning_rate': '2.536e-05', 'epoch': '0.03107'}
+{'loss': '0.9601', 'grad_norm': '2.171', 'learning_rate': '2.607e-05', 'epoch': '0.03193'}
+{'loss': '0.9564', 'grad_norm': '1.964', 'learning_rate': '2.679e-05', 'epoch': '0.03279'}
+{'loss': '0.9692', 'grad_norm': '2.791', 'learning_rate': '2.75e-05', 'epoch': '0.03366'}
+{'loss': '0.9474', 'grad_norm': '2.138', 'learning_rate': '2.821e-05', 'epoch': '0.03452'}
+{'loss': '0.96', 'grad_norm': '2.583', 'learning_rate': '2.893e-05', 'epoch': '0.03538'}
+{'loss': '0.881', 'grad_norm': '2.136', 'learning_rate': '2.964e-05', 'epoch': '0.03625'}
+{'loss': '0.8347', 'grad_norm': '1.89', 'learning_rate': '3.036e-05', 'epoch': '0.03711'}
+{'loss': '0.9489', 'grad_norm': '2.685', 'learning_rate': '3.107e-05', 'epoch': '0.03797'}
+{'loss': '0.8981', 'grad_norm': '2.344', 'learning_rate': '3.179e-05', 'epoch': '0.03883'}
+{'loss': '0.8588', 'grad_norm': '2.402', 'learning_rate': '3.25e-05', 'epoch': '0.0397'}
+{'loss': '0.8564', 'grad_norm': '1.965', 'learning_rate': '3.321e-05', 'epoch': '0.04056'}
+{'loss': '1.014', 'grad_norm': '2.42', 'learning_rate': '3.393e-05', 'epoch': '0.04142'}
+{'loss': '0.8932', 'grad_norm': '3.049', 'learning_rate': '3.464e-05', 'epoch': '0.04229'}
+{'loss': '0.9056', 'grad_norm': '1.82', 'learning_rate': '3.536e-05', 'epoch': '0.04315'}
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 921, in <module>                                   
+{'eval_loss': '0.936', 'eval_runtime': '834.9', 'eval_samples_per_second': '2.467', 'eval_steps_per_second': '2.467', 'epoch': '0.04315'}
+{'loss': '0.9161', 'grad_norm': '2.02', 'learning_rate': '3.607e-05', 'epoch': '0.04401'}
+{'loss': '0.8526', 'grad_norm': '2.063', 'learning_rate': '3.679e-05', 'epoch': '0.04488'}
+{'loss': '0.9063', 'grad_norm': '2.217', 'learning_rate': '3.75e-05', 'epoch': '0.04574'}
+{'loss': '0.932', 'grad_norm': '1.944', 'learning_rate': '3.821e-05', 'epoch': '0.0466'}
+{'loss': '0.962', 'grad_norm': '1.944', 'learning_rate': '3.893e-05', 'epoch': '0.04746'}
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 896, in main
+    trainer.train(resume_from_checkpoint=resume_from)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 2168, in train
+    return inner_training_loop(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 2535, in _inner_training_loop
+    tr_loss_step = self.training_step(model, inputs, num_items_in_batch)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 3807, in training_step
+    loss = self.compute_loss(model, inputs, num_items_in_batch=num_items_in_batch)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 3878, in compute_loss
+    outputs = model(**inputs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/utils/operations.py", line 819, in forward
+    return model_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/utils/operations.py", line 807, in __call__
+    return convert_to_fp32(self.model_forward(*args, **kwargs))
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/amp/autocast_mode.py", line 44, in decorate_autocast
+    return func(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/peft_model.py", line 1923, in forward
+    return self.base_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/tuners/tuners_utils.py", line 308, in forward
+    return self.model.forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/hooks.py", line 175, in new_forward
+    output = module._old_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/utils/generic.py", line 810, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/mistral3/modeling_mistral3.py", line 449, in forward
+    outputs = self.model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/utils/generic.py", line 810, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/mistral3/modeling_mistral3.py", line 331, in forward
+    outputs = self.language_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/utils/generic.py", line 965, in wrapper
+    outputs = func(self, *args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/ministral3/modeling_ministral3.py", line 409, in forward
+    hidden_states = decoder_layer(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/modeling_layers.py", line 93, in __call__
+    return self._gradient_checkpointing_func(partial(super().__call__, **kwargs), *args)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/_compile.py", line 53, in inner
+    return disable_fn(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 1044, in _fn
+    return fn(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 496, in checkpoint
+    return CheckpointFunction.apply(function, preserve, *args)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/autograd/function.py", line 581, in apply
+    return super().apply(*args, **kwargs)  # type: ignore[misc]
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 262, in forward
+    outputs = run_function(*args)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/hooks.py", line 175, in new_forward
+    output = module._old_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/ministral3/modeling_ministral3.py", line 242, in forward
+    hidden_states, _ = self.self_attn(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/hooks.py", line 175, in new_forward
+    output = module._old_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/ministral3/modeling_ministral3.py", line 146, in forward
+    value_states = self.v_proj(hidden_states).view(hidden_shape).transpose(1, 2)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/tuners/lora/layer.py", line 805, in forward
+    x = self._cast_input_dtype(x, lora_A.weight.dtype)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/tuners/tuners_utils.py", line 1547, in _cast_input_dtype
+    cast_input_dtype_enabled = getattr(self, "cast_input_dtype_enabled", True)
+KeyboardInterrupt
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 921, in <module>
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 896, in main
+    trainer.train(resume_from_checkpoint=resume_from)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 2168, in train
+    return inner_training_loop(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 2535, in _inner_training_loop
+    tr_loss_step = self.training_step(model, inputs, num_items_in_batch)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 3807, in training_step
+    loss = self.compute_loss(model, inputs, num_items_in_batch=num_items_in_batch)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/trainer.py", line 3878, in compute_loss
+    outputs = model(**inputs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/utils/operations.py", line 819, in forward
+    return model_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/utils/operations.py", line 807, in __call__
+    return convert_to_fp32(self.model_forward(*args, **kwargs))
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/amp/autocast_mode.py", line 44, in decorate_autocast
+    return func(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/peft_model.py", line 1923, in forward
+    return self.base_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/tuners/tuners_utils.py", line 308, in forward
+    return self.model.forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/hooks.py", line 175, in new_forward
+    output = module._old_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/utils/generic.py", line 810, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/mistral3/modeling_mistral3.py", line 449, in forward
+    outputs = self.model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/utils/generic.py", line 810, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/mistral3/modeling_mistral3.py", line 331, in forward
+    outputs = self.language_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/utils/generic.py", line 965, in wrapper
+    outputs = func(self, *args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/ministral3/modeling_ministral3.py", line 409, in forward
+    hidden_states = decoder_layer(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/modeling_layers.py", line 93, in __call__
+    return self._gradient_checkpointing_func(partial(super().__call__, **kwargs), *args)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/_compile.py", line 53, in inner
+    return disable_fn(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 1044, in _fn
+    return fn(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 496, in checkpoint
+    return CheckpointFunction.apply(function, preserve, *args)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/autograd/function.py", line 581, in apply
+    return super().apply(*args, **kwargs)  # type: ignore[misc]
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 262, in forward
+    outputs = run_function(*args)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/hooks.py", line 175, in new_forward
+    output = module._old_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/ministral3/modeling_ministral3.py", line 242, in forward
+    hidden_states, _ = self.self_attn(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/accelerate/hooks.py", line 175, in new_forward
+    output = module._old_forward(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/ministral3/modeling_ministral3.py", line 146, in forward
+    value_states = self.v_proj(hidden_states).view(hidden_shape).transpose(1, 2)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/tuners/lora/layer.py", line 805, in forward
+    x = self._cast_input_dtype(x, lora_A.weight.dtype)
+  File "/workspace/CPT-env/lib/python3.10/site-packages/peft/tuners/tuners_utils.py", line 1547, in _cast_input_dtype
+    cast_input_dtype_enabled = getattr(self, "cast_input_dtype_enabled", True)
+KeyboardInterrupt
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/requirements.txt b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/wandb-metadata.json b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/wandb-metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..c468300ce9d49690f93719450212d1a9b2673e84
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/wandb-metadata.json
@@ -0,0 +1,51 @@
+{
+  "os":  "Linux-6.12.46+-x86_64-with-glibc2.35",
+  "python":  "CPython 3.10.12",
+  "startedAt":  "2025-12-23T13:46:18.997931Z",
+  "args":  [
+    "--config",
+    "./SFT/config_instruct.yaml"
+  ],
+  "program":  "/workspace/Trainer-kit/SFT/run_instruct.py",
+  "codePath":  "SFT/run_instruct.py",
+  "codePathLocal":  "SFT/run_instruct.py",
+  "git":  {
+    "remote":  "https://github.com/Ashx098/Trainer-kit.git",
+    "commit":  "92382202abe01938bfeff39a3bf7b82e6e786325"
+  },
+  "email":  "shaiksirajuddin9949@gmail.com",
+  "root":  "runs/instruct_run_24b",
+  "host":  "a100-2gpu-shell-session-9c959f5f-mx8rm",
+  "executable":  "/workspace/CPT-env/bin/python",
+  "cpu_count":  12,
+  "cpu_count_logical":  24,
+  "gpu":  "NVIDIA A100-SXM4-80GB",
+  "gpu_count":  2,
+  "disk":  {
+    "/":  {
+      "total":  "791251738624",
+      "used":  "329776107520"
+    }
+  },
+  "memory":  {
+    "total":  "359047892992"
+  },
+  "gpu_nvidia":  [
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba"
+    },
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40"
+    }
+  ],
+  "cudaVersion":  "13.0",
+  "writerId":  "s1ye6j5wjt18kccvxoax2tmmzadckaes"
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/wandb-summary.json b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/wandb-summary.json
new file mode 100644
index 0000000000000000000000000000000000000000..e60f93baf6969187a487506d304c1089989e575b
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/files/wandb-summary.json
@@ -0,0 +1 @@
+{"eval/loss":0.9360476136207581,"train/loss":0.9619514346122742,"_wandb":{"runtime":1914},"train/global_step":110,"_runtime":1914,"_step":55,"train/learning_rate":3.892857142857143e-05,"eval/samples_per_second":2.467,"train/epoch":0.04746494066882417,"train/grad_norm":1.9441817998886108,"_timestamp":1.766499493333307e+09,"eval/steps_per_second":2.467,"eval/runtime":834.9}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug-core.log b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..c1ff6758d1836b78f0283fe4865b8cf19bb72823
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T13:46:19.252514162Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpmsurnhv8/port-905748.txt","pid":905748,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T13:46:19.253124037Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":905748}
+{"time":"2025-12-23T13:46:19.253106968Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-905748-905851-3403929552/socket","Net":"unix"}}
+{"time":"2025-12-23T13:46:19.438032709Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T13:46:19.446294923Z","level":"INFO","msg":"handleInformInit: received","streamId":"9jed4peb","id":"1(@)"}
+{"time":"2025-12-23T13:46:19.599480078Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"9jed4peb","id":"1(@)"}
+{"time":"2025-12-23T14:18:14.700284741Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T14:18:14.700390753Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T14:18:14.700460679Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T14:18:14.700410352Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T14:18:14.70055921Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-905748-905851-3403929552/socket","Net":"unix"}}
+{"time":"2025-12-23T14:18:15.143103436Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T14:18:15.143144225Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T14:18:15.143157553Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug-internal.log b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..41cb61d2f2b9e9512e6b1fe9aa3be18d1f780cea
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug-internal.log
@@ -0,0 +1,11 @@
+{"time":"2025-12-23T13:46:19.446458079Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T13:46:19.599297934Z","level":"INFO","msg":"stream: created new stream","id":"9jed4peb"}
+{"time":"2025-12-23T13:46:19.59939412Z","level":"INFO","msg":"handler: started","stream_id":"9jed4peb"}
+{"time":"2025-12-23T13:46:19.599469815Z","level":"INFO","msg":"stream: started","id":"9jed4peb"}
+{"time":"2025-12-23T13:46:19.599497631Z","level":"INFO","msg":"writer: started","stream_id":"9jed4peb"}
+{"time":"2025-12-23T13:46:19.599547576Z","level":"INFO","msg":"sender: started","stream_id":"9jed4peb"}
+{"time":"2025-12-23T14:18:14.700385192Z","level":"INFO","msg":"stream: closing","id":"9jed4peb"}
+{"time":"2025-12-23T14:18:14.93233186Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2025-12-23T14:18:15.142062605Z","level":"INFO","msg":"handler: closed","stream_id":"9jed4peb"}
+{"time":"2025-12-23T14:18:15.142193511Z","level":"INFO","msg":"sender: closed","stream_id":"9jed4peb"}
+{"time":"2025-12-23T14:18:15.142208896Z","level":"INFO","msg":"stream: closed","id":"9jed4peb"}
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug.log b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..0210b15b2a2a0563494b37d2b1413e3fcdc0162c
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/logs/debug.log
@@ -0,0 +1,26 @@
+2025-12-23 13:46:18,999 INFO    MainThread:905748 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 13:46:18,999 INFO    MainThread:905748 [wandb_setup.py:_flush():80] Configure stats pid to 905748
+2025-12-23 13:46:18,999 INFO    MainThread:905748 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 13:46:18,999 INFO    MainThread:905748 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 13:46:18,999 INFO    MainThread:905748 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 13:46:18,999 INFO    MainThread:905748 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/run-20251223_134618-9jed4peb/logs/debug.log
+2025-12-23 13:46:19,000 INFO    MainThread:905748 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/run-20251223_134618-9jed4peb/logs/debug-internal.log
+2025-12-23 13:46:19,000 INFO    MainThread:905748 [wandb_init.py:init():841] calling init triggers
+2025-12-23 13:46:19,000 INFO    MainThread:905748 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 64, 'lora_alpha': 128, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '5e-5', 'weight_decay': 0.0, 'warmup_ratio': 0.01, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 1.0, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 13:46:19,000 INFO    MainThread:905748 [wandb_init.py:init():889] starting backend
+2025-12-23 13:46:19,438 INFO    MainThread:905748 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 13:46:19,443 INFO    MainThread:905748 [wandb_init.py:init():900] backend started and connected
+2025-12-23 13:46:19,446 INFO    MainThread:905748 [wandb_init.py:init():970] updated telemetry
+2025-12-23 13:46:19,452 INFO    MainThread:905748 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 13:46:19,909 INFO    MainThread:905748 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 13:46:20,043 INFO    MainThread:905748 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 13:46:20,043 INFO    MainThread:905748 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 13:46:20,044 INFO    MainThread:905748 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 13:46:20,044 INFO    MainThread:905748 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 13:46:20,048 INFO    MainThread:905748 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 13:46:42,700 INFO    MainThread:905748 [wandb_run.py:_config_callback():1396] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.0', 'base_model_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'inference_mode': False, 'r': 64, 'target_modules': ['v_proj', 'k_proj', 'o_proj', 'q_proj'], 'exclude_modules': None, 'lora_alpha': 128, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': None, 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False}}, 'image_token_index': 10, 'projector_hidden_act': 'gelu', 'vision_feature_layer': -1, 'vision_config': {'hidden_size': 1024, 'intermediate_size': 4096, 'num_hidden_layers': 24, 'num_attention_heads': 16, 'num_channels': 3, 'patch_size': 14, 'image_size': 1540, 'attention_dropout': 0.0, 'hidden_act': 'silu', 'head_dim': 64, 'initializer_range': 0.02, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'pixtral', 'output_attentions': False}, 'text_config': {'vocab_size': 131072, 'max_position_embeddings': 393216, 'hidden_size': 5120, 'intermediate_size': 32768, 'num_hidden_layers': 40, 'num_attention_heads': 32, 'sliding_window': None, 'head_dim': 128, 'num_key_value_heads': 8, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'use_cache': True, 'attention_dropout': 0.0, 'rope_parameters': {'beta_fast': 32.0, 'beta_slow': 1.0, 'factor': 48.0, 'llama_4_scaling_beta': 0.1, 'mscale': 1.0, 'mscale_all_dim': 1.0, 'original_max_position_embeddings': 8192, 'rope_theta': 100000000.0, 'rope_type': 'yarn', 'type': 'yarn'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 11, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'ministral3', 'output_attentions': False}, 'multimodal_projector_bias': False, 'spatial_merge_size': 2, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': ['Mistral3ForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'transformers_version': '5.0.0.dev0', 'model_type': 'mistral3', 'use_cache': False, 'output_attentions': False, 'output_dir': 'runs/instruct_run_24b/checkpoints', 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 5e-05, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 6.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.01, 'warmup_steps': 0.01, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': None, 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 2, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'enable_jit_checkpoint': False, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'use_cpu': False, 'seed': 42, 'data_seed': None, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': -1, 'ddp_backend': None, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': 100, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'run_name': None, 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': True, 'metric_for_best_model': 'eval_loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'auto_find_batch_size': False, 'full_determinism': False, 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True}
+2025-12-23 13:46:42,710 INFO    MainThread:905748 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 24102587392 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x79a67434b550>>
+2025-12-23 13:46:42,710 INFO    MainThread:905748 [wandb_run.py:_config_callback():1396] config_cb model/num_parameters 24102587392 None
+2025-12-23 14:18:14,700 INFO    wandb-AsyncioManager-main:905748 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 14:18:14,700 INFO    wandb-AsyncioManager-main:905748 [mailbox.py:close():137] Closing mailbox, abandoning 1 handles.
diff --git a/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/run-9jed4peb.wandb b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/run-9jed4peb.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..1823ae51626d39a4b15a86ae74535269ceadda35
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_134618-9jed4peb/run-9jed4peb.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d1737f18862e6f71558b5c08031429b53af2ac88099c18b1e5f72ec3d4e6777
+size 1180471
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/config.yaml b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/config.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..f5e6281ca4aaa211c72aadc2497ee8f68a319462
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/config.yaml
@@ -0,0 +1,176 @@
+_wandb:
+    value:
+        cli_version: 0.23.1
+        e:
+            7eidhw1dhxga1ovk8tpqqdaeudz1qfd9:
+                args:
+                    - --config
+                    - ./SFT/config_instruct.yaml
+                codePath: SFT/run_instruct.py
+                codePathLocal: SFT/run_instruct.py
+                cpu_count: 12
+                cpu_count_logical: 24
+                cudaVersion: "13.0"
+                disk:
+                    /:
+                        total: "791251738624"
+                        used: "329782554624"
+                email: shaiksirajuddin9949@gmail.com
+                executable: /workspace/CPT-env/bin/python
+                git:
+                    commit: 92382202abe01938bfeff39a3bf7b82e6e786325
+                    remote: https://github.com/Ashx098/Trainer-kit.git
+                gpu: NVIDIA A100-SXM4-80GB
+                gpu_count: 2
+                gpu_nvidia:
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40
+                host: a100-2gpu-shell-session-9c959f5f-mx8rm
+                memory:
+                    total: "359047892992"
+                os: Linux-6.12.46+-x86_64-with-glibc2.35
+                program: /workspace/Trainer-kit/SFT/run_instruct.py
+                python: CPython 3.10.12
+                root: runs/instruct_run_24b
+                startedAt: "2025-12-23T14:22:08.868345Z"
+                writerId: 7eidhw1dhxga1ovk8tpqqdaeudz1qfd9
+        m: []
+        python_version: 3.10.12
+        t:
+            "1":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "2":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "3":
+                - 15
+                - 16
+            "4": 3.10.12
+            "5": 0.23.1
+            "6": 5.0.0.dev0
+            "12": 0.23.1
+            "13": linux-x86_64
+data:
+    value:
+        custom_template: |-
+            ##INSTRUCTION
+            {instruction}<|im_end|>
+            ##TASK
+            {input}<|im_end|>
+            ##OUTPUT
+            {output}<|im_end|>
+        eval_jsonl: null
+        eval_split_ratio: 0.1
+        format_type: custom
+        input_field: input
+        instruction_field: instruction
+        max_length: 2048
+        num_proc: 4
+        output_field: output
+        shuffle: true
+        system_prompt: |
+            You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.
+
+            ## Output Format
+
+            ##OUTPUT
+            Explain the data flow and why each component must change:
+            - Flow: [Input → Processing → Output with arrows]
+            - For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"
+            - Explain coupling between components
+
+            ##SELECT
+            modify::crates/path/to/file.rs::impl::ComponentName
+            add::crates/another/file.rs::function::AnotherComponent
+            <EOS>
+
+            ## Rules
+
+            1. Use full paths: `remove::crates/folder/file.rs::Type::Name`
+            2. Use `::` for nested items: `status::StructName::Type::Name`
+            3. Always explain "must change because" and "without this"
+            3. Types of components: function, struct, enum, impl, trait
+            4. If there is extra information (e.g., enum variants), include that too.
+            5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>
+
+            ## Example
+
+            ##TASK
+            Add webhook subscription support
+
+            ##OUTPUT
+            The webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don't trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.
+
+            ##SELECT
+            crates/common_enums/src/enums.rs::EventClass
+            crates/common_enums/src/transformers.rs::SubscriptionStatus
+            <EOS>
+        train_jsonl: ../sft_dataset.jsonl
+model:
+    value:
+        attn_implementation: null
+        base_local_dir: base_model
+        bnb_4bit_compute_dtype: bfloat16
+        bnb_4bit_quant_type: nf4
+        bnb_4bit_use_double_quant: false
+        device_map: auto
+        repo_id: ./CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+        revision: null
+        tokenizer_use_fast: true
+        torch_dtype: bfloat16
+        trust_remote_code: true
+        use_4bit: false
+peft:
+    value:
+        bias: none
+        enabled: true
+        lora_alpha: 16
+        lora_dropout: 0.05
+        r: 8
+        target_modules: auto
+run_dir:
+    value: runs/instruct_run_24b
+train:
+    value:
+        early_stopping:
+            enabled: true
+            metric: eval_loss
+            min_delta: 0.001
+            mode: min
+            patience: 3
+        eval_steps: 100
+        evaluation_strategy: steps
+        gradient_accumulation_steps: 8
+        gradient_checkpointing: true
+        learning_rate: "1e-4"
+        load_best_model_at_end: true
+        logging_steps: 2
+        lr_scheduler_type: cosine
+        max_grad_norm: 0.8
+        num_train_epochs: 6
+        optim: adamw_torch
+        per_device_eval_batch_size: 1
+        per_device_train_batch_size: 1
+        resume_from_checkpoint: auto
+        save_steps: 500
+        save_strategy: steps
+        save_total_limit: 20
+        warmup_ratio: 0.1
+        weight_decay: 0
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/output.log b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/output.log
new file mode 100644
index 0000000000000000000000000000000000000000..4c6a825b7e4b5234499f3caae444f5071703cc80
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/output.log
@@ -0,0 +1,51 @@
+Wandb initialized: project='sft-training', name='auto-generated'
+[info] Detected Mistral3 model architecture, loading with specific class
+Loading weights:  29%|▎| 172/585 [00:06<00:14, 28.17it/s, Materializing param=model.language_model.layers.18.
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 921, in <module>
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 787, in main
+    model, tokenizer = load_base_model_and_tokenizer(cfg, base_dir)
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 548, in load_base_model_and_tokenizer
+    model = Mistral3ForConditionalGeneration.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/modeling_utils.py", line 3991, in from_pretrained
+    model, missing_keys, unexpected_keys, mismatched_keys, offload_index, error_msgs = cls._load_pretrained_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/modeling_utils.py", line 4134, in _load_pretrained_model
+    convert_and_load_state_dict_in_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 1108, in convert_and_load_state_dict_in_model
+    realized_value, conversion_errors = mapping.convert(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 581, in convert
+    collected_tensors = self.materialize_tensors()
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 556, in materialize_tensors
+    tensors = [future.result() for future in tensors]
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 556, in <listcomp>
+    tensors = [future.result() for future in tensors]
+  File "/usr/lib/python3.10/concurrent/futures/_base.py", line 453, in result
+    self._condition.wait(timeout)
+  File "/usr/lib/python3.10/threading.py", line 320, in wait
+    waiter.acquire()
+KeyboardInterrupt
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 921, in <module>
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 787, in main
+    model, tokenizer = load_base_model_and_tokenizer(cfg, base_dir)
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 548, in load_base_model_and_tokenizer
+    model = Mistral3ForConditionalGeneration.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/modeling_utils.py", line 3991, in from_pretrained
+    model, missing_keys, unexpected_keys, mismatched_keys, offload_index, error_msgs = cls._load_pretrained_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/modeling_utils.py", line 4134, in _load_pretrained_model
+    convert_and_load_state_dict_in_model(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 1108, in convert_and_load_state_dict_in_model
+    realized_value, conversion_errors = mapping.convert(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 581, in convert
+    collected_tensors = self.materialize_tensors()
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 556, in materialize_tensors
+    tensors = [future.result() for future in tensors]
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/core_model_loading.py", line 556, in <listcomp>
+    tensors = [future.result() for future in tensors]
+  File "/usr/lib/python3.10/concurrent/futures/_base.py", line 453, in result
+    self._condition.wait(timeout)
+  File "/usr/lib/python3.10/threading.py", line 320, in wait
+    waiter.acquire()
+KeyboardInterrupt
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/requirements.txt b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/wandb-metadata.json b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/wandb-metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..75942473c7339a4e38428423329722f3844c75ed
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/wandb-metadata.json
@@ -0,0 +1,51 @@
+{
+  "os":  "Linux-6.12.46+-x86_64-with-glibc2.35",
+  "python":  "CPython 3.10.12",
+  "startedAt":  "2025-12-23T14:22:08.868345Z",
+  "args":  [
+    "--config",
+    "./SFT/config_instruct.yaml"
+  ],
+  "program":  "/workspace/Trainer-kit/SFT/run_instruct.py",
+  "codePath":  "SFT/run_instruct.py",
+  "codePathLocal":  "SFT/run_instruct.py",
+  "git":  {
+    "remote":  "https://github.com/Ashx098/Trainer-kit.git",
+    "commit":  "92382202abe01938bfeff39a3bf7b82e6e786325"
+  },
+  "email":  "shaiksirajuddin9949@gmail.com",
+  "root":  "runs/instruct_run_24b",
+  "host":  "a100-2gpu-shell-session-9c959f5f-mx8rm",
+  "executable":  "/workspace/CPT-env/bin/python",
+  "cpu_count":  12,
+  "cpu_count_logical":  24,
+  "gpu":  "NVIDIA A100-SXM4-80GB",
+  "gpu_count":  2,
+  "disk":  {
+    "/":  {
+      "total":  "791251738624",
+      "used":  "329782554624"
+    }
+  },
+  "memory":  {
+    "total":  "359047892992"
+  },
+  "gpu_nvidia":  [
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba"
+    },
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40"
+    }
+  ],
+  "cudaVersion":  "13.0",
+  "writerId":  "7eidhw1dhxga1ovk8tpqqdaeudz1qfd9"
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/wandb-summary.json b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/wandb-summary.json
new file mode 100644
index 0000000000000000000000000000000000000000..f4ec3806c8e5abda8d70342c64624c8cf8f81552
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/files/wandb-summary.json
@@ -0,0 +1 @@
+{"_wandb":{"runtime":9},"_runtime":9}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug-core.log b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..fbbe3ea17d9101beb8f4db39bea4d5342e2cef0e
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T14:22:09.132908208Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp4d4ayp08/port-924352.txt","pid":924352,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T14:22:09.133527882Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":924352}
+{"time":"2025-12-23T14:22:09.133532329Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-924352-924448-1422975440/socket","Net":"unix"}}
+{"time":"2025-12-23T14:22:09.319013245Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T14:22:09.325436557Z","level":"INFO","msg":"handleInformInit: received","streamId":"zha9b4tx","id":"1(@)"}
+{"time":"2025-12-23T14:22:09.580526533Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"zha9b4tx","id":"1(@)"}
+{"time":"2025-12-23T14:22:19.588930108Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-23T14:22:19.589052783Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-23T14:22:19.58903178Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-23T14:22:19.589256112Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-23T14:22:19.589261472Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-924352-924448-1422975440/socket","Net":"unix"}}
+{"time":"2025-12-23T14:22:19.883686154Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-23T14:22:19.883722856Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-23T14:22:19.883740737Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug-internal.log b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..905138e46925f18990188bc357728081780f9fff
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug-internal.log
@@ -0,0 +1,11 @@
+{"time":"2025-12-23T14:22:09.325575087Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T14:22:09.580192455Z","level":"INFO","msg":"stream: created new stream","id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:09.58031765Z","level":"INFO","msg":"handler: started","stream_id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:09.580513233Z","level":"INFO","msg":"stream: started","id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:09.580533537Z","level":"INFO","msg":"writer: started","stream_id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:09.580560479Z","level":"INFO","msg":"sender: started","stream_id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:19.589038754Z","level":"INFO","msg":"stream: closing","id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:19.760406906Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2025-12-23T14:22:19.882920539Z","level":"INFO","msg":"handler: closed","stream_id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:19.883028625Z","level":"INFO","msg":"sender: closed","stream_id":"zha9b4tx"}
+{"time":"2025-12-23T14:22:19.883039662Z","level":"INFO","msg":"stream: closed","id":"zha9b4tx"}
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug.log b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..06ecfd1a143219c9cf396008f264cb2d76c3e5fa
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/logs/debug.log
@@ -0,0 +1,23 @@
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_setup.py:_flush():80] Configure stats pid to 924352
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/run-20251223_142208-zha9b4tx/logs/debug.log
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/run-20251223_142208-zha9b4tx/logs/debug-internal.log
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_init.py:init():841] calling init triggers
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 8, 'lora_alpha': 16, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '1e-4', 'weight_decay': 0.0, 'warmup_ratio': 0.1, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 0.8, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 14:22:08,870 INFO    MainThread:924352 [wandb_init.py:init():889] starting backend
+2025-12-23 14:22:09,319 INFO    MainThread:924352 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 14:22:09,323 INFO    MainThread:924352 [wandb_init.py:init():900] backend started and connected
+2025-12-23 14:22:09,325 INFO    MainThread:924352 [wandb_init.py:init():970] updated telemetry
+2025-12-23 14:22:09,329 INFO    MainThread:924352 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 14:22:09,831 INFO    MainThread:924352 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 14:22:09,965 INFO    MainThread:924352 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 14:22:09,965 INFO    MainThread:924352 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 14:22:09,965 INFO    MainThread:924352 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 14:22:09,965 INFO    MainThread:924352 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 14:22:09,969 INFO    MainThread:924352 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 14:22:19,588 INFO    wandb-AsyncioManager-main:924352 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-23 14:22:19,589 INFO    wandb-AsyncioManager-main:924352 [mailbox.py:close():137] Closing mailbox, abandoning 1 handles.
diff --git a/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/run-zha9b4tx.wandb b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/run-zha9b4tx.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..1ab9d1b237378fdfad24608efbc7df28c1492788
Binary files /dev/null and b/sft_devstral_24B/wandb/run-20251223_142208-zha9b4tx/run-zha9b4tx.wandb differ
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/config.yaml b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/config.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..03859fd8b841f3321396d95d4d33e825376ee419
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/config.yaml
@@ -0,0 +1,632 @@
+_name_or_path:
+    value: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+_wandb:
+    value:
+        cli_version: 0.23.1
+        e:
+            dkyx2p7rbyqdypnvz73q5ob1r5e53xmv:
+                args:
+                    - --config
+                    - ./SFT/config_instruct.yaml
+                codePath: SFT/run_instruct.py
+                codePathLocal: SFT/run_instruct.py
+                cpu_count: 12
+                cpu_count_logical: 24
+                cudaVersion: "13.0"
+                disk:
+                    /:
+                        total: "791251738624"
+                        used: "329782792192"
+                email: shaiksirajuddin9949@gmail.com
+                executable: /workspace/CPT-env/bin/python
+                git:
+                    commit: 92382202abe01938bfeff39a3bf7b82e6e786325
+                    remote: https://github.com/Ashx098/Trainer-kit.git
+                gpu: NVIDIA A100-SXM4-80GB
+                gpu_count: 2
+                gpu_nvidia:
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba
+                    - architecture: Ampere
+                      cudaCores: 6912
+                      memoryTotal: "85899345920"
+                      name: NVIDIA A100-SXM4-80GB
+                      uuid: GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40
+                host: a100-2gpu-shell-session-9c959f5f-mx8rm
+                memory:
+                    total: "359047892992"
+                os: Linux-6.12.46+-x86_64-with-glibc2.35
+                program: /workspace/Trainer-kit/SFT/run_instruct.py
+                python: CPython 3.10.12
+                root: runs/instruct_run_24b
+                startedAt: "2025-12-23T14:22:35.730910Z"
+                writerId: dkyx2p7rbyqdypnvz73q5ob1r5e53xmv
+        m:
+            - "1": train/global_step
+              "6":
+                - 3
+              "7": []
+            - "2": '*'
+              "5": 1
+              "6":
+                - 1
+              "7": []
+        python_version: 3.10.12
+        t:
+            "1":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "2":
+                - 1
+                - 11
+                - 49
+                - 51
+                - 71
+                - 98
+            "3":
+                - 7
+                - 15
+                - 16
+                - 19
+                - 62
+                - 66
+            "4": 3.10.12
+            "5": 0.23.1
+            "6": 5.0.0.dev0
+            "9":
+                "1": transformers_trainer
+            "12": 0.23.1
+            "13": linux-x86_64
+accelerator_config:
+    value:
+        dispatch_batches: null
+        even_batches: true
+        gradient_accumulation_kwargs: null
+        non_blocking: false
+        split_batches: false
+        use_seedable_sampler: true
+adam_beta1:
+    value: 0.9
+adam_beta2:
+    value: 0.999
+adam_epsilon:
+    value: 1e-08
+add_cross_attention:
+    value: false
+architectures:
+    value:
+        - Mistral3ForConditionalGeneration
+auto_find_batch_size:
+    value: false
+average_tokens_across_devices:
+    value: true
+batch_eval_metrics:
+    value: false
+bf16:
+    value: true
+bf16_full_eval:
+    value: false
+bos_token_id:
+    value: null
+chunk_size_feed_forward:
+    value: 0
+cross_attention_hidden_size:
+    value: null
+data:
+    value:
+        custom_template: |-
+            ##INSTRUCTION
+            {instruction}<|im_end|>
+            ##TASK
+            {input}<|im_end|>
+            ##OUTPUT
+            {output}<|im_end|>
+        eval_jsonl: null
+        eval_split_ratio: 0.1
+        format_type: custom
+        input_field: input
+        instruction_field: instruction
+        max_length: 2048
+        num_proc: 4
+        output_field: output
+        shuffle: true
+        system_prompt: |
+            You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.
+
+            ## Output Format
+
+            ##OUTPUT
+            Explain the data flow and why each component must change:
+            - Flow: [Input → Processing → Output with arrows]
+            - For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"
+            - Explain coupling between components
+
+            ##SELECT
+            modify::crates/path/to/file.rs::impl::ComponentName
+            add::crates/another/file.rs::function::AnotherComponent
+            <EOS>
+
+            ## Rules
+
+            1. Use full paths: `remove::crates/folder/file.rs::Type::Name`
+            2. Use `::` for nested items: `status::StructName::Type::Name`
+            3. Always explain "must change because" and "without this"
+            3. Types of components: function, struct, enum, impl, trait
+            4. If there is extra information (e.g., enum variants), include that too.
+            5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>
+
+            ## Example
+
+            ##TASK
+            Add webhook subscription support
+
+            ##OUTPUT
+            The webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don't trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.
+
+            ##SELECT
+            crates/common_enums/src/enums.rs::EventClass
+            crates/common_enums/src/transformers.rs::SubscriptionStatus
+            <EOS>
+        train_jsonl: ../sft_dataset.jsonl
+data_seed:
+    value: null
+dataloader_drop_last:
+    value: false
+dataloader_num_workers:
+    value: 0
+dataloader_persistent_workers:
+    value: false
+dataloader_pin_memory:
+    value: true
+dataloader_prefetch_factor:
+    value: null
+ddp_backend:
+    value: null
+ddp_broadcast_buffers:
+    value: null
+ddp_bucket_cap_mb:
+    value: null
+ddp_find_unused_parameters:
+    value: null
+ddp_timeout:
+    value: 1800
+debug:
+    value: []
+decoder_start_token_id:
+    value: null
+deepspeed:
+    value: null
+disable_tqdm:
+    value: false
+do_eval:
+    value: true
+do_predict:
+    value: false
+do_train:
+    value: false
+dtype:
+    value: bfloat16
+enable_jit_checkpoint:
+    value: false
+eos_token_id:
+    value: null
+eval_accumulation_steps:
+    value: null
+eval_delay:
+    value: 0
+eval_do_concat_batches:
+    value: true
+eval_on_start:
+    value: false
+eval_steps:
+    value: 100
+eval_strategy:
+    value: steps
+eval_use_gather_object:
+    value: false
+finetuning_task:
+    value: null
+fp16:
+    value: false
+fp16_full_eval:
+    value: false
+fsdp:
+    value: []
+fsdp_config:
+    value:
+        min_num_params: 0
+        xla: false
+        xla_fsdp_grad_ckpt: false
+        xla_fsdp_v2: false
+full_determinism:
+    value: false
+gradient_accumulation_steps:
+    value: 8
+gradient_checkpointing:
+    value: false
+gradient_checkpointing_kwargs:
+    value: null
+greater_is_better:
+    value: false
+group_by_length:
+    value: false
+hub_always_push:
+    value: false
+hub_model_id:
+    value: null
+hub_private_repo:
+    value: null
+hub_revision:
+    value: null
+hub_strategy:
+    value: every_save
+hub_token:
+    value: <HUB_TOKEN>
+id2label:
+    value:
+        "0": LABEL_0
+        "1": LABEL_1
+ignore_data_skip:
+    value: false
+image_token_index:
+    value: 10
+include_for_metrics:
+    value: []
+include_num_input_tokens_seen:
+    value: "no"
+is_decoder:
+    value: false
+is_encoder_decoder:
+    value: false
+label_names:
+    value: null
+label_smoothing_factor:
+    value: 0
+label2id:
+    value:
+        LABEL_0: 0
+        LABEL_1: 1
+learning_rate:
+    value: 0.0001
+length_column_name:
+    value: length
+liger_kernel_config:
+    value: null
+load_best_model_at_end:
+    value: true
+local_rank:
+    value: -1
+log_level:
+    value: passive
+log_level_replica:
+    value: warning
+log_on_each_node:
+    value: true
+logging_dir:
+    value: null
+logging_first_step:
+    value: false
+logging_nan_inf_filter:
+    value: true
+logging_steps:
+    value: 2
+logging_strategy:
+    value: steps
+lr_scheduler_kwargs:
+    value: null
+lr_scheduler_type:
+    value: cosine
+max_grad_norm:
+    value: 0.8
+max_steps:
+    value: -1
+metric_for_best_model:
+    value: eval_loss
+model:
+    value:
+        attn_implementation: null
+        base_local_dir: base_model
+        bnb_4bit_compute_dtype: bfloat16
+        bnb_4bit_quant_type: nf4
+        bnb_4bit_use_double_quant: false
+        device_map: auto
+        repo_id: ./CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+        revision: null
+        tokenizer_use_fast: true
+        torch_dtype: bfloat16
+        trust_remote_code: true
+        use_4bit: false
+model/num_parameters:
+    value: 24022764544
+model_type:
+    value: mistral3
+multimodal_projector_bias:
+    value: false
+neftune_noise_alpha:
+    value: null
+num_train_epochs:
+    value: 6
+optim:
+    value: adamw_torch
+optim_args:
+    value: null
+optim_target_modules:
+    value: null
+output_attentions:
+    value: false
+output_dir:
+    value: runs/instruct_run_24b/checkpoints
+output_hidden_states:
+    value: false
+pad_token_id:
+    value: null
+parallelism_config:
+    value: null
+peft:
+    value:
+        bias: none
+        enabled: true
+        lora_alpha: 16
+        lora_dropout: 0.05
+        r: 8
+        target_modules: auto
+peft_config:
+    value:
+        default:
+            alora_invocation_tokens: null
+            arrow_config: null
+            auto_mapping: null
+            base_model_name_or_path: CPT/runs/cpt_run_v1/merged_24b_cpt_lora
+            bias: none
+            corda_config: null
+            ensure_weight_tying: false
+            eva_config: null
+            exclude_modules: null
+            fan_in_fan_out: false
+            inference_mode: false
+            init_lora_weights: true
+            layer_replication: null
+            layers_pattern: null
+            layers_to_transform: null
+            lora_alpha: 16
+            lora_bias: false
+            lora_dropout: 0.05
+            megatron_config: null
+            megatron_core: megatron.core
+            modules_to_save: null
+            peft_type: LORA
+            peft_version: 0.18.0
+            qalora_group_size: 16
+            r: 8
+            revision: null
+            runtime_config:
+                ephemeral_gpu_offload: false
+            target_modules:
+                - v_proj
+                - k_proj
+                - q_proj
+                - o_proj
+            target_parameters: null
+            task_type: CAUSAL_LM
+            trainable_token_indices: null
+            use_dora: false
+            use_qalora: false
+            use_rslora: false
+per_device_eval_batch_size:
+    value: 1
+per_device_train_batch_size:
+    value: 1
+prediction_loss_only:
+    value: false
+prefix:
+    value: null
+problem_type:
+    value: null
+project:
+    value: huggingface
+projector_hidden_act:
+    value: gelu
+push_to_hub:
+    value: false
+remove_unused_columns:
+    value: false
+report_to:
+    value:
+        - wandb
+restore_callback_states_from_checkpoint:
+    value: false
+resume_from_checkpoint:
+    value: null
+return_dict:
+    value: true
+run_dir:
+    value: runs/instruct_run_24b
+run_name:
+    value: null
+save_on_each_node:
+    value: false
+save_only_model:
+    value: false
+save_steps:
+    value: 500
+save_strategy:
+    value: steps
+save_total_limit:
+    value: 20
+seed:
+    value: 42
+sep_token_id:
+    value: null
+skip_memory_metrics:
+    value: true
+spatial_merge_size:
+    value: 2
+task_specific_params:
+    value: null
+text_config:
+    value:
+        _name_or_path: ""
+        add_cross_attention: false
+        architectures: null
+        attention_dropout: 0
+        bos_token_id: 1
+        chunk_size_feed_forward: 0
+        cross_attention_hidden_size: null
+        decoder_start_token_id: null
+        dtype: bfloat16
+        eos_token_id: 2
+        finetuning_task: null
+        head_dim: 128
+        hidden_act: silu
+        hidden_size: 5120
+        id2label:
+            "0": LABEL_0
+            "1": LABEL_1
+        initializer_range: 0.02
+        intermediate_size: 32768
+        is_decoder: false
+        is_encoder_decoder: false
+        label2id:
+            LABEL_0: 0
+            LABEL_1: 1
+        max_position_embeddings: 393216
+        model_type: ministral3
+        num_attention_heads: 32
+        num_hidden_layers: 40
+        num_key_value_heads: 8
+        output_attentions: false
+        output_hidden_states: false
+        pad_token_id: 11
+        prefix: null
+        problem_type: null
+        return_dict: true
+        rms_norm_eps: 1e-05
+        rope_parameters:
+            beta_fast: 32
+            beta_slow: 1
+            factor: 48
+            llama_4_scaling_beta: 0.1
+            mscale: 1
+            mscale_all_dim: 1
+            original_max_position_embeddings: 8192
+            rope_theta: 1e+08
+            rope_type: yarn
+            type: yarn
+        sep_token_id: null
+        sliding_window: null
+        task_specific_params: null
+        tie_word_embeddings: false
+        tokenizer_class: null
+        use_cache: true
+        vocab_size: 131072
+tf32:
+    value: null
+tie_word_embeddings:
+    value: false
+tokenizer_class:
+    value: null
+torch_compile:
+    value: false
+torch_compile_backend:
+    value: null
+torch_compile_mode:
+    value: null
+torch_empty_cache_steps:
+    value: null
+trackio_space_id:
+    value: trackio
+train:
+    value:
+        early_stopping:
+            enabled: true
+            metric: eval_loss
+            min_delta: 0.001
+            mode: min
+            patience: 3
+        eval_steps: 100
+        evaluation_strategy: steps
+        gradient_accumulation_steps: 8
+        gradient_checkpointing: true
+        learning_rate: "1e-4"
+        load_best_model_at_end: true
+        logging_steps: 2
+        lr_scheduler_type: cosine
+        max_grad_norm: 0.8
+        num_train_epochs: 6
+        optim: adamw_torch
+        per_device_eval_batch_size: 1
+        per_device_train_batch_size: 1
+        resume_from_checkpoint: auto
+        save_steps: 500
+        save_strategy: steps
+        save_total_limit: 20
+        warmup_ratio: 0.08
+        weight_decay: 0
+transformers_version:
+    value: 5.0.0.dev0
+use_cache:
+    value: false
+use_cpu:
+    value: false
+use_liger_kernel:
+    value: false
+vision_config:
+    value:
+        _name_or_path: ""
+        add_cross_attention: false
+        architectures: null
+        attention_dropout: 0
+        bos_token_id: null
+        chunk_size_feed_forward: 0
+        cross_attention_hidden_size: null
+        decoder_start_token_id: null
+        dtype: bfloat16
+        eos_token_id: null
+        finetuning_task: null
+        head_dim: 64
+        hidden_act: silu
+        hidden_size: 1024
+        id2label:
+            "0": LABEL_0
+            "1": LABEL_1
+        image_size: 1540
+        initializer_range: 0.02
+        intermediate_size: 4096
+        is_decoder: false
+        is_encoder_decoder: false
+        label2id:
+            LABEL_0: 0
+            LABEL_1: 1
+        model_type: pixtral
+        num_attention_heads: 16
+        num_channels: 3
+        num_hidden_layers: 24
+        output_attentions: false
+        output_hidden_states: false
+        pad_token_id: null
+        patch_size: 14
+        prefix: null
+        problem_type: null
+        return_dict: true
+        rope_parameters:
+            rope_theta: 10000
+            rope_type: default
+        sep_token_id: null
+        task_specific_params: null
+        tie_word_embeddings: true
+        tokenizer_class: null
+vision_feature_layer:
+    value: -1
+warmup_ratio:
+    value: 0.08
+warmup_steps:
+    value: 0.08
+weight_decay:
+    value: 0
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/output.log b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/output.log
new file mode 100644
index 0000000000000000000000000000000000000000..4b41e8eb7ea44d99ee762809c0c0528874c759e7
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/output.log
@@ -0,0 +1,2489 @@
+Wandb initialized: project='sft-training', name='auto-generated'
+[info] Detected Mistral3 model architecture, loading with specific class
+Loading weights: 100%|█| 585/585 [00:12<00:00, 46.23it/s, Materializing param=model.vision_tower.transformer.
+[info] Ensuring all parameters are materialized...
+Formatting train instructions (num_proc=4): 100%|█████████████| 18540/18540 [00:02<00:00, 7793.04 examples/s]
+Formatting eval instructions (num_proc=4): 100%|█████████████████| 2060/2060 [00:02<00:00, 907.54 examples/s]
+Tokenizing and masking train (num_proc=4): 100%|██████████████| 18540/18540 [00:15<00:00, 1159.03 examples/s]
+Tokenizing and masking eval (num_proc=4): 100%|██████████████████| 2060/2060 [00:03<00:00, 586.83 examples/s]
+warmup_ratio is deprecated and will be removed in v5.2. Use `warmup_steps` instead.
+Early stopping enabled: patience=3, min_delta=0.001
+Starting instruction fine-tuning...
+  0%|                                                                              | 0/13908 [00:00<?, ?it/s]`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
+                                                                                                             
+{'loss': '1.399', 'grad_norm': '1.912', 'learning_rate': '8.985e-08', 'epoch': '0.000863'}
+{'loss': '1.37', 'grad_norm': '1.785', 'learning_rate': '2.695e-07', 'epoch': '0.001726'}
+{'loss': '1.449', 'grad_norm': '1.686', 'learning_rate': '4.492e-07', 'epoch': '0.002589'}
+{'loss': '1.336', 'grad_norm': '1.735', 'learning_rate': '6.289e-07', 'epoch': '0.003452'}
+{'loss': '1.429', 'grad_norm': '1.777', 'learning_rate': '8.086e-07', 'epoch': '0.004315'}
+{'loss': '1.393', 'grad_norm': '1.777', 'learning_rate': '9.883e-07', 'epoch': '0.005178'}
+{'loss': '1.414', 'grad_norm': '1.5', 'learning_rate': '1.168e-06', 'epoch': '0.006041'}
+{'loss': '1.387', 'grad_norm': '1.726', 'learning_rate': '1.348e-06', 'epoch': '0.006904'}
+{'loss': '1.456', 'grad_norm': '1.907', 'learning_rate': '1.527e-06', 'epoch': '0.007767'}
+{'loss': '1.426', 'grad_norm': '1.725', 'learning_rate': '1.707e-06', 'epoch': '0.00863'}
+{'loss': '1.372', 'grad_norm': '1.486', 'learning_rate': '1.887e-06', 'epoch': '0.009493'}
+{'loss': '1.361', 'grad_norm': '1.752', 'learning_rate': '2.066e-06', 'epoch': '0.01036'}
+{'loss': '1.465', 'grad_norm': '1.755', 'learning_rate': '2.246e-06', 'epoch': '0.01122'}
+{'loss': '1.321', 'grad_norm': '1.52', 'learning_rate': '2.426e-06', 'epoch': '0.01208'}
+{'loss': '1.475', 'grad_norm': '2.03', 'learning_rate': '2.606e-06', 'epoch': '0.01294'}
+{'loss': '1.398', 'grad_norm': '1.615', 'learning_rate': '2.785e-06', 'epoch': '0.01381'}
+{'loss': '1.406', 'grad_norm': '1.703', 'learning_rate': '2.965e-06', 'epoch': '0.01467'}
+{'loss': '1.294', 'grad_norm': '1.572', 'learning_rate': '3.145e-06', 'epoch': '0.01553'}
+{'loss': '1.422', 'grad_norm': '1.707', 'learning_rate': '3.324e-06', 'epoch': '0.0164'}
+{'loss': '1.42', 'grad_norm': '1.605', 'learning_rate': '3.504e-06', 'epoch': '0.01726'}
+{'loss': '1.349', 'grad_norm': '1.511', 'learning_rate': '3.684e-06', 'epoch': '0.01812'}
+{'loss': '1.424', 'grad_norm': '1.556', 'learning_rate': '3.863e-06', 'epoch': '0.01899'}
+{'loss': '1.39', 'grad_norm': '1.581', 'learning_rate': '4.043e-06', 'epoch': '0.01985'}
+{'loss': '1.373', 'grad_norm': '1.533', 'learning_rate': '4.223e-06', 'epoch': '0.02071'}
+{'loss': '1.254', 'grad_norm': '1.567', 'learning_rate': '4.403e-06', 'epoch': '0.02157'}
+{'loss': '1.231', 'grad_norm': '1.371', 'learning_rate': '4.582e-06', 'epoch': '0.02244'}
+{'loss': '1.356', 'grad_norm': '1.567', 'learning_rate': '4.762e-06', 'epoch': '0.0233'}
+{'loss': '1.225', 'grad_norm': '1.355', 'learning_rate': '4.942e-06', 'epoch': '0.02416'}
+{'loss': '1.24', 'grad_norm': '1.442', 'learning_rate': '5.121e-06', 'epoch': '0.02503'}
+{'loss': '1.323', 'grad_norm': '1.464', 'learning_rate': '5.301e-06', 'epoch': '0.02589'}
+{'loss': '1.108', 'grad_norm': '1.22', 'learning_rate': '5.481e-06', 'epoch': '0.02675'}
+{'loss': '1.237', 'grad_norm': '1.264', 'learning_rate': '5.66e-06', 'epoch': '0.02762'}
+{'loss': '1.357', 'grad_norm': '1.464', 'learning_rate': '5.84e-06', 'epoch': '0.02848'}
+{'loss': '1.36', 'grad_norm': '1.581', 'learning_rate': '6.02e-06', 'epoch': '0.02934'}
+{'loss': '1.218', 'grad_norm': '1.435', 'learning_rate': '6.199e-06', 'epoch': '0.0302'}
+{'loss': '1.208', 'grad_norm': '1.456', 'learning_rate': '6.379e-06', 'epoch': '0.03107'}
+{'loss': '1.248', 'grad_norm': '1.287', 'learning_rate': '6.559e-06', 'epoch': '0.03193'}
+{'loss': '1.22', 'grad_norm': '1.335', 'learning_rate': '6.739e-06', 'epoch': '0.03279'}
+{'loss': '1.25', 'grad_norm': '1.482', 'learning_rate': '6.918e-06', 'epoch': '0.03366'}
+{'loss': '1.199', 'grad_norm': '1.242', 'learning_rate': '7.098e-06', 'epoch': '0.03452'}
+{'loss': '1.218', 'grad_norm': '1.504', 'learning_rate': '7.278e-06', 'epoch': '0.03538'}
+{'loss': '1.118', 'grad_norm': '1.305', 'learning_rate': '7.457e-06', 'epoch': '0.03625'}
+{'loss': '1.047', 'grad_norm': '1.33', 'learning_rate': '7.637e-06', 'epoch': '0.03711'}
+{'loss': '1.182', 'grad_norm': '1.452', 'learning_rate': '7.817e-06', 'epoch': '0.03797'}
+{'loss': '1.117', 'grad_norm': '1.544', 'learning_rate': '7.996e-06', 'epoch': '0.03883'}
+{'loss': '1.067', 'grad_norm': '1.527', 'learning_rate': '8.176e-06', 'epoch': '0.0397'}
+{'loss': '1.075', 'grad_norm': '2.44', 'learning_rate': '8.356e-06', 'epoch': '0.04056'}
+{'loss': '1.252', 'grad_norm': '1.77', 'learning_rate': '8.535e-06', 'epoch': '0.04142'}
+{'loss': '1.103', 'grad_norm': '1.432', 'learning_rate': '8.715e-06', 'epoch': '0.04229'}
+{'loss': '1.082', 'grad_norm': '1.326', 'learning_rate': '8.895e-06', 'epoch': '0.04315'}
+                                                                                                             
+{'eval_loss': '1.129', 'eval_runtime': '825', 'eval_samples_per_second': '2.497', 'eval_steps_per_second': '2.497', 'epoch': '0.04315'}
+{'loss': '1.098', 'grad_norm': '1.583', 'learning_rate': '9.075e-06', 'epoch': '0.04401'}
+{'loss': '1.053', 'grad_norm': '1.529', 'learning_rate': '9.254e-06', 'epoch': '0.04488'}
+{'loss': '1.09', 'grad_norm': '1.451', 'learning_rate': '9.434e-06', 'epoch': '0.04574'}
+{'loss': '1.12', 'grad_norm': '1.385', 'learning_rate': '9.614e-06', 'epoch': '0.0466'}
+{'loss': '1.131', 'grad_norm': '1.409', 'learning_rate': '9.793e-06', 'epoch': '0.04746'}
+{'loss': '1.014', 'grad_norm': '1.767', 'learning_rate': '9.973e-06', 'epoch': '0.04833'}
+{'loss': '1.058', 'grad_norm': '1.608', 'learning_rate': '1.015e-05', 'epoch': '0.04919'}
+{'loss': '1.112', 'grad_norm': '1.748', 'learning_rate': '1.033e-05', 'epoch': '0.05005'}
+{'loss': '1.083', 'grad_norm': '1.826', 'learning_rate': '1.051e-05', 'epoch': '0.05092'}
+{'loss': '1.109', 'grad_norm': '1.588', 'learning_rate': '1.069e-05', 'epoch': '0.05178'}
+{'loss': '1.119', 'grad_norm': '1.655', 'learning_rate': '1.087e-05', 'epoch': '0.05264'}
+{'loss': '1.032', 'grad_norm': '1.368', 'learning_rate': '1.105e-05', 'epoch': '0.05351'}
+{'loss': '0.9964', 'grad_norm': '1.534', 'learning_rate': '1.123e-05', 'epoch': '0.05437'}
+{'loss': '1.114', 'grad_norm': '1.593', 'learning_rate': '1.141e-05', 'epoch': '0.05523'}
+{'loss': '0.9954', 'grad_norm': '1.641', 'learning_rate': '1.159e-05', 'epoch': '0.05609'}
+{'loss': '0.994', 'grad_norm': '1.608', 'learning_rate': '1.177e-05', 'epoch': '0.05696'}
+{'loss': '1.046', 'grad_norm': '1.733', 'learning_rate': '1.195e-05', 'epoch': '0.05782'}
+{'loss': '0.9933', 'grad_norm': '1.607', 'learning_rate': '1.213e-05', 'epoch': '0.05868'}
+{'loss': '0.9584', 'grad_norm': '1.506', 'learning_rate': '1.231e-05', 'epoch': '0.05955'}
+{'loss': '0.9733', 'grad_norm': '1.545', 'learning_rate': '1.249e-05', 'epoch': '0.06041'}
+{'loss': '0.9971', 'grad_norm': '1.546', 'learning_rate': '1.267e-05', 'epoch': '0.06127'}
+{'loss': '1.099', 'grad_norm': '1.526', 'learning_rate': '1.285e-05', 'epoch': '0.06214'}
+{'loss': '1.02', 'grad_norm': '1.564', 'learning_rate': '1.303e-05', 'epoch': '0.063'}
+{'loss': '1.022', 'grad_norm': '1.618', 'learning_rate': '1.321e-05', 'epoch': '0.06386'}
+{'loss': '1.021', 'grad_norm': '1.954', 'learning_rate': '1.339e-05', 'epoch': '0.06472'}
+{'loss': '1.025', 'grad_norm': '1.7', 'learning_rate': '1.357e-05', 'epoch': '0.06559'}
+{'loss': '1.008', 'grad_norm': '1.983', 'learning_rate': '1.375e-05', 'epoch': '0.06645'}
+{'loss': '1.005', 'grad_norm': '1.638', 'learning_rate': '1.393e-05', 'epoch': '0.06731'}
+{'loss': '1.033', 'grad_norm': '1.763', 'learning_rate': '1.411e-05', 'epoch': '0.06818'}
+{'loss': '0.973', 'grad_norm': '1.57', 'learning_rate': '1.429e-05', 'epoch': '0.06904'}
+{'loss': '0.9967', 'grad_norm': '1.613', 'learning_rate': '1.447e-05', 'epoch': '0.0699'}
+{'loss': '1.079', 'grad_norm': '1.99', 'learning_rate': '1.465e-05', 'epoch': '0.07077'}
+{'loss': '0.9076', 'grad_norm': '1.82', 'learning_rate': '1.482e-05', 'epoch': '0.07163'}
+{'loss': '0.9269', 'grad_norm': '1.691', 'learning_rate': '1.5e-05', 'epoch': '0.07249'}
+{'loss': '1.036', 'grad_norm': '1.565', 'learning_rate': '1.518e-05', 'epoch': '0.07335'}
+{'loss': '0.9198', 'grad_norm': '1.829', 'learning_rate': '1.536e-05', 'epoch': '0.07422'}
+{'loss': '0.9662', 'grad_norm': '1.688', 'learning_rate': '1.554e-05', 'epoch': '0.07508'}
+{'loss': '1.001', 'grad_norm': '1.696', 'learning_rate': '1.572e-05', 'epoch': '0.07594'}
+{'loss': '1.01', 'grad_norm': '2.013', 'learning_rate': '1.59e-05', 'epoch': '0.07681'}
+{'loss': '0.9727', 'grad_norm': '1.683', 'learning_rate': '1.608e-05', 'epoch': '0.07767'}
+{'loss': '0.9233', 'grad_norm': '1.694', 'learning_rate': '1.626e-05', 'epoch': '0.07853'}
+{'loss': '0.9812', 'grad_norm': '2.085', 'learning_rate': '1.644e-05', 'epoch': '0.0794'}
+{'loss': '0.9763', 'grad_norm': '1.377', 'learning_rate': '1.662e-05', 'epoch': '0.08026'}
+{'loss': '0.9906', 'grad_norm': '1.853', 'learning_rate': '1.68e-05', 'epoch': '0.08112'}
+{'loss': '0.8457', 'grad_norm': '1.654', 'learning_rate': '1.698e-05', 'epoch': '0.08198'}
+{'loss': '1.017', 'grad_norm': '1.772', 'learning_rate': '1.716e-05', 'epoch': '0.08285'}
+{'loss': '0.9768', 'grad_norm': '1.716', 'learning_rate': '1.734e-05', 'epoch': '0.08371'}
+{'loss': '0.9786', 'grad_norm': '1.794', 'learning_rate': '1.752e-05', 'epoch': '0.08457'}
+{'loss': '0.9941', 'grad_norm': '1.944', 'learning_rate': '1.77e-05', 'epoch': '0.08544'}
+{'loss': '0.8924', 'grad_norm': '1.817', 'learning_rate': '1.788e-05', 'epoch': '0.0863'}
+{'eval_loss': '0.9811', 'eval_runtime': '826.4', 'eval_samples_per_second': '2.493', 'eval_steps_per_second': '2.493', 'epoch': '0.0863'}
+{'loss': '0.9908', 'grad_norm': '1.823', 'learning_rate': '1.806e-05', 'epoch': '0.08716'}
+{'loss': '0.9658', 'grad_norm': '1.822', 'learning_rate': '1.824e-05', 'epoch': '0.08803'}
+{'loss': '0.9397', 'grad_norm': '1.811', 'learning_rate': '1.842e-05', 'epoch': '0.08889'}
+{'loss': '0.877', 'grad_norm': '1.58', 'learning_rate': '1.86e-05', 'epoch': '0.08975'}
+{'loss': '0.9228', 'grad_norm': '1.592', 'learning_rate': '1.878e-05', 'epoch': '0.09061'}
+{'loss': '0.9008', 'grad_norm': '1.933', 'learning_rate': '1.896e-05', 'epoch': '0.09148'}
+{'loss': '0.9261', 'grad_norm': '1.862', 'learning_rate': '1.914e-05', 'epoch': '0.09234'}
+{'loss': '0.9319', 'grad_norm': '1.752', 'learning_rate': '1.932e-05', 'epoch': '0.0932'}
+{'loss': '0.9476', 'grad_norm': '2.183', 'learning_rate': '1.95e-05', 'epoch': '0.09407'}
+{'loss': '0.9214', 'grad_norm': '1.897', 'learning_rate': '1.968e-05', 'epoch': '0.09493'}
+{'loss': '0.9058', 'grad_norm': '2', 'learning_rate': '1.986e-05', 'epoch': '0.09579'}
+{'loss': '0.8918', 'grad_norm': '1.937', 'learning_rate': '2.004e-05', 'epoch': '0.09666'}
+{'loss': '1.009', 'grad_norm': '2.193', 'learning_rate': '2.022e-05', 'epoch': '0.09752'}
+{'loss': '0.9544', 'grad_norm': '1.872', 'learning_rate': '2.04e-05', 'epoch': '0.09838'}
+{'loss': '0.9189', 'grad_norm': '2.011', 'learning_rate': '2.058e-05', 'epoch': '0.09924'}
+{'loss': '0.845', 'grad_norm': '1.842', 'learning_rate': '2.075e-05', 'epoch': '0.1001'}
+{'loss': '0.9865', 'grad_norm': '2.209', 'learning_rate': '2.093e-05', 'epoch': '0.101'}
+{'loss': '0.9288', 'grad_norm': '1.612', 'learning_rate': '2.111e-05', 'epoch': '0.1018'}
+{'loss': '0.8854', 'grad_norm': '1.906', 'learning_rate': '2.129e-05', 'epoch': '0.1027'}
+{'loss': '0.9498', 'grad_norm': '2.046', 'learning_rate': '2.147e-05', 'epoch': '0.1036'}
+{'loss': '0.9583', 'grad_norm': '2.161', 'learning_rate': '2.165e-05', 'epoch': '0.1044'}
+{'loss': '1.009', 'grad_norm': '2.14', 'learning_rate': '2.183e-05', 'epoch': '0.1053'}
+{'loss': '0.9627', 'grad_norm': '1.984', 'learning_rate': '2.201e-05', 'epoch': '0.1061'}
+{'loss': '0.926', 'grad_norm': '1.969', 'learning_rate': '2.219e-05', 'epoch': '0.107'}
+{'loss': '1.065', 'grad_norm': '2.192', 'learning_rate': '2.237e-05', 'epoch': '0.1079'}
+{'loss': '0.9681', 'grad_norm': '2.027', 'learning_rate': '2.255e-05', 'epoch': '0.1087'}
+{'loss': '1.003', 'grad_norm': '1.983', 'learning_rate': '2.273e-05', 'epoch': '0.1096'}
+{'loss': '1.005', 'grad_norm': '2.011', 'learning_rate': '2.291e-05', 'epoch': '0.1105'}
+{'loss': '0.8955', 'grad_norm': '1.915', 'learning_rate': '2.309e-05', 'epoch': '0.1113'}
+{'loss': '0.9413', 'grad_norm': '2.101', 'learning_rate': '2.327e-05', 'epoch': '0.1122'}
+{'loss': '0.9293', 'grad_norm': '1.838', 'learning_rate': '2.345e-05', 'epoch': '0.1131'}
+{'loss': '0.9461', 'grad_norm': '1.912', 'learning_rate': '2.363e-05', 'epoch': '0.1139'}
+{'loss': '0.873', 'grad_norm': '1.917', 'learning_rate': '2.381e-05', 'epoch': '0.1148'}
+{'loss': '0.9194', 'grad_norm': '2.118', 'learning_rate': '2.399e-05', 'epoch': '0.1156'}
+{'loss': '0.8994', 'grad_norm': '1.91', 'learning_rate': '2.417e-05', 'epoch': '0.1165'}
+{'loss': '0.851', 'grad_norm': '1.991', 'learning_rate': '2.435e-05', 'epoch': '0.1174'}
+{'loss': '1.022', 'grad_norm': '2.598', 'learning_rate': '2.453e-05', 'epoch': '0.1182'}
+{'loss': '0.8703', 'grad_norm': '1.896', 'learning_rate': '2.471e-05', 'epoch': '0.1191'}
+{'loss': '0.9472', 'grad_norm': '2.116', 'learning_rate': '2.489e-05', 'epoch': '0.12'}
+{'loss': '0.8052', 'grad_norm': '2.026', 'learning_rate': '2.507e-05', 'epoch': '0.1208'}
+{'loss': '0.9008', 'grad_norm': '1.939', 'learning_rate': '2.525e-05', 'epoch': '0.1217'}
+{'loss': '0.9506', 'grad_norm': '2.202', 'learning_rate': '2.543e-05', 'epoch': '0.1225'}
+{'loss': '0.9673', 'grad_norm': '4.048', 'learning_rate': '2.561e-05', 'epoch': '0.1234'}
+{'loss': '0.8251', 'grad_norm': '1.812', 'learning_rate': '2.579e-05', 'epoch': '0.1243'}
+{'loss': '0.9398', 'grad_norm': '1.93', 'learning_rate': '2.597e-05', 'epoch': '0.1251'}
+{'loss': '0.959', 'grad_norm': '2.086', 'learning_rate': '2.615e-05', 'epoch': '0.126'}
+{'loss': '0.9548', 'grad_norm': '2.113', 'learning_rate': '2.633e-05', 'epoch': '0.1269'}
+{'loss': '0.9352', 'grad_norm': '2.067', 'learning_rate': '2.65e-05', 'epoch': '0.1277'}
+{'loss': '0.9268', 'grad_norm': '2.068', 'learning_rate': '2.668e-05', 'epoch': '0.1286'}
+{'loss': '0.8522', 'grad_norm': '2.029', 'learning_rate': '2.686e-05', 'epoch': '0.1294'}
+{'eval_loss': '0.9385', 'eval_runtime': '835.5', 'eval_samples_per_second': '2.466', 'eval_steps_per_second': '2.466', 'epoch': '0.1294'}
+{'loss': '0.946', 'grad_norm': '2.193', 'learning_rate': '2.704e-05', 'epoch': '0.1303'}
+{'loss': '0.9642', 'grad_norm': '2.276', 'learning_rate': '2.722e-05', 'epoch': '0.1312'}
+{'loss': '0.8699', 'grad_norm': '1.912', 'learning_rate': '2.74e-05', 'epoch': '0.132'}
+{'loss': '0.8595', 'grad_norm': '1.968', 'learning_rate': '2.758e-05', 'epoch': '0.1329'}
+{'loss': '0.8541', 'grad_norm': '1.79', 'learning_rate': '2.776e-05', 'epoch': '0.1338'}
+{'loss': '1.015', 'grad_norm': '2.161', 'learning_rate': '2.794e-05', 'epoch': '0.1346'}
+{'loss': '0.9228', 'grad_norm': '2.081', 'learning_rate': '2.812e-05', 'epoch': '0.1355'}
+{'loss': '0.8527', 'grad_norm': '1.963', 'learning_rate': '2.83e-05', 'epoch': '0.1364'}
+{'loss': '0.8137', 'grad_norm': '1.957', 'learning_rate': '2.848e-05', 'epoch': '0.1372'}
+{'loss': '0.9', 'grad_norm': '2.302', 'learning_rate': '2.866e-05', 'epoch': '0.1381'}
+{'loss': '0.9752', 'grad_norm': '2.025', 'learning_rate': '2.884e-05', 'epoch': '0.1389'}
+{'loss': '0.8742', 'grad_norm': '2.02', 'learning_rate': '2.902e-05', 'epoch': '0.1398'}
+{'loss': '0.9481', 'grad_norm': '2.35', 'learning_rate': '2.92e-05', 'epoch': '0.1407'}
+{'loss': '1.009', 'grad_norm': '2.544', 'learning_rate': '2.938e-05', 'epoch': '0.1415'}
+{'loss': '0.9492', 'grad_norm': '3.165', 'learning_rate': '2.956e-05', 'epoch': '0.1424'}
+{'loss': '0.9077', 'grad_norm': '2.055', 'learning_rate': '2.974e-05', 'epoch': '0.1433'}
+{'loss': '0.884', 'grad_norm': '1.679', 'learning_rate': '2.992e-05', 'epoch': '0.1441'}
+{'loss': '0.8558', 'grad_norm': '1.777', 'learning_rate': '3.01e-05', 'epoch': '0.145'}
+{'loss': '0.8477', 'grad_norm': '2.031', 'learning_rate': '3.028e-05', 'epoch': '0.1458'}
+{'loss': '0.9139', 'grad_norm': '2.418', 'learning_rate': '3.046e-05', 'epoch': '0.1467'}
+{'loss': '0.8836', 'grad_norm': '2.095', 'learning_rate': '3.064e-05', 'epoch': '0.1476'}
+{'loss': '0.9319', 'grad_norm': '2.005', 'learning_rate': '3.082e-05', 'epoch': '0.1484'}
+{'loss': '0.8916', 'grad_norm': '2.407', 'learning_rate': '3.1e-05', 'epoch': '0.1493'}
+{'loss': '0.9077', 'grad_norm': '2.119', 'learning_rate': '3.118e-05', 'epoch': '0.1502'}
+{'loss': '0.9211', 'grad_norm': '2.721', 'learning_rate': '3.136e-05', 'epoch': '0.151'}
+{'loss': '0.9134', 'grad_norm': '2.499', 'learning_rate': '3.154e-05', 'epoch': '0.1519'}
+{'loss': '0.9128', 'grad_norm': '2.608', 'learning_rate': '3.172e-05', 'epoch': '0.1528'}
+{'loss': '0.802', 'grad_norm': '1.988', 'learning_rate': '3.19e-05', 'epoch': '0.1536'}
+{'loss': '0.8417', 'grad_norm': '2.059', 'learning_rate': '3.208e-05', 'epoch': '0.1545'}
+{'loss': '0.9097', 'grad_norm': '2.051', 'learning_rate': '3.226e-05', 'epoch': '0.1553'}
+{'loss': '0.9415', 'grad_norm': '2.32', 'learning_rate': '3.243e-05', 'epoch': '0.1562'}
+{'loss': '0.8631', 'grad_norm': '1.966', 'learning_rate': '3.261e-05', 'epoch': '0.1571'}
+{'loss': '0.8997', 'grad_norm': '2.222', 'learning_rate': '3.279e-05', 'epoch': '0.1579'}
+{'loss': '0.8186', 'grad_norm': '1.982', 'learning_rate': '3.297e-05', 'epoch': '0.1588'}
+{'loss': '0.9953', 'grad_norm': '2.32', 'learning_rate': '3.315e-05', 'epoch': '0.1597'}
+{'loss': '0.7604', 'grad_norm': '1.792', 'learning_rate': '3.333e-05', 'epoch': '0.1605'}
+{'loss': '0.7868', 'grad_norm': '1.956', 'learning_rate': '3.351e-05', 'epoch': '0.1614'}
+{'loss': '0.9107', 'grad_norm': '2.41', 'learning_rate': '3.369e-05', 'epoch': '0.1622'}
+{'loss': '0.8946', 'grad_norm': '2.252', 'learning_rate': '3.387e-05', 'epoch': '0.1631'}
+{'loss': '0.895', 'grad_norm': '2.358', 'learning_rate': '3.405e-05', 'epoch': '0.164'}
+{'loss': '0.9966', 'grad_norm': '2.347', 'learning_rate': '3.423e-05', 'epoch': '0.1648'}
+{'loss': '0.9081', 'grad_norm': '2.055', 'learning_rate': '3.441e-05', 'epoch': '0.1657'}
+{'loss': '0.874', 'grad_norm': '2.155', 'learning_rate': '3.459e-05', 'epoch': '0.1666'}
+{'loss': '0.9112', 'grad_norm': '1.693', 'learning_rate': '3.477e-05', 'epoch': '0.1674'}
+{'loss': '0.8916', 'grad_norm': '2.01', 'learning_rate': '3.495e-05', 'epoch': '0.1683'}
+{'loss': '1.016', 'grad_norm': '4.216', 'learning_rate': '3.513e-05', 'epoch': '0.1691'}
+{'loss': '0.7812', 'grad_norm': '1.864', 'learning_rate': '3.531e-05', 'epoch': '0.17'}
+{'loss': '0.9315', 'grad_norm': '1.978', 'learning_rate': '3.549e-05', 'epoch': '0.1709'}
+{'loss': '0.9409', 'grad_norm': '2.076', 'learning_rate': '3.567e-05', 'epoch': '0.1717'}
+{'loss': '0.8159', 'grad_norm': '1.585', 'learning_rate': '3.585e-05', 'epoch': '0.1726'}
+{'eval_loss': '0.9124', 'eval_runtime': '832.6', 'eval_samples_per_second': '2.474', 'eval_steps_per_second': '2.474', 'epoch': '0.1726'}
+{'loss': '0.8901', 'grad_norm': '1.816', 'learning_rate': '3.603e-05', 'epoch': '0.1735'}
+{'loss': '0.8331', 'grad_norm': '1.945', 'learning_rate': '3.621e-05', 'epoch': '0.1743'}
+{'loss': '0.8705', 'grad_norm': '2.028', 'learning_rate': '3.639e-05', 'epoch': '0.1752'}
+{'loss': '0.8065', 'grad_norm': '2.201', 'learning_rate': '3.657e-05', 'epoch': '0.1761'}
+{'loss': '0.8457', 'grad_norm': '2.276', 'learning_rate': '3.675e-05', 'epoch': '0.1769'}
+{'loss': '0.7479', 'grad_norm': '2.097', 'learning_rate': '3.693e-05', 'epoch': '0.1778'}
+{'loss': '0.8634', 'grad_norm': '1.892', 'learning_rate': '3.711e-05', 'epoch': '0.1786'}
+{'loss': '0.8655', 'grad_norm': '2.043', 'learning_rate': '3.729e-05', 'epoch': '0.1795'}
+{'loss': '0.8295', 'grad_norm': '2.005', 'learning_rate': '3.747e-05', 'epoch': '0.1804'}
+{'loss': '0.8729', 'grad_norm': '2.385', 'learning_rate': '3.765e-05', 'epoch': '0.1812'}
+{'loss': '0.8217', 'grad_norm': '1.873', 'learning_rate': '3.783e-05', 'epoch': '0.1821'}
+{'loss': '0.887', 'grad_norm': '2.282', 'learning_rate': '3.801e-05', 'epoch': '0.183'}
+{'loss': '0.9732', 'grad_norm': '2.065', 'learning_rate': '3.819e-05', 'epoch': '0.1838'}
+{'loss': '0.9012', 'grad_norm': '1.999', 'learning_rate': '3.836e-05', 'epoch': '0.1847'}
+{'loss': '0.8675', 'grad_norm': '2.21', 'learning_rate': '3.854e-05', 'epoch': '0.1855'}
+{'loss': '0.8697', 'grad_norm': '2.124', 'learning_rate': '3.872e-05', 'epoch': '0.1864'}
+{'loss': '0.8713', 'grad_norm': '1.771', 'learning_rate': '3.89e-05', 'epoch': '0.1873'}
+{'loss': '0.8638', 'grad_norm': '2.134', 'learning_rate': '3.908e-05', 'epoch': '0.1881'}
+{'loss': '0.8255', 'grad_norm': '1.935', 'learning_rate': '3.926e-05', 'epoch': '0.189'}
+{'loss': '0.9652', 'grad_norm': '1.956', 'learning_rate': '3.944e-05', 'epoch': '0.1899'}
+{'loss': '0.8961', 'grad_norm': '1.973', 'learning_rate': '3.962e-05', 'epoch': '0.1907'}
+{'loss': '0.9555', 'grad_norm': '2.167', 'learning_rate': '3.98e-05', 'epoch': '0.1916'}
+{'loss': '0.8239', 'grad_norm': '2.006', 'learning_rate': '3.998e-05', 'epoch': '0.1924'}
+{'loss': '0.8701', 'grad_norm': '1.98', 'learning_rate': '4.016e-05', 'epoch': '0.1933'}
+{'loss': '0.8928', 'grad_norm': '1.88', 'learning_rate': '4.034e-05', 'epoch': '0.1942'}
+{'loss': '0.8636', 'grad_norm': '1.856', 'learning_rate': '4.052e-05', 'epoch': '0.195'}
+{'loss': '0.9254', 'grad_norm': '2.512', 'learning_rate': '4.07e-05', 'epoch': '0.1959'}
+{'loss': '0.8917', 'grad_norm': '2.083', 'learning_rate': '4.088e-05', 'epoch': '0.1968'}
+{'loss': '0.9098', 'grad_norm': '2.441', 'learning_rate': '4.106e-05', 'epoch': '0.1976'}
+{'loss': '0.8285', 'grad_norm': '2.504', 'learning_rate': '4.124e-05', 'epoch': '0.1985'}
+{'loss': '0.8743', 'grad_norm': '1.936', 'learning_rate': '4.142e-05', 'epoch': '0.1994'}
+{'loss': '0.8379', 'grad_norm': '1.923', 'learning_rate': '4.16e-05', 'epoch': '0.2002'}
+{'loss': '0.8319', 'grad_norm': '2.324', 'learning_rate': '4.178e-05', 'epoch': '0.2011'}
+{'loss': '0.8012', 'grad_norm': '1.88', 'learning_rate': '4.196e-05', 'epoch': '0.2019'}
+{'loss': '0.8671', 'grad_norm': '2.3', 'learning_rate': '4.214e-05', 'epoch': '0.2028'}
+{'loss': '0.883', 'grad_norm': '2.098', 'learning_rate': '4.232e-05', 'epoch': '0.2037'}
+{'loss': '0.9045', 'grad_norm': '2.059', 'learning_rate': '4.25e-05', 'epoch': '0.2045'}
+{'loss': '0.9788', 'grad_norm': '1.883', 'learning_rate': '4.268e-05', 'epoch': '0.2054'}
+{'loss': '0.8875', 'grad_norm': '2.345', 'learning_rate': '4.286e-05', 'epoch': '0.2063'}
+{'loss': '0.7942', 'grad_norm': '1.809', 'learning_rate': '4.304e-05', 'epoch': '0.2071'}
+{'loss': '0.7927', 'grad_norm': '1.84', 'learning_rate': '4.322e-05', 'epoch': '0.208'}
+{'loss': '0.9222', 'grad_norm': '2.394', 'learning_rate': '4.34e-05', 'epoch': '0.2088'}
+{'loss': '0.8716', 'grad_norm': '2.265', 'learning_rate': '4.358e-05', 'epoch': '0.2097'}
+{'loss': '0.8518', 'grad_norm': '2.068', 'learning_rate': '4.376e-05', 'epoch': '0.2106'}
+{'loss': '0.8699', 'grad_norm': '2.265', 'learning_rate': '4.394e-05', 'epoch': '0.2114'}
+{'loss': '0.876', 'grad_norm': '2.082', 'learning_rate': '4.412e-05', 'epoch': '0.2123'}
+{'loss': '0.8957', 'grad_norm': '1.949', 'learning_rate': '4.429e-05', 'epoch': '0.2132'}
+{'loss': '0.9812', 'grad_norm': '2.093', 'learning_rate': '4.447e-05', 'epoch': '0.214'}
+{'loss': '0.8295', 'grad_norm': '2.326', 'learning_rate': '4.465e-05', 'epoch': '0.2149'}
+{'loss': '0.8463', 'grad_norm': '2.251', 'learning_rate': '4.483e-05', 'epoch': '0.2157'}
+{'eval_loss': '0.8945', 'eval_runtime': '823', 'eval_samples_per_second': '2.503', 'eval_steps_per_second': '2.503', 'epoch': '0.2157'}
+{'loss': '0.9099', 'grad_norm': '1.811', 'learning_rate': '4.501e-05', 'epoch': '0.2166'}
+{'loss': '0.8751', 'grad_norm': '2.098', 'learning_rate': '4.519e-05', 'epoch': '0.2175'}
+{'loss': '0.8512', 'grad_norm': '1.865', 'learning_rate': '4.537e-05', 'epoch': '0.2183'}
+{'loss': '0.8408', 'grad_norm': '1.755', 'learning_rate': '4.555e-05', 'epoch': '0.2192'}
+{'loss': '0.8825', 'grad_norm': '1.987', 'learning_rate': '4.573e-05', 'epoch': '0.2201'}
+{'loss': '0.9949', 'grad_norm': '2.46', 'learning_rate': '4.591e-05', 'epoch': '0.2209'}
+{'loss': '0.8979', 'grad_norm': '2.122', 'learning_rate': '4.609e-05', 'epoch': '0.2218'}
+{'loss': '0.8728', 'grad_norm': '1.681', 'learning_rate': '4.627e-05', 'epoch': '0.2227'}
+{'loss': '0.8083', 'grad_norm': '1.651', 'learning_rate': '4.645e-05', 'epoch': '0.2235'}
+{'loss': '0.8042', 'grad_norm': '1.645', 'learning_rate': '4.663e-05', 'epoch': '0.2244'}
+{'loss': '0.9264', 'grad_norm': '2.097', 'learning_rate': '4.681e-05', 'epoch': '0.2252'}
+{'loss': '0.909', 'grad_norm': '1.902', 'learning_rate': '4.699e-05', 'epoch': '0.2261'}
+{'loss': '0.8311', 'grad_norm': '1.984', 'learning_rate': '4.717e-05', 'epoch': '0.227'}
+{'loss': '0.8839', 'grad_norm': '1.788', 'learning_rate': '4.735e-05', 'epoch': '0.2278'}
+{'loss': '0.9431', 'grad_norm': '1.853', 'learning_rate': '4.753e-05', 'epoch': '0.2287'}
+{'loss': '0.7988', 'grad_norm': '1.894', 'learning_rate': '4.771e-05', 'epoch': '0.2296'}
+{'loss': '0.8477', 'grad_norm': '1.92', 'learning_rate': '4.789e-05', 'epoch': '0.2304'}
+{'loss': '0.8482', 'grad_norm': '1.802', 'learning_rate': '4.807e-05', 'epoch': '0.2313'}
+{'loss': '0.8554', 'grad_norm': '1.863', 'learning_rate': '4.825e-05', 'epoch': '0.2321'}
+{'loss': '0.8288', 'grad_norm': '1.864', 'learning_rate': '4.843e-05', 'epoch': '0.233'}
+{'loss': '0.8984', 'grad_norm': '2.163', 'learning_rate': '4.861e-05', 'epoch': '0.2339'}
+{'loss': '0.7253', 'grad_norm': '1.453', 'learning_rate': '4.879e-05', 'epoch': '0.2347'}
+{'loss': '0.9128', 'grad_norm': '2.215', 'learning_rate': '4.897e-05', 'epoch': '0.2356'}
+{'loss': '0.7821', 'grad_norm': '1.764', 'learning_rate': '4.915e-05', 'epoch': '0.2365'}
+{'loss': '0.8212', 'grad_norm': '1.66', 'learning_rate': '4.933e-05', 'epoch': '0.2373'}
+{'loss': '0.9033', 'grad_norm': '1.944', 'learning_rate': '4.951e-05', 'epoch': '0.2382'}
+{'loss': '0.9196', 'grad_norm': '2.033', 'learning_rate': '4.969e-05', 'epoch': '0.2391'}
+{'loss': '0.8828', 'grad_norm': '1.922', 'learning_rate': '4.987e-05', 'epoch': '0.2399'}
+{'loss': '0.909', 'grad_norm': '1.983', 'learning_rate': '5.004e-05', 'epoch': '0.2408'}
+{'loss': '0.8748', 'grad_norm': '1.99', 'learning_rate': '5.022e-05', 'epoch': '0.2416'}
+{'loss': '0.8421', 'grad_norm': '1.796', 'learning_rate': '5.04e-05', 'epoch': '0.2425'}
+{'loss': '0.867', 'grad_norm': '1.98', 'learning_rate': '5.058e-05', 'epoch': '0.2434'}
+{'loss': '0.8597', 'grad_norm': '2.014', 'learning_rate': '5.076e-05', 'epoch': '0.2442'}
+{'loss': '0.8289', 'grad_norm': '2.069', 'learning_rate': '5.094e-05', 'epoch': '0.2451'}
+{'loss': '0.8321', 'grad_norm': '2.691', 'learning_rate': '5.112e-05', 'epoch': '0.246'}
+{'loss': '0.8427', 'grad_norm': '1.806', 'learning_rate': '5.13e-05', 'epoch': '0.2468'}
+{'loss': '0.8431', 'grad_norm': '1.574', 'learning_rate': '5.148e-05', 'epoch': '0.2477'}
+{'loss': '0.9046', 'grad_norm': '2.033', 'learning_rate': '5.166e-05', 'epoch': '0.2485'}
+{'loss': '0.8109', 'grad_norm': '1.764', 'learning_rate': '5.184e-05', 'epoch': '0.2494'}
+{'loss': '0.8621', 'grad_norm': '2.109', 'learning_rate': '5.202e-05', 'epoch': '0.2503'}
+{'loss': '0.8774', 'grad_norm': '2.072', 'learning_rate': '5.22e-05', 'epoch': '0.2511'}
+{'loss': '0.8633', 'grad_norm': '1.809', 'learning_rate': '5.238e-05', 'epoch': '0.252'}
+{'loss': '0.8494', 'grad_norm': '1.712', 'learning_rate': '5.256e-05', 'epoch': '0.2529'}
+{'loss': '0.8279', 'grad_norm': '2.042', 'learning_rate': '5.274e-05', 'epoch': '0.2537'}
+{'loss': '0.8389', 'grad_norm': '1.902', 'learning_rate': '5.292e-05', 'epoch': '0.2546'}
+{'loss': '0.8667', 'grad_norm': '1.807', 'learning_rate': '5.31e-05', 'epoch': '0.2554'}
+{'loss': '0.8114', 'grad_norm': '1.71', 'learning_rate': '5.328e-05', 'epoch': '0.2563'}
+{'loss': '0.8843', 'grad_norm': '1.813', 'learning_rate': '5.346e-05', 'epoch': '0.2572'}
+{'loss': '0.7954', 'grad_norm': '1.958', 'learning_rate': '5.364e-05', 'epoch': '0.258'}
+{'loss': '0.9125', 'grad_norm': '1.832', 'learning_rate': '5.382e-05', 'epoch': '0.2589'}
+{'eval_loss': '0.8771', 'eval_runtime': '827.9', 'eval_samples_per_second': '2.488', 'eval_steps_per_second': '2.488', 'epoch': '0.2589'}
+{'loss': '0.8437', 'grad_norm': '1.757', 'learning_rate': '5.4e-05', 'epoch': '0.2598'}
+{'loss': '0.9384', 'grad_norm': '1.92', 'learning_rate': '5.418e-05', 'epoch': '0.2606'}
+{'loss': '0.8696', 'grad_norm': '1.761', 'learning_rate': '5.436e-05', 'epoch': '0.2615'}
+{'loss': '0.8463', 'grad_norm': '1.772', 'learning_rate': '5.454e-05', 'epoch': '0.2624'}
+{'loss': '0.9075', 'grad_norm': '1.896', 'learning_rate': '5.472e-05', 'epoch': '0.2632'}
+{'loss': '0.8603', 'grad_norm': '1.867', 'learning_rate': '5.49e-05', 'epoch': '0.2641'}
+{'loss': '0.8925', 'grad_norm': '1.89', 'learning_rate': '5.508e-05', 'epoch': '0.2649'}
+{'loss': '0.7873', 'grad_norm': '1.717', 'learning_rate': '5.526e-05', 'epoch': '0.2658'}
+{'loss': '0.7951', 'grad_norm': '1.781', 'learning_rate': '5.544e-05', 'epoch': '0.2667'}
+{'loss': '0.8599', 'grad_norm': '1.871', 'learning_rate': '5.562e-05', 'epoch': '0.2675'}
+{'loss': '0.8953', 'grad_norm': '1.961', 'learning_rate': '5.58e-05', 'epoch': '0.2684'}
+{'loss': '0.8346', 'grad_norm': '1.719', 'learning_rate': '5.597e-05', 'epoch': '0.2693'}
+{'loss': '0.8546', 'grad_norm': '1.981', 'learning_rate': '5.615e-05', 'epoch': '0.2701'}
+{'loss': '0.9064', 'grad_norm': '1.919', 'learning_rate': '5.633e-05', 'epoch': '0.271'}
+{'loss': '0.8831', 'grad_norm': '1.994', 'learning_rate': '5.651e-05', 'epoch': '0.2718'}
+{'loss': '0.8778', 'grad_norm': '1.8', 'learning_rate': '5.669e-05', 'epoch': '0.2727'}
+{'loss': '0.8528', 'grad_norm': '1.916', 'learning_rate': '5.687e-05', 'epoch': '0.2736'}
+{'loss': '0.8811', 'grad_norm': '1.749', 'learning_rate': '5.705e-05', 'epoch': '0.2744'}
+{'loss': '0.8547', 'grad_norm': '1.943', 'learning_rate': '5.723e-05', 'epoch': '0.2753'}
+{'loss': '0.8429', 'grad_norm': '1.754', 'learning_rate': '5.741e-05', 'epoch': '0.2762'}
+{'loss': '0.8587', 'grad_norm': '1.905', 'learning_rate': '5.759e-05', 'epoch': '0.277'}
+{'loss': '0.7796', 'grad_norm': '1.553', 'learning_rate': '5.777e-05', 'epoch': '0.2779'}
+{'loss': '0.7658', 'grad_norm': '1.705', 'learning_rate': '5.795e-05', 'epoch': '0.2787'}
+{'loss': '0.812', 'grad_norm': '1.843', 'learning_rate': '5.813e-05', 'epoch': '0.2796'}
+{'loss': '0.8846', 'grad_norm': '1.95', 'learning_rate': '5.831e-05', 'epoch': '0.2805'}
+{'loss': '0.8543', 'grad_norm': '1.963', 'learning_rate': '5.849e-05', 'epoch': '0.2813'}
+{'loss': '0.7909', 'grad_norm': '1.767', 'learning_rate': '5.867e-05', 'epoch': '0.2822'}
+{'loss': '0.8156', 'grad_norm': '1.887', 'learning_rate': '5.885e-05', 'epoch': '0.2831'}
+{'loss': '0.9833', 'grad_norm': '1.954', 'learning_rate': '5.903e-05', 'epoch': '0.2839'}
+{'loss': '0.7991', 'grad_norm': '1.867', 'learning_rate': '5.921e-05', 'epoch': '0.2848'}
+{'loss': '0.8233', 'grad_norm': '1.849', 'learning_rate': '5.939e-05', 'epoch': '0.2857'}
+{'loss': '0.7653', 'grad_norm': '1.502', 'learning_rate': '5.957e-05', 'epoch': '0.2865'}
+{'loss': '0.822', 'grad_norm': '2.727', 'learning_rate': '5.975e-05', 'epoch': '0.2874'}
+{'loss': '0.7535', 'grad_norm': '1.675', 'learning_rate': '5.993e-05', 'epoch': '0.2882'}
+{'loss': '0.8406', 'grad_norm': '1.995', 'learning_rate': '6.011e-05', 'epoch': '0.2891'}
+{'loss': '0.8143', 'grad_norm': '1.798', 'learning_rate': '6.029e-05', 'epoch': '0.29'}
+{'loss': '0.9602', 'grad_norm': '1.731', 'learning_rate': '6.047e-05', 'epoch': '0.2908'}
+{'loss': '0.8629', 'grad_norm': '1.991', 'learning_rate': '6.065e-05', 'epoch': '0.2917'}
+{'loss': '0.78', 'grad_norm': '1.748', 'learning_rate': '6.083e-05', 'epoch': '0.2926'}
+{'loss': '0.8722', 'grad_norm': '1.798', 'learning_rate': '6.101e-05', 'epoch': '0.2934'}
+{'loss': '0.8896', 'grad_norm': '2.146', 'learning_rate': '6.119e-05', 'epoch': '0.2943'}
+{'loss': '0.8579', 'grad_norm': '1.926', 'learning_rate': '6.137e-05', 'epoch': '0.2951'}
+{'loss': '0.7811', 'grad_norm': '1.446', 'learning_rate': '6.155e-05', 'epoch': '0.296'}
+{'loss': '0.7828', 'grad_norm': '1.62', 'learning_rate': '6.173e-05', 'epoch': '0.2969'}
+{'loss': '0.8186', 'grad_norm': '2.052', 'learning_rate': '6.19e-05', 'epoch': '0.2977'}
+{'loss': '0.8691', 'grad_norm': '1.787', 'learning_rate': '6.208e-05', 'epoch': '0.2986'}
+{'loss': '0.8233', 'grad_norm': '1.961', 'learning_rate': '6.226e-05', 'epoch': '0.2995'}
+{'loss': '0.8493', 'grad_norm': '1.704', 'learning_rate': '6.244e-05', 'epoch': '0.3003'}
+{'loss': '0.8114', 'grad_norm': '1.94', 'learning_rate': '6.262e-05', 'epoch': '0.3012'}
+{'loss': '0.8996', 'grad_norm': '1.853', 'learning_rate': '6.28e-05', 'epoch': '0.302'}
+{'eval_loss': '0.8631', 'eval_runtime': '839.6', 'eval_samples_per_second': '2.453', 'eval_steps_per_second': '2.453', 'epoch': '0.302'}
+{'loss': '0.7978', 'grad_norm': '1.728', 'learning_rate': '6.298e-05', 'epoch': '0.3029'}
+{'loss': '0.812', 'grad_norm': '1.96', 'learning_rate': '6.316e-05', 'epoch': '0.3038'}
+{'loss': '0.8006', 'grad_norm': '1.838', 'learning_rate': '6.334e-05', 'epoch': '0.3046'}
+{'loss': '0.8967', 'grad_norm': '1.821', 'learning_rate': '6.352e-05', 'epoch': '0.3055'}
+{'loss': '0.8272', 'grad_norm': '3.536', 'learning_rate': '6.37e-05', 'epoch': '0.3064'}
+{'loss': '0.8759', 'grad_norm': '1.794', 'learning_rate': '6.388e-05', 'epoch': '0.3072'}
+{'loss': '0.7872', 'grad_norm': '1.548', 'learning_rate': '6.406e-05', 'epoch': '0.3081'}
+{'loss': '0.85', 'grad_norm': '1.804', 'learning_rate': '6.424e-05', 'epoch': '0.309'}
+{'loss': '0.8293', 'grad_norm': '1.548', 'learning_rate': '6.442e-05', 'epoch': '0.3098'}
+{'loss': '0.8797', 'grad_norm': '1.643', 'learning_rate': '6.46e-05', 'epoch': '0.3107'}
+{'loss': '0.8062', 'grad_norm': '1.634', 'learning_rate': '6.478e-05', 'epoch': '0.3115'}
+{'loss': '0.8027', 'grad_norm': '1.771', 'learning_rate': '6.496e-05', 'epoch': '0.3124'}
+{'loss': '0.8736', 'grad_norm': '1.916', 'learning_rate': '6.514e-05', 'epoch': '0.3133'}
+{'loss': '0.7934', 'grad_norm': '1.865', 'learning_rate': '6.532e-05', 'epoch': '0.3141'}
+{'loss': '0.7773', 'grad_norm': '1.592', 'learning_rate': '6.55e-05', 'epoch': '0.315'}
+{'loss': '0.8732', 'grad_norm': '1.856', 'learning_rate': '6.568e-05', 'epoch': '0.3159'}
+{'loss': '0.8168', 'grad_norm': '1.604', 'learning_rate': '6.586e-05', 'epoch': '0.3167'}
+{'loss': '0.781', 'grad_norm': '1.622', 'learning_rate': '6.604e-05', 'epoch': '0.3176'}
+{'loss': '0.8735', 'grad_norm': '1.853', 'learning_rate': '6.622e-05', 'epoch': '0.3184'}
+{'loss': '0.8232', 'grad_norm': '1.473', 'learning_rate': '6.64e-05', 'epoch': '0.3193'}
+{'loss': '0.8401', 'grad_norm': '1.773', 'learning_rate': '6.658e-05', 'epoch': '0.3202'}
+{'loss': '0.7371', 'grad_norm': '1.725', 'learning_rate': '6.676e-05', 'epoch': '0.321'}
+{'loss': '0.8057', 'grad_norm': '1.623', 'learning_rate': '6.694e-05', 'epoch': '0.3219'}
+{'loss': '0.8471', 'grad_norm': '1.607', 'learning_rate': '6.712e-05', 'epoch': '0.3228'}
+{'loss': '0.8251', 'grad_norm': '1.76', 'learning_rate': '6.73e-05', 'epoch': '0.3236'}
+{'loss': '0.8216', 'grad_norm': '1.801', 'learning_rate': '6.748e-05', 'epoch': '0.3245'}
+{'loss': '0.8113', 'grad_norm': '1.77', 'learning_rate': '6.765e-05', 'epoch': '0.3254'}
+{'loss': '0.8894', 'grad_norm': '1.755', 'learning_rate': '6.783e-05', 'epoch': '0.3262'}
+{'loss': '0.8165', 'grad_norm': '1.572', 'learning_rate': '6.801e-05', 'epoch': '0.3271'}
+{'loss': '0.88', 'grad_norm': '2.189', 'learning_rate': '6.819e-05', 'epoch': '0.3279'}
+{'loss': '0.7269', 'grad_norm': '1.634', 'learning_rate': '6.837e-05', 'epoch': '0.3288'}
+{'loss': '0.7696', 'grad_norm': '1.584', 'learning_rate': '6.855e-05', 'epoch': '0.3297'}
+{'loss': '0.9006', 'grad_norm': '1.875', 'learning_rate': '6.873e-05', 'epoch': '0.3305'}
+{'loss': '0.796', 'grad_norm': '1.42', 'learning_rate': '6.891e-05', 'epoch': '0.3314'}
+{'loss': '0.8515', 'grad_norm': '1.667', 'learning_rate': '6.909e-05', 'epoch': '0.3323'}
+{'loss': '0.8403', 'grad_norm': '1.727', 'learning_rate': '6.927e-05', 'epoch': '0.3331'}
+{'loss': '0.7582', 'grad_norm': '1.649', 'learning_rate': '6.945e-05', 'epoch': '0.334'}
+{'loss': '0.825', 'grad_norm': '1.896', 'learning_rate': '6.963e-05', 'epoch': '0.3348'}
+{'loss': '0.7916', 'grad_norm': '1.475', 'learning_rate': '6.981e-05', 'epoch': '0.3357'}
+{'loss': '0.7432', 'grad_norm': '1.489', 'learning_rate': '6.999e-05', 'epoch': '0.3366'}
+{'loss': '0.737', 'grad_norm': '1.515', 'learning_rate': '7.017e-05', 'epoch': '0.3374'}
+{'loss': '0.8348', 'grad_norm': '2.07', 'learning_rate': '7.035e-05', 'epoch': '0.3383'}
+{'loss': '0.7758', 'grad_norm': '1.833', 'learning_rate': '7.053e-05', 'epoch': '0.3392'}
+{'loss': '0.7966', 'grad_norm': '1.681', 'learning_rate': '7.071e-05', 'epoch': '0.34'}
+{'loss': '0.7978', 'grad_norm': '1.581', 'learning_rate': '7.089e-05', 'epoch': '0.3409'}
+{'loss': '0.7843', 'grad_norm': '1.657', 'learning_rate': '7.107e-05', 'epoch': '0.3417'}
+{'loss': '0.7586', 'grad_norm': '1.614', 'learning_rate': '7.125e-05', 'epoch': '0.3426'}
+{'loss': '0.9011', 'grad_norm': '1.92', 'learning_rate': '7.143e-05', 'epoch': '0.3435'}
+{'loss': '0.7857', 'grad_norm': '1.621', 'learning_rate': '7.161e-05', 'epoch': '0.3443'}
+{'loss': '0.9088', 'grad_norm': '2.109', 'learning_rate': '7.179e-05', 'epoch': '0.3452'}
+{'eval_loss': '0.8476', 'eval_runtime': '831.7', 'eval_samples_per_second': '2.477', 'eval_steps_per_second': '2.477', 'epoch': '0.3452'}
+{'loss': '0.7562', 'grad_norm': '1.633', 'learning_rate': '7.197e-05', 'epoch': '0.3461'}
+{'loss': '0.794', 'grad_norm': '1.804', 'learning_rate': '7.215e-05', 'epoch': '0.3469'}
+{'loss': '0.8003', 'grad_norm': '1.494', 'learning_rate': '7.233e-05', 'epoch': '0.3478'}
+{'loss': '0.8211', 'grad_norm': '1.521', 'learning_rate': '7.251e-05', 'epoch': '0.3487'}
+{'loss': '0.869', 'grad_norm': '1.663', 'learning_rate': '7.269e-05', 'epoch': '0.3495'}
+{'loss': '0.826', 'grad_norm': '1.445', 'learning_rate': '7.287e-05', 'epoch': '0.3504'}
+{'loss': '0.7973', 'grad_norm': '1.603', 'learning_rate': '7.305e-05', 'epoch': '0.3512'}
+{'loss': '0.8258', 'grad_norm': '1.594', 'learning_rate': '7.323e-05', 'epoch': '0.3521'}
+{'loss': '0.857', 'grad_norm': '1.628', 'learning_rate': '7.341e-05', 'epoch': '0.353'}
+{'loss': '0.7809', 'grad_norm': '1.399', 'learning_rate': '7.358e-05', 'epoch': '0.3538'}
+{'loss': '0.7514', 'grad_norm': '1.623', 'learning_rate': '7.376e-05', 'epoch': '0.3547'}
+{'loss': '0.8392', 'grad_norm': '3.161', 'learning_rate': '7.394e-05', 'epoch': '0.3556'}
+{'loss': '0.7574', 'grad_norm': '1.517', 'learning_rate': '7.412e-05', 'epoch': '0.3564'}
+{'loss': '0.8395', 'grad_norm': '1.688', 'learning_rate': '7.43e-05', 'epoch': '0.3573'}
+{'loss': '0.7761', 'grad_norm': '1.631', 'learning_rate': '7.448e-05', 'epoch': '0.3581'}
+{'loss': '0.8146', 'grad_norm': '1.482', 'learning_rate': '7.466e-05', 'epoch': '0.359'}
+{'loss': '0.7824', 'grad_norm': '1.963', 'learning_rate': '7.484e-05', 'epoch': '0.3599'}
+{'loss': '0.8955', 'grad_norm': '1.806', 'learning_rate': '7.502e-05', 'epoch': '0.3607'}
+{'loss': '0.7638', 'grad_norm': '1.606', 'learning_rate': '7.52e-05', 'epoch': '0.3616'}
+{'loss': '0.8465', 'grad_norm': '1.533', 'learning_rate': '7.538e-05', 'epoch': '0.3625'}
+{'loss': '0.781', 'grad_norm': '1.564', 'learning_rate': '7.556e-05', 'epoch': '0.3633'}
+{'loss': '0.7432', 'grad_norm': '1.384', 'learning_rate': '7.574e-05', 'epoch': '0.3642'}
+{'loss': '0.8985', 'grad_norm': '1.844', 'learning_rate': '7.592e-05', 'epoch': '0.365'}
+{'loss': '0.781', 'grad_norm': '1.708', 'learning_rate': '7.61e-05', 'epoch': '0.3659'}
+{'loss': '0.7577', 'grad_norm': '1.353', 'learning_rate': '7.628e-05', 'epoch': '0.3668'}
+{'loss': '0.7667', 'grad_norm': '1.439', 'learning_rate': '7.646e-05', 'epoch': '0.3676'}
+{'loss': '0.8356', 'grad_norm': '1.627', 'learning_rate': '7.664e-05', 'epoch': '0.3685'}
+{'loss': '0.7766', 'grad_norm': '1.555', 'learning_rate': '7.682e-05', 'epoch': '0.3694'}
+{'loss': '0.7647', 'grad_norm': '1.468', 'learning_rate': '7.7e-05', 'epoch': '0.3702'}
+{'loss': '0.8717', 'grad_norm': '1.645', 'learning_rate': '7.718e-05', 'epoch': '0.3711'}
+{'loss': '0.8562', 'grad_norm': '1.6', 'learning_rate': '7.736e-05', 'epoch': '0.372'}
+{'loss': '0.8795', 'grad_norm': '1.632', 'learning_rate': '7.754e-05', 'epoch': '0.3728'}
+{'loss': '0.878', 'grad_norm': '1.717', 'learning_rate': '7.772e-05', 'epoch': '0.3737'}
+{'loss': '0.8812', 'grad_norm': '1.859', 'learning_rate': '7.79e-05', 'epoch': '0.3745'}
+{'loss': '0.8392', 'grad_norm': '1.805', 'learning_rate': '7.808e-05', 'epoch': '0.3754'}
+{'loss': '0.8426', 'grad_norm': '2.293', 'learning_rate': '7.826e-05', 'epoch': '0.3763'}
+{'loss': '0.841', 'grad_norm': '1.552', 'learning_rate': '7.844e-05', 'epoch': '0.3771'}
+{'loss': '0.7442', 'grad_norm': '1.466', 'learning_rate': '7.862e-05', 'epoch': '0.378'}
+{'loss': '0.885', 'grad_norm': '1.789', 'learning_rate': '7.88e-05', 'epoch': '0.3789'}
+{'loss': '0.7991', 'grad_norm': '1.474', 'learning_rate': '7.898e-05', 'epoch': '0.3797'}
+{'loss': '0.7904', 'grad_norm': '1.684', 'learning_rate': '7.916e-05', 'epoch': '0.3806'}
+{'loss': '0.8122', 'grad_norm': '1.675', 'learning_rate': '7.934e-05', 'epoch': '0.3814'}
+{'loss': '0.7819', 'grad_norm': '1.655', 'learning_rate': '7.951e-05', 'epoch': '0.3823'}
+{'loss': '0.8223', 'grad_norm': '1.375', 'learning_rate': '7.969e-05', 'epoch': '0.3832'}
+{'loss': '0.7171', 'grad_norm': '1.422', 'learning_rate': '7.987e-05', 'epoch': '0.384'}
+{'loss': '0.8534', 'grad_norm': '1.713', 'learning_rate': '8.005e-05', 'epoch': '0.3849'}
+{'loss': '0.88', 'grad_norm': '1.552', 'learning_rate': '8.023e-05', 'epoch': '0.3858'}
+{'loss': '0.7969', 'grad_norm': '1.261', 'learning_rate': '8.041e-05', 'epoch': '0.3866'}
+{'loss': '0.7911', 'grad_norm': '1.477', 'learning_rate': '8.059e-05', 'epoch': '0.3875'}
+{'loss': '0.7873', 'grad_norm': '1.464', 'learning_rate': '8.077e-05', 'epoch': '0.3883'}
+{'eval_loss': '0.8352', 'eval_runtime': '823.6', 'eval_samples_per_second': '2.501', 'eval_steps_per_second': '2.501', 'epoch': '0.3883'}
+{'loss': '0.8184', 'grad_norm': '1.378', 'learning_rate': '8.095e-05', 'epoch': '0.3892'}
+{'loss': '0.7861', 'grad_norm': '1.59', 'learning_rate': '8.113e-05', 'epoch': '0.3901'}
+{'loss': '0.8638', 'grad_norm': '1.634', 'learning_rate': '8.131e-05', 'epoch': '0.3909'}
+{'loss': '0.7028', 'grad_norm': '1.399', 'learning_rate': '8.149e-05', 'epoch': '0.3918'}
+{'loss': '0.7845', 'grad_norm': '1.37', 'learning_rate': '8.167e-05', 'epoch': '0.3927'}
+{'loss': '0.8117', 'grad_norm': '1.419', 'learning_rate': '8.185e-05', 'epoch': '0.3935'}
+{'loss': '0.8229', 'grad_norm': '1.472', 'learning_rate': '8.203e-05', 'epoch': '0.3944'}
+{'loss': '0.6988', 'grad_norm': '1.513', 'learning_rate': '8.221e-05', 'epoch': '0.3953'}
+{'loss': '0.7694', 'grad_norm': '1.476', 'learning_rate': '8.239e-05', 'epoch': '0.3961'}
+{'loss': '0.7793', 'grad_norm': '1.53', 'learning_rate': '8.257e-05', 'epoch': '0.397'}
+{'loss': '0.7137', 'grad_norm': '1.475', 'learning_rate': '8.275e-05', 'epoch': '0.3978'}
+{'loss': '0.787', 'grad_norm': '1.402', 'learning_rate': '8.293e-05', 'epoch': '0.3987'}
+{'loss': '0.904', 'grad_norm': '1.573', 'learning_rate': '8.311e-05', 'epoch': '0.3996'}
+{'loss': '0.759', 'grad_norm': '1.61', 'learning_rate': '8.329e-05', 'epoch': '0.4004'}
+{'loss': '0.8726', 'grad_norm': '1.779', 'learning_rate': '8.347e-05', 'epoch': '0.4013'}
+{'loss': '0.828', 'grad_norm': '1.633', 'learning_rate': '8.365e-05', 'epoch': '0.4022'}
+{'loss': '0.8293', 'grad_norm': '1.644', 'learning_rate': '8.383e-05', 'epoch': '0.403'}
+{'loss': '0.7628', 'grad_norm': '1.282', 'learning_rate': '8.401e-05', 'epoch': '0.4039'}
+{'loss': '0.8951', 'grad_norm': '1.785', 'learning_rate': '8.419e-05', 'epoch': '0.4047'}
+{'loss': '0.7678', 'grad_norm': '1.801', 'learning_rate': '8.437e-05', 'epoch': '0.4056'}
+{'loss': '0.8553', 'grad_norm': '1.5', 'learning_rate': '8.455e-05', 'epoch': '0.4065'}
+{'loss': '0.7111', 'grad_norm': '1.34', 'learning_rate': '8.473e-05', 'epoch': '0.4073'}
+{'loss': '0.938', 'grad_norm': '1.894', 'learning_rate': '8.491e-05', 'epoch': '0.4082'}
+{'loss': '0.7493', 'grad_norm': '1.265', 'learning_rate': '8.509e-05', 'epoch': '0.4091'}
+{'loss': '0.7922', 'grad_norm': '1.559', 'learning_rate': '8.527e-05', 'epoch': '0.4099'}
+{'loss': '0.7494', 'grad_norm': '1.504', 'learning_rate': '8.544e-05', 'epoch': '0.4108'}
+{'loss': '0.7984', 'grad_norm': '1.434', 'learning_rate': '8.562e-05', 'epoch': '0.4117'}
+{'loss': '0.8133', 'grad_norm': '1.614', 'learning_rate': '8.58e-05', 'epoch': '0.4125'}
+{'loss': '0.7795', 'grad_norm': '1.376', 'learning_rate': '8.598e-05', 'epoch': '0.4134'}
+{'loss': '0.7545', 'grad_norm': '1.515', 'learning_rate': '8.616e-05', 'epoch': '0.4142'}
+{'loss': '0.8151', 'grad_norm': '1.37', 'learning_rate': '8.634e-05', 'epoch': '0.4151'}
+{'loss': '0.7997', 'grad_norm': '1.551', 'learning_rate': '8.652e-05', 'epoch': '0.416'}
+{'loss': '0.7678', 'grad_norm': '1.297', 'learning_rate': '8.67e-05', 'epoch': '0.4168'}
+{'loss': '0.7537', 'grad_norm': '1.244', 'learning_rate': '8.688e-05', 'epoch': '0.4177'}
+{'loss': '0.7796', 'grad_norm': '1.439', 'learning_rate': '8.706e-05', 'epoch': '0.4186'}
+{'loss': '0.7813', 'grad_norm': '1.482', 'learning_rate': '8.724e-05', 'epoch': '0.4194'}
+{'loss': '0.7396', 'grad_norm': '1.377', 'learning_rate': '8.742e-05', 'epoch': '0.4203'}
+{'loss': '0.8391', 'grad_norm': '1.553', 'learning_rate': '8.76e-05', 'epoch': '0.4211'}
+{'loss': '0.8369', 'grad_norm': '1.543', 'learning_rate': '8.778e-05', 'epoch': '0.422'}
+{'loss': '0.8203', 'grad_norm': '1.358', 'learning_rate': '8.796e-05', 'epoch': '0.4229'}
+{'loss': '0.7623', 'grad_norm': '1.554', 'learning_rate': '8.814e-05', 'epoch': '0.4237'}
+{'loss': '0.7765', 'grad_norm': '1.573', 'learning_rate': '8.832e-05', 'epoch': '0.4246'}
+{'loss': '0.8136', 'grad_norm': '1.584', 'learning_rate': '8.85e-05', 'epoch': '0.4255'}
+{'loss': '0.8388', 'grad_norm': '1.572', 'learning_rate': '8.868e-05', 'epoch': '0.4263'}
+{'loss': '0.8829', 'grad_norm': '1.668', 'learning_rate': '8.886e-05', 'epoch': '0.4272'}
+{'loss': '0.8545', 'grad_norm': '1.71', 'learning_rate': '8.904e-05', 'epoch': '0.428'}
+{'loss': '0.8234', 'grad_norm': '1.516', 'learning_rate': '8.922e-05', 'epoch': '0.4289'}
+{'loss': '0.7577', 'grad_norm': '1.458', 'learning_rate': '8.94e-05', 'epoch': '0.4298'}
+{'loss': '0.809', 'grad_norm': '1.638', 'learning_rate': '8.958e-05', 'epoch': '0.4306'}
+{'loss': '0.7792', 'grad_norm': '1.415', 'learning_rate': '8.976e-05', 'epoch': '0.4315'}
+{'eval_loss': '0.8266', 'eval_runtime': '835', 'eval_samples_per_second': '2.467', 'eval_steps_per_second': '2.467', 'epoch': '0.4315'}
+{'loss': '0.7444', 'grad_norm': '1.548', 'learning_rate': '8.994e-05', 'epoch': '0.4324'}
+{'loss': '0.8211', 'grad_norm': '1.462', 'learning_rate': '9.012e-05', 'epoch': '0.4332'}
+{'loss': '0.8131', 'grad_norm': '1.365', 'learning_rate': '9.03e-05', 'epoch': '0.4341'}
+{'loss': '0.7915', 'grad_norm': '1.359', 'learning_rate': '9.048e-05', 'epoch': '0.435'}
+{'loss': '0.8287', 'grad_norm': '1.465', 'learning_rate': '9.066e-05', 'epoch': '0.4358'}
+{'loss': '0.7173', 'grad_norm': '1.391', 'learning_rate': '9.084e-05', 'epoch': '0.4367'}
+{'loss': '0.7294', 'grad_norm': '1.205', 'learning_rate': '9.102e-05', 'epoch': '0.4375'}
+{'loss': '0.7367', 'grad_norm': '1.345', 'learning_rate': '9.119e-05', 'epoch': '0.4384'}
+{'loss': '0.8551', 'grad_norm': '2.901', 'learning_rate': '9.137e-05', 'epoch': '0.4393'}
+{'loss': '0.8944', 'grad_norm': '1.665', 'learning_rate': '9.155e-05', 'epoch': '0.4401'}
+{'loss': '0.7644', 'grad_norm': '1.305', 'learning_rate': '9.173e-05', 'epoch': '0.441'}
+{'loss': '0.8645', 'grad_norm': '1.464', 'learning_rate': '9.191e-05', 'epoch': '0.4419'}
+{'loss': '0.8281', 'grad_norm': '1.469', 'learning_rate': '9.209e-05', 'epoch': '0.4427'}
+{'loss': '0.7432', 'grad_norm': '1.332', 'learning_rate': '9.227e-05', 'epoch': '0.4436'}
+{'loss': '0.8144', 'grad_norm': '1.434', 'learning_rate': '9.245e-05', 'epoch': '0.4444'}
+{'loss': '0.8102', 'grad_norm': '1.747', 'learning_rate': '9.263e-05', 'epoch': '0.4453'}
+{'loss': '0.7768', 'grad_norm': '1.327', 'learning_rate': '9.281e-05', 'epoch': '0.4462'}
+{'loss': '0.7887', 'grad_norm': '1.502', 'learning_rate': '9.299e-05', 'epoch': '0.447'}
+{'loss': '0.8152', 'grad_norm': '1.671', 'learning_rate': '9.317e-05', 'epoch': '0.4479'}
+{'loss': '0.8339', 'grad_norm': '1.85', 'learning_rate': '9.335e-05', 'epoch': '0.4488'}
+{'loss': '0.7799', 'grad_norm': '1.433', 'learning_rate': '9.353e-05', 'epoch': '0.4496'}
+{'loss': '0.7677', 'grad_norm': '1.336', 'learning_rate': '9.371e-05', 'epoch': '0.4505'}
+{'loss': '0.7729', 'grad_norm': '1.339', 'learning_rate': '9.389e-05', 'epoch': '0.4513'}
+{'loss': '0.7924', 'grad_norm': '1.435', 'learning_rate': '9.407e-05', 'epoch': '0.4522'}
+{'loss': '0.8069', 'grad_norm': '1.445', 'learning_rate': '9.425e-05', 'epoch': '0.4531'}
+{'loss': '0.7533', 'grad_norm': '1.19', 'learning_rate': '9.443e-05', 'epoch': '0.4539'}
+{'loss': '0.8237', 'grad_norm': '1.41', 'learning_rate': '9.461e-05', 'epoch': '0.4548'}
+{'loss': '0.7533', 'grad_norm': '1.376', 'learning_rate': '9.479e-05', 'epoch': '0.4557'}
+{'loss': '0.8506', 'grad_norm': '1.461', 'learning_rate': '9.497e-05', 'epoch': '0.4565'}
+{'loss': '0.7825', 'grad_norm': '1.721', 'learning_rate': '9.515e-05', 'epoch': '0.4574'}
+{'loss': '0.8674', 'grad_norm': '1.595', 'learning_rate': '9.533e-05', 'epoch': '0.4583'}
+{'loss': '0.7852', 'grad_norm': '1.355', 'learning_rate': '9.551e-05', 'epoch': '0.4591'}
+{'loss': '0.8891', 'grad_norm': '1.415', 'learning_rate': '9.569e-05', 'epoch': '0.46'}
+{'loss': '0.8227', 'grad_norm': '1.332', 'learning_rate': '9.587e-05', 'epoch': '0.4608'}
+{'loss': '0.7663', 'grad_norm': '1.414', 'learning_rate': '9.605e-05', 'epoch': '0.4617'}
+{'loss': '0.7123', 'grad_norm': '1.325', 'learning_rate': '9.623e-05', 'epoch': '0.4626'}
+{'loss': '0.7178', 'grad_norm': '1.205', 'learning_rate': '9.641e-05', 'epoch': '0.4634'}
+{'loss': '0.7619', 'grad_norm': '1.522', 'learning_rate': '9.659e-05', 'epoch': '0.4643'}
+{'loss': '0.7484', 'grad_norm': '1.352', 'learning_rate': '9.677e-05', 'epoch': '0.4652'}
+{'loss': '0.8508', 'grad_norm': '1.748', 'learning_rate': '9.695e-05', 'epoch': '0.466'}
+{'loss': '0.7894', 'grad_norm': '1.418', 'learning_rate': '9.712e-05', 'epoch': '0.4669'}
+{'loss': '0.7891', 'grad_norm': '1.629', 'learning_rate': '9.73e-05', 'epoch': '0.4677'}
+{'loss': '0.841', 'grad_norm': '1.661', 'learning_rate': '9.748e-05', 'epoch': '0.4686'}
+{'loss': '0.8099', 'grad_norm': '1.389', 'learning_rate': '9.766e-05', 'epoch': '0.4695'}
+{'loss': '0.7178', 'grad_norm': '1.39', 'learning_rate': '9.784e-05', 'epoch': '0.4703'}
+{'loss': '0.8286', 'grad_norm': '1.334', 'learning_rate': '9.802e-05', 'epoch': '0.4712'}
+{'loss': '0.7583', 'grad_norm': '1.545', 'learning_rate': '9.82e-05', 'epoch': '0.4721'}
+{'loss': '0.7293', 'grad_norm': '1.54', 'learning_rate': '9.838e-05', 'epoch': '0.4729'}
+{'loss': '0.8858', 'grad_norm': '1.501', 'learning_rate': '9.856e-05', 'epoch': '0.4738'}
+{'loss': '0.7176', 'grad_norm': '1.356', 'learning_rate': '9.874e-05', 'epoch': '0.4746'}
+{'eval_loss': '0.8152', 'eval_runtime': '835', 'eval_samples_per_second': '2.467', 'eval_steps_per_second': '2.467', 'epoch': '0.4746'}
+{'loss': '0.8049', 'grad_norm': '1.523', 'learning_rate': '9.892e-05', 'epoch': '0.4755'}
+{'loss': '0.7954', 'grad_norm': '1.354', 'learning_rate': '9.91e-05', 'epoch': '0.4764'}
+{'loss': '0.804', 'grad_norm': '1.517', 'learning_rate': '9.928e-05', 'epoch': '0.4772'}
+{'loss': '0.7652', 'grad_norm': '1.401', 'learning_rate': '9.946e-05', 'epoch': '0.4781'}
+{'loss': '0.8501', 'grad_norm': '1.416', 'learning_rate': '9.964e-05', 'epoch': '0.479'}
+{'loss': '0.9024', 'grad_norm': '1.741', 'learning_rate': '9.982e-05', 'epoch': '0.4798'}
+{'loss': '0.7308', 'grad_norm': '1.149', 'learning_rate': '0.0001', 'epoch': '0.4807'}
+{'loss': '0.7274', 'grad_norm': '1.251', 'learning_rate': '0.0001', 'epoch': '0.4816'}
+{'loss': '0.8072', 'grad_norm': '1.394', 'learning_rate': '0.0001', 'epoch': '0.4824'}
+{'loss': '0.6587', 'grad_norm': '1.221', 'learning_rate': '0.0001', 'epoch': '0.4833'}
+{'loss': '0.8529', 'grad_norm': '1.733', 'learning_rate': '0.0001', 'epoch': '0.4841'}
+{'loss': '0.753', 'grad_norm': '1.331', 'learning_rate': '0.0001', 'epoch': '0.485'}
+{'loss': '0.678', 'grad_norm': '1.22', 'learning_rate': '0.0001', 'epoch': '0.4859'}
+{'loss': '0.7942', 'grad_norm': '1.975', 'learning_rate': '0.0001', 'epoch': '0.4867'}
+{'loss': '0.7895', 'grad_norm': '1.494', 'learning_rate': '0.0001', 'epoch': '0.4876'}
+{'loss': '0.7662', 'grad_norm': '1.424', 'learning_rate': '0.0001', 'epoch': '0.4885'}
+{'loss': '0.8079', 'grad_norm': '1.25', 'learning_rate': '0.0001', 'epoch': '0.4893'}
+{'loss': '0.7616', 'grad_norm': '1.225', 'learning_rate': '0.0001', 'epoch': '0.4902'}
+{'loss': '0.8577', 'grad_norm': '1.218', 'learning_rate': '0.0001', 'epoch': '0.491'}
+{'loss': '0.806', 'grad_norm': '1.211', 'learning_rate': '0.0001', 'epoch': '0.4919'}
+{'loss': '0.7533', 'grad_norm': '1.634', 'learning_rate': '0.0001', 'epoch': '0.4928'}
+{'loss': '0.7242', 'grad_norm': '1.262', 'learning_rate': '0.0001', 'epoch': '0.4936'}
+{'loss': '0.7866', 'grad_norm': '1.268', 'learning_rate': '0.0001', 'epoch': '0.4945'}
+{'loss': '0.7423', 'grad_norm': '1.419', 'learning_rate': '0.0001', 'epoch': '0.4954'}
+{'loss': '0.8178', 'grad_norm': '1.322', 'learning_rate': '0.0001', 'epoch': '0.4962'}
+{'loss': '0.8016', 'grad_norm': '1.97', 'learning_rate': '0.0001', 'epoch': '0.4971'}
+{'loss': '0.7824', 'grad_norm': '1.223', 'learning_rate': '0.0001', 'epoch': '0.498'}
+{'loss': '0.7236', 'grad_norm': '1.275', 'learning_rate': '0.0001', 'epoch': '0.4988'}
+{'loss': '0.7686', 'grad_norm': '1.353', 'learning_rate': '0.0001', 'epoch': '0.4997'}
+{'loss': '0.7843', 'grad_norm': '1.424', 'learning_rate': '0.0001', 'epoch': '0.5005'}
+{'loss': '0.8324', 'grad_norm': '1.502', 'learning_rate': '0.0001', 'epoch': '0.5014'}
+{'loss': '0.7798', 'grad_norm': '1.54', 'learning_rate': '0.0001', 'epoch': '0.5023'}
+{'loss': '0.7667', 'grad_norm': '1.408', 'learning_rate': '0.0001', 'epoch': '0.5031'}
+{'loss': '0.8153', 'grad_norm': '1.441', 'learning_rate': '0.0001', 'epoch': '0.504'}
+{'loss': '0.8549', 'grad_norm': '1.42', 'learning_rate': '0.0001', 'epoch': '0.5049'}
+{'loss': '0.7978', 'grad_norm': '1.503', 'learning_rate': '9.999e-05', 'epoch': '0.5057'}
+{'loss': '0.799', 'grad_norm': '1.484', 'learning_rate': '9.999e-05', 'epoch': '0.5066'}
+{'loss': '0.8073', 'grad_norm': '1.542', 'learning_rate': '9.999e-05', 'epoch': '0.5074'}
+{'loss': '0.7216', 'grad_norm': '1.13', 'learning_rate': '9.999e-05', 'epoch': '0.5083'}
+{'loss': '0.8637', 'grad_norm': '1.321', 'learning_rate': '9.999e-05', 'epoch': '0.5092'}
+{'loss': '0.8561', 'grad_norm': '1.524', 'learning_rate': '9.999e-05', 'epoch': '0.51'}
+{'loss': '0.7515', 'grad_norm': '1.319', 'learning_rate': '9.999e-05', 'epoch': '0.5109'}
+{'loss': '0.8635', 'grad_norm': '1.313', 'learning_rate': '9.999e-05', 'epoch': '0.5118'}
+{'loss': '0.7494', 'grad_norm': '1.268', 'learning_rate': '9.999e-05', 'epoch': '0.5126'}
+{'loss': '0.8981', 'grad_norm': '1.39', 'learning_rate': '9.999e-05', 'epoch': '0.5135'}
+{'loss': '0.6771', 'grad_norm': '1.232', 'learning_rate': '9.999e-05', 'epoch': '0.5143'}
+{'loss': '0.7899', 'grad_norm': '1.13', 'learning_rate': '9.999e-05', 'epoch': '0.5152'}
+{'loss': '0.6656', 'grad_norm': '1.582', 'learning_rate': '9.999e-05', 'epoch': '0.5161'}
+{'loss': '0.7956', 'grad_norm': '1.344', 'learning_rate': '9.999e-05', 'epoch': '0.5169'}
+{'loss': '0.777', 'grad_norm': '1.451', 'learning_rate': '9.999e-05', 'epoch': '0.5178'}
+{'eval_loss': '0.807', 'eval_runtime': '823.9', 'eval_samples_per_second': '2.5', 'eval_steps_per_second': '2.5', 'epoch': '0.5178'}
+{'loss': '0.7527', 'grad_norm': '1.379', 'learning_rate': '9.999e-05', 'epoch': '0.5187'}
+{'loss': '0.8022', 'grad_norm': '1.257', 'learning_rate': '9.999e-05', 'epoch': '0.5195'}
+{'loss': '0.7766', 'grad_norm': '1.331', 'learning_rate': '9.999e-05', 'epoch': '0.5204'}
+{'loss': '0.7381', 'grad_norm': '1.233', 'learning_rate': '9.999e-05', 'epoch': '0.5213'}
+{'loss': '0.6596', 'grad_norm': '1.048', 'learning_rate': '9.999e-05', 'epoch': '0.5221'}
+{'loss': '0.8389', 'grad_norm': '1.546', 'learning_rate': '9.999e-05', 'epoch': '0.523'}
+{'loss': '0.7594', 'grad_norm': '1.373', 'learning_rate': '9.998e-05', 'epoch': '0.5238'}
+{'loss': '0.7797', 'grad_norm': '1.267', 'learning_rate': '9.998e-05', 'epoch': '0.5247'}
+{'loss': '0.7494', 'grad_norm': '1.118', 'learning_rate': '9.998e-05', 'epoch': '0.5256'}
+{'loss': '0.7791', 'grad_norm': '1.122', 'learning_rate': '9.998e-05', 'epoch': '0.5264'}
+{'loss': '0.8299', 'grad_norm': '1.498', 'learning_rate': '9.998e-05', 'epoch': '0.5273'}
+{'loss': '0.7911', 'grad_norm': '1.227', 'learning_rate': '9.998e-05', 'epoch': '0.5282'}
+{'loss': '0.6488', 'grad_norm': '1.176', 'learning_rate': '9.998e-05', 'epoch': '0.529'}
+{'loss': '0.7332', 'grad_norm': '1.17', 'learning_rate': '9.998e-05', 'epoch': '0.5299'}
+{'loss': '0.8198', 'grad_norm': '1.409', 'learning_rate': '9.998e-05', 'epoch': '0.5307'}
+{'loss': '0.71', 'grad_norm': '1.863', 'learning_rate': '9.998e-05', 'epoch': '0.5316'}
+{'loss': '0.7417', 'grad_norm': '1.472', 'learning_rate': '9.998e-05', 'epoch': '0.5325'}
+{'loss': '0.7583', 'grad_norm': '1.243', 'learning_rate': '9.998e-05', 'epoch': '0.5333'}
+{'loss': '0.7639', 'grad_norm': '1.25', 'learning_rate': '9.998e-05', 'epoch': '0.5342'}
+{'loss': '0.8304', 'grad_norm': '1.358', 'learning_rate': '9.998e-05', 'epoch': '0.5351'}
+{'loss': '0.7332', 'grad_norm': '1.099', 'learning_rate': '9.998e-05', 'epoch': '0.5359'}
+{'loss': '0.8173', 'grad_norm': '1.461', 'learning_rate': '9.997e-05', 'epoch': '0.5368'}
+{'loss': '0.7677', 'grad_norm': '1.29', 'learning_rate': '9.997e-05', 'epoch': '0.5376'}
+{'loss': '0.8524', 'grad_norm': '1.458', 'learning_rate': '9.997e-05', 'epoch': '0.5385'}
+{'loss': '0.7921', 'grad_norm': '1.203', 'learning_rate': '9.997e-05', 'epoch': '0.5394'}
+{'loss': '0.789', 'grad_norm': '1.109', 'learning_rate': '9.997e-05', 'epoch': '0.5402'}
+{'loss': '0.6594', 'grad_norm': '1.193', 'learning_rate': '9.997e-05', 'epoch': '0.5411'}
+{'loss': '0.7768', 'grad_norm': '1.367', 'learning_rate': '9.997e-05', 'epoch': '0.542'}
+{'loss': '0.7392', 'grad_norm': '1.29', 'learning_rate': '9.997e-05', 'epoch': '0.5428'}
+{'loss': '0.6961', 'grad_norm': '1.367', 'learning_rate': '9.997e-05', 'epoch': '0.5437'}
+{'loss': '0.8015', 'grad_norm': '1.293', 'learning_rate': '9.997e-05', 'epoch': '0.5446'}
+{'loss': '0.7287', 'grad_norm': '1.197', 'learning_rate': '9.997e-05', 'epoch': '0.5454'}
+{'loss': '0.7791', 'grad_norm': '1.42', 'learning_rate': '9.997e-05', 'epoch': '0.5463'}
+{'loss': '0.6665', 'grad_norm': '1.354', 'learning_rate': '9.996e-05', 'epoch': '0.5471'}
+{'loss': '0.7903', 'grad_norm': '1.627', 'learning_rate': '9.996e-05', 'epoch': '0.548'}
+{'loss': '0.8484', 'grad_norm': '1.426', 'learning_rate': '9.996e-05', 'epoch': '0.5489'}
+{'loss': '0.7432', 'grad_norm': '1.318', 'learning_rate': '9.996e-05', 'epoch': '0.5497'}
+{'loss': '0.6722', 'grad_norm': '0.9959', 'learning_rate': '9.996e-05', 'epoch': '0.5506'}
+{'loss': '0.777', 'grad_norm': '1.372', 'learning_rate': '9.996e-05', 'epoch': '0.5515'}
+{'loss': '0.8165', 'grad_norm': '1.191', 'learning_rate': '9.996e-05', 'epoch': '0.5523'}
+{'loss': '0.7122', 'grad_norm': '1.262', 'learning_rate': '9.996e-05', 'epoch': '0.5532'}
+{'loss': '0.8476', 'grad_norm': '1.455', 'learning_rate': '9.996e-05', 'epoch': '0.554'}
+{'loss': '0.7595', 'grad_norm': '1.443', 'learning_rate': '9.996e-05', 'epoch': '0.5549'}
+{'loss': '0.7413', 'grad_norm': '1.151', 'learning_rate': '9.995e-05', 'epoch': '0.5558'}
+{'loss': '0.7243', 'grad_norm': '1.316', 'learning_rate': '9.995e-05', 'epoch': '0.5566'}
+{'loss': '0.778', 'grad_norm': '1.275', 'learning_rate': '9.995e-05', 'epoch': '0.5575'}
+{'loss': '0.8076', 'grad_norm': '1.364', 'learning_rate': '9.995e-05', 'epoch': '0.5584'}
+{'loss': '0.6941', 'grad_norm': '1.025', 'learning_rate': '9.995e-05', 'epoch': '0.5592'}
+{'loss': '0.7991', 'grad_norm': '1.338', 'learning_rate': '9.995e-05', 'epoch': '0.5601'}
+{'loss': '0.8039', 'grad_norm': '1.377', 'learning_rate': '9.995e-05', 'epoch': '0.5609'}
+{'eval_loss': '0.7963', 'eval_runtime': '827.9', 'eval_samples_per_second': '2.488', 'eval_steps_per_second': '2.488', 'epoch': '0.5609'}
+{'loss': '0.8553', 'grad_norm': '1.542', 'learning_rate': '9.995e-05', 'epoch': '0.5618'}
+{'loss': '0.665', 'grad_norm': '1.105', 'learning_rate': '9.995e-05', 'epoch': '0.5627'}
+{'loss': '0.7945', 'grad_norm': '1.453', 'learning_rate': '9.994e-05', 'epoch': '0.5635'}
+{'loss': '0.7241', 'grad_norm': '1.259', 'learning_rate': '9.994e-05', 'epoch': '0.5644'}
+{'loss': '0.7781', 'grad_norm': '1.498', 'learning_rate': '9.994e-05', 'epoch': '0.5653'}
+{'loss': '0.7772', 'grad_norm': '1.426', 'learning_rate': '9.994e-05', 'epoch': '0.5661'}
+{'loss': '0.7444', 'grad_norm': '1.199', 'learning_rate': '9.994e-05', 'epoch': '0.567'}
+{'loss': '0.7203', 'grad_norm': '1.149', 'learning_rate': '9.994e-05', 'epoch': '0.5679'}
+{'loss': '0.8039', 'grad_norm': '1.359', 'learning_rate': '9.994e-05', 'epoch': '0.5687'}
+{'loss': '0.8273', 'grad_norm': '1.416', 'learning_rate': '9.994e-05', 'epoch': '0.5696'}
+{'loss': '0.8183', 'grad_norm': '1.273', 'learning_rate': '9.993e-05', 'epoch': '0.5704'}
+{'loss': '0.8268', 'grad_norm': '1.379', 'learning_rate': '9.993e-05', 'epoch': '0.5713'}
+{'loss': '0.7435', 'grad_norm': '1.264', 'learning_rate': '9.993e-05', 'epoch': '0.5722'}
+{'loss': '0.7637', 'grad_norm': '1.243', 'learning_rate': '9.993e-05', 'epoch': '0.573'}
+{'loss': '0.7946', 'grad_norm': '1.328', 'learning_rate': '9.993e-05', 'epoch': '0.5739'}
+{'loss': '0.7659', 'grad_norm': '1.24', 'learning_rate': '9.993e-05', 'epoch': '0.5748'}
+{'loss': '0.8483', 'grad_norm': '1.414', 'learning_rate': '9.993e-05', 'epoch': '0.5756'}
+{'loss': '0.861', 'grad_norm': '1.239', 'learning_rate': '9.993e-05', 'epoch': '0.5765'}
+{'loss': '0.8021', 'grad_norm': '1.331', 'learning_rate': '9.992e-05', 'epoch': '0.5773'}
+{'loss': '0.7536', 'grad_norm': '1.293', 'learning_rate': '9.992e-05', 'epoch': '0.5782'}
+{'loss': '0.7519', 'grad_norm': '1.151', 'learning_rate': '9.992e-05', 'epoch': '0.5791'}
+{'loss': '0.7844', 'grad_norm': '1.591', 'learning_rate': '9.992e-05', 'epoch': '0.5799'}
+{'loss': '0.8218', 'grad_norm': '1.142', 'learning_rate': '9.992e-05', 'epoch': '0.5808'}
+{'loss': '0.7447', 'grad_norm': '1.077', 'learning_rate': '9.992e-05', 'epoch': '0.5817'}
+{'loss': '0.8426', 'grad_norm': '1.148', 'learning_rate': '9.992e-05', 'epoch': '0.5825'}
+{'loss': '0.8186', 'grad_norm': '1.662', 'learning_rate': '9.991e-05', 'epoch': '0.5834'}
+{'loss': '0.7522', 'grad_norm': '1.415', 'learning_rate': '9.991e-05', 'epoch': '0.5843'}
+{'loss': '0.7517', 'grad_norm': '1.298', 'learning_rate': '9.991e-05', 'epoch': '0.5851'}
+{'loss': '0.7493', 'grad_norm': '1.324', 'learning_rate': '9.991e-05', 'epoch': '0.586'}
+{'loss': '0.6884', 'grad_norm': '1.195', 'learning_rate': '9.991e-05', 'epoch': '0.5868'}
+{'loss': '0.786', 'grad_norm': '1.393', 'learning_rate': '9.991e-05', 'epoch': '0.5877'}
+{'loss': '0.7313', 'grad_norm': '1.221', 'learning_rate': '9.991e-05', 'epoch': '0.5886'}
+{'loss': '0.8008', 'grad_norm': '1.537', 'learning_rate': '9.99e-05', 'epoch': '0.5894'}
+{'loss': '0.773', 'grad_norm': '1.183', 'learning_rate': '9.99e-05', 'epoch': '0.5903'}
+{'loss': '0.7469', 'grad_norm': '1.118', 'learning_rate': '9.99e-05', 'epoch': '0.5912'}
+{'loss': '0.7789', 'grad_norm': '1.374', 'learning_rate': '9.99e-05', 'epoch': '0.592'}
+{'loss': '0.8118', 'grad_norm': '1.191', 'learning_rate': '9.99e-05', 'epoch': '0.5929'}
+{'loss': '0.7537', 'grad_norm': '1.327', 'learning_rate': '9.99e-05', 'epoch': '0.5937'}
+{'loss': '0.8603', 'grad_norm': '1.451', 'learning_rate': '9.989e-05', 'epoch': '0.5946'}
+{'loss': '0.7495', 'grad_norm': '1.114', 'learning_rate': '9.989e-05', 'epoch': '0.5955'}
+{'loss': '0.7295', 'grad_norm': '1.286', 'learning_rate': '9.989e-05', 'epoch': '0.5963'}
+{'loss': '0.7798', 'grad_norm': '1.292', 'learning_rate': '9.989e-05', 'epoch': '0.5972'}
+{'loss': '0.7824', 'grad_norm': '1.125', 'learning_rate': '9.989e-05', 'epoch': '0.5981'}
+{'loss': '0.7148', 'grad_norm': '1.238', 'learning_rate': '9.989e-05', 'epoch': '0.5989'}
+{'loss': '0.7073', 'grad_norm': '1.195', 'learning_rate': '9.989e-05', 'epoch': '0.5998'}
+{'loss': '0.7076', 'grad_norm': '1.341', 'learning_rate': '9.988e-05', 'epoch': '0.6006'}
+{'loss': '0.8237', 'grad_norm': '1.458', 'learning_rate': '9.988e-05', 'epoch': '0.6015'}
+{'loss': '0.7992', 'grad_norm': '1.384', 'learning_rate': '9.988e-05', 'epoch': '0.6024'}
+{'loss': '0.6936', 'grad_norm': '1.218', 'learning_rate': '9.988e-05', 'epoch': '0.6032'}
+{'loss': '0.7835', 'grad_norm': '1.171', 'learning_rate': '9.988e-05', 'epoch': '0.6041'}
+{'eval_loss': '0.7867', 'eval_runtime': '836.1', 'eval_samples_per_second': '2.464', 'eval_steps_per_second': '2.464', 'epoch': '0.6041'}
+{'loss': '0.7651', 'grad_norm': '1.593', 'learning_rate': '9.988e-05', 'epoch': '0.605'}
+{'loss': '0.7521', 'grad_norm': '1.231', 'learning_rate': '9.987e-05', 'epoch': '0.6058'}
+{'loss': '0.7475', 'grad_norm': '1.315', 'learning_rate': '9.987e-05', 'epoch': '0.6067'}
+{'loss': '0.777', 'grad_norm': '1.264', 'learning_rate': '9.987e-05', 'epoch': '0.6076'}
+{'loss': '0.6682', 'grad_norm': '1.085', 'learning_rate': '9.987e-05', 'epoch': '0.6084'}
+{'loss': '0.7766', 'grad_norm': '1.425', 'learning_rate': '9.987e-05', 'epoch': '0.6093'}
+{'loss': '0.743', 'grad_norm': '1.299', 'learning_rate': '9.986e-05', 'epoch': '0.6101'}
+{'loss': '0.7838', 'grad_norm': '2.615', 'learning_rate': '9.986e-05', 'epoch': '0.611'}
+{'loss': '0.7679', 'grad_norm': '1.431', 'learning_rate': '9.986e-05', 'epoch': '0.6119'}
+{'loss': '0.689', 'grad_norm': '1.141', 'learning_rate': '9.986e-05', 'epoch': '0.6127'}
+{'loss': '0.7659', 'grad_norm': '1.372', 'learning_rate': '9.986e-05', 'epoch': '0.6136'}
+{'loss': '0.7405', 'grad_norm': '1.263', 'learning_rate': '9.986e-05', 'epoch': '0.6145'}
+{'loss': '0.8082', 'grad_norm': '1.565', 'learning_rate': '9.985e-05', 'epoch': '0.6153'}
+{'loss': '0.7774', 'grad_norm': '1.227', 'learning_rate': '9.985e-05', 'epoch': '0.6162'}
+{'loss': '0.7413', 'grad_norm': '1.219', 'learning_rate': '9.985e-05', 'epoch': '0.617'}
+{'loss': '0.7258', 'grad_norm': '1.482', 'learning_rate': '9.985e-05', 'epoch': '0.6179'}
+{'loss': '0.7151', 'grad_norm': '1.03', 'learning_rate': '9.985e-05', 'epoch': '0.6188'}
+{'loss': '0.7856', 'grad_norm': '1.24', 'learning_rate': '9.984e-05', 'epoch': '0.6196'}
+{'loss': '0.636', 'grad_norm': '1.295', 'learning_rate': '9.984e-05', 'epoch': '0.6205'}
+{'loss': '0.728', 'grad_norm': '1.541', 'learning_rate': '9.984e-05', 'epoch': '0.6214'}
+{'loss': '0.7553', 'grad_norm': '1.304', 'learning_rate': '9.984e-05', 'epoch': '0.6222'}
+{'loss': '0.7329', 'grad_norm': '1.145', 'learning_rate': '9.984e-05', 'epoch': '0.6231'}
+{'loss': '0.7532', 'grad_norm': '1.265', 'learning_rate': '9.983e-05', 'epoch': '0.6239'}
+{'loss': '0.7379', 'grad_norm': '1.21', 'learning_rate': '9.983e-05', 'epoch': '0.6248'}
+{'loss': '0.791', 'grad_norm': '1.286', 'learning_rate': '9.983e-05', 'epoch': '0.6257'}
+{'loss': '0.7853', 'grad_norm': '1.39', 'learning_rate': '9.983e-05', 'epoch': '0.6265'}
+{'loss': '0.7542', 'grad_norm': '1.24', 'learning_rate': '9.983e-05', 'epoch': '0.6274'}
+{'loss': '0.8082', 'grad_norm': '1.29', 'learning_rate': '9.982e-05', 'epoch': '0.6283'}
+{'loss': '0.7887', 'grad_norm': '1.224', 'learning_rate': '9.982e-05', 'epoch': '0.6291'}
+{'loss': '0.8441', 'grad_norm': '1.304', 'learning_rate': '9.982e-05', 'epoch': '0.63'}
+{'loss': '0.7784', 'grad_norm': '1.642', 'learning_rate': '9.982e-05', 'epoch': '0.6309'}
+{'loss': '0.747', 'grad_norm': '1.244', 'learning_rate': '9.982e-05', 'epoch': '0.6317'}
+{'loss': '0.668', 'grad_norm': '1.032', 'learning_rate': '9.981e-05', 'epoch': '0.6326'}
+{'loss': '0.7103', 'grad_norm': '1.224', 'learning_rate': '9.981e-05', 'epoch': '0.6334'}
+{'loss': '0.7427', 'grad_norm': '1.265', 'learning_rate': '9.981e-05', 'epoch': '0.6343'}
+{'loss': '0.7665', 'grad_norm': '1.192', 'learning_rate': '9.981e-05', 'epoch': '0.6352'}
+{'loss': '0.7348', 'grad_norm': '1.05', 'learning_rate': '9.98e-05', 'epoch': '0.636'}
+{'loss': '0.6877', 'grad_norm': '1.222', 'learning_rate': '9.98e-05', 'epoch': '0.6369'}
+{'loss': '0.7903', 'grad_norm': '1.147', 'learning_rate': '9.98e-05', 'epoch': '0.6378'}
+{'loss': '0.7921', 'grad_norm': '1.258', 'learning_rate': '9.98e-05', 'epoch': '0.6386'}
+{'loss': '0.703', 'grad_norm': '1.262', 'learning_rate': '9.98e-05', 'epoch': '0.6395'}
+{'loss': '0.7489', 'grad_norm': '1.397', 'learning_rate': '9.979e-05', 'epoch': '0.6403'}
+{'loss': '0.6733', 'grad_norm': '1.006', 'learning_rate': '9.979e-05', 'epoch': '0.6412'}
+{'loss': '0.8222', 'grad_norm': '1.499', 'learning_rate': '9.979e-05', 'epoch': '0.6421'}
+{'loss': '0.8829', 'grad_norm': '1.284', 'learning_rate': '9.979e-05', 'epoch': '0.6429'}
+{'loss': '0.7606', 'grad_norm': '1.176', 'learning_rate': '9.978e-05', 'epoch': '0.6438'}
+{'loss': '0.7618', 'grad_norm': '1.199', 'learning_rate': '9.978e-05', 'epoch': '0.6447'}
+{'loss': '0.7071', 'grad_norm': '1.168', 'learning_rate': '9.978e-05', 'epoch': '0.6455'}
+{'loss': '0.756', 'grad_norm': '1.263', 'learning_rate': '9.978e-05', 'epoch': '0.6464'}
+{'loss': '0.7254', 'grad_norm': '1.211', 'learning_rate': '9.978e-05', 'epoch': '0.6472'}
+{'eval_loss': '0.7785', 'eval_runtime': '836.2', 'eval_samples_per_second': '2.463', 'eval_steps_per_second': '2.463', 'epoch': '0.6472'}
+{'loss': '0.8213', 'grad_norm': '1.479', 'learning_rate': '9.977e-05', 'epoch': '0.6481'}
+{'loss': '0.722', 'grad_norm': '1.212', 'learning_rate': '9.977e-05', 'epoch': '0.649'}
+{'loss': '0.7791', 'grad_norm': '1.313', 'learning_rate': '9.977e-05', 'epoch': '0.6498'}
+{'loss': '0.7052', 'grad_norm': '1.146', 'learning_rate': '9.977e-05', 'epoch': '0.6507'}
+{'loss': '0.7135', 'grad_norm': '1.19', 'learning_rate': '9.976e-05', 'epoch': '0.6516'}
+{'loss': '0.8115', 'grad_norm': '1.411', 'learning_rate': '9.976e-05', 'epoch': '0.6524'}
+{'loss': '0.7502', 'grad_norm': '1.246', 'learning_rate': '9.976e-05', 'epoch': '0.6533'}
+{'loss': '0.7697', 'grad_norm': '1.292', 'learning_rate': '9.976e-05', 'epoch': '0.6542'}
+{'loss': '0.7457', 'grad_norm': '1.27', 'learning_rate': '9.975e-05', 'epoch': '0.655'}
+{'loss': '0.8368', 'grad_norm': '1.397', 'learning_rate': '9.975e-05', 'epoch': '0.6559'}
+{'loss': '0.7356', 'grad_norm': '1.216', 'learning_rate': '9.975e-05', 'epoch': '0.6567'}
+{'loss': '0.8023', 'grad_norm': '1.13', 'learning_rate': '9.975e-05', 'epoch': '0.6576'}
+{'loss': '0.7965', 'grad_norm': '1.111', 'learning_rate': '9.974e-05', 'epoch': '0.6585'}
+{'loss': '0.7177', 'grad_norm': '1.574', 'learning_rate': '9.974e-05', 'epoch': '0.6593'}
+{'loss': '0.7114', 'grad_norm': '1.144', 'learning_rate': '9.974e-05', 'epoch': '0.6602'}
+{'loss': '0.7547', 'grad_norm': '1.274', 'learning_rate': '9.974e-05', 'epoch': '0.6611'}
+{'loss': '0.6721', 'grad_norm': '3.987', 'learning_rate': '9.973e-05', 'epoch': '0.6619'}
+{'loss': '0.793', 'grad_norm': '1.251', 'learning_rate': '9.973e-05', 'epoch': '0.6628'}
+{'loss': '0.7806', 'grad_norm': '1.433', 'learning_rate': '9.973e-05', 'epoch': '0.6636'}
+{'loss': '0.7176', 'grad_norm': '1.479', 'learning_rate': '9.973e-05', 'epoch': '0.6645'}
+{'loss': '0.7146', 'grad_norm': '1.24', 'learning_rate': '9.972e-05', 'epoch': '0.6654'}
+{'loss': '0.8026', 'grad_norm': '1.131', 'learning_rate': '9.972e-05', 'epoch': '0.6662'}
+{'loss': '0.7809', 'grad_norm': '1.248', 'learning_rate': '9.972e-05', 'epoch': '0.6671'}
+{'loss': '0.7097', 'grad_norm': '1.254', 'learning_rate': '9.972e-05', 'epoch': '0.668'}
+{'loss': '0.6529', 'grad_norm': '1.143', 'learning_rate': '9.971e-05', 'epoch': '0.6688'}
+{'loss': '0.778', 'grad_norm': '1.597', 'learning_rate': '9.971e-05', 'epoch': '0.6697'}
+{'loss': '0.6889', 'grad_norm': '1.186', 'learning_rate': '9.971e-05', 'epoch': '0.6706'}
+{'loss': '0.7155', 'grad_norm': '1.192', 'learning_rate': '9.971e-05', 'epoch': '0.6714'}
+{'loss': '0.7419', 'grad_norm': '1.301', 'learning_rate': '9.97e-05', 'epoch': '0.6723'}
+{'loss': '0.7857', 'grad_norm': '1.154', 'learning_rate': '9.97e-05', 'epoch': '0.6731'}
+{'loss': '0.6834', 'grad_norm': '1.306', 'learning_rate': '9.97e-05', 'epoch': '0.674'}
+{'loss': '0.8491', 'grad_norm': '1.29', 'learning_rate': '9.97e-05', 'epoch': '0.6749'}
+{'loss': '0.7463', 'grad_norm': '1.218', 'learning_rate': '9.969e-05', 'epoch': '0.6757'}
+{'loss': '0.7145', 'grad_norm': '1.123', 'learning_rate': '9.969e-05', 'epoch': '0.6766'}
+{'loss': '0.7912', 'grad_norm': '1.178', 'learning_rate': '9.969e-05', 'epoch': '0.6775'}
+{'loss': '0.6947', 'grad_norm': '1.354', 'learning_rate': '9.968e-05', 'epoch': '0.6783'}
+{'loss': '0.7597', 'grad_norm': '1.307', 'learning_rate': '9.968e-05', 'epoch': '0.6792'}
+{'loss': '0.6763', 'grad_norm': '1.321', 'learning_rate': '9.968e-05', 'epoch': '0.68'}
+{'loss': '0.7561', 'grad_norm': '1.173', 'learning_rate': '9.968e-05', 'epoch': '0.6809'}
+{'loss': '0.805', 'grad_norm': '1.446', 'learning_rate': '9.967e-05', 'epoch': '0.6818'}
+{'loss': '0.6628', 'grad_norm': '1.119', 'learning_rate': '9.967e-05', 'epoch': '0.6826'}
+{'loss': '0.7684', 'grad_norm': '1.383', 'learning_rate': '9.967e-05', 'epoch': '0.6835'}
+{'loss': '0.7084', 'grad_norm': '1.242', 'learning_rate': '9.966e-05', 'epoch': '0.6844'}
+{'loss': '0.6817', 'grad_norm': '1.214', 'learning_rate': '9.966e-05', 'epoch': '0.6852'}
+{'loss': '0.8422', 'grad_norm': '1.4', 'learning_rate': '9.966e-05', 'epoch': '0.6861'}
+{'loss': '0.7433', 'grad_norm': '1.435', 'learning_rate': '9.966e-05', 'epoch': '0.6869'}
+{'loss': '0.7571', 'grad_norm': '1.192', 'learning_rate': '9.965e-05', 'epoch': '0.6878'}
+{'loss': '0.8024', 'grad_norm': '1.283', 'learning_rate': '9.965e-05', 'epoch': '0.6887'}
+{'loss': '0.8113', 'grad_norm': '1.165', 'learning_rate': '9.965e-05', 'epoch': '0.6895'}
+{'loss': '0.7507', 'grad_norm': '1.171', 'learning_rate': '9.964e-05', 'epoch': '0.6904'}
+{'eval_loss': '0.7706', 'eval_runtime': '821.5', 'eval_samples_per_second': '2.508', 'eval_steps_per_second': '2.508', 'epoch': '0.6904'}
+{'loss': '0.7423', 'grad_norm': '1.175', 'learning_rate': '9.964e-05', 'epoch': '0.6913'}
+{'loss': '0.7126', 'grad_norm': '1.642', 'learning_rate': '9.964e-05', 'epoch': '0.6921'}
+{'loss': '0.7579', 'grad_norm': '1.326', 'learning_rate': '9.964e-05', 'epoch': '0.693'}
+{'loss': '0.6849', 'grad_norm': '1.2', 'learning_rate': '9.963e-05', 'epoch': '0.6939'}
+{'loss': '0.7636', 'grad_norm': '1.039', 'learning_rate': '9.963e-05', 'epoch': '0.6947'}
+{'loss': '0.7209', 'grad_norm': '1.111', 'learning_rate': '9.963e-05', 'epoch': '0.6956'}
+{'loss': '0.6718', 'grad_norm': '1.157', 'learning_rate': '9.962e-05', 'epoch': '0.6964'}
+{'loss': '0.7649', 'grad_norm': '1.216', 'learning_rate': '9.962e-05', 'epoch': '0.6973'}
+{'loss': '0.8491', 'grad_norm': '1.48', 'learning_rate': '9.962e-05', 'epoch': '0.6982'}
+{'loss': '0.5951', 'grad_norm': '1.034', 'learning_rate': '9.961e-05', 'epoch': '0.699'}
+{'loss': '0.7288', 'grad_norm': '1.111', 'learning_rate': '9.961e-05', 'epoch': '0.6999'}
+{'loss': '0.7075', 'grad_norm': '1.097', 'learning_rate': '9.961e-05', 'epoch': '0.7008'}
+{'loss': '0.7349', 'grad_norm': '1.311', 'learning_rate': '9.961e-05', 'epoch': '0.7016'}
+{'loss': '0.7658', 'grad_norm': '1.34', 'learning_rate': '9.96e-05', 'epoch': '0.7025'}
+{'loss': '0.81', 'grad_norm': '1.371', 'learning_rate': '9.96e-05', 'epoch': '0.7033'}
+{'loss': '0.6835', 'grad_norm': '1.228', 'learning_rate': '9.96e-05', 'epoch': '0.7042'}
+{'loss': '0.7416', 'grad_norm': '1.051', 'learning_rate': '9.959e-05', 'epoch': '0.7051'}
+{'loss': '0.7304', 'grad_norm': '1.131', 'learning_rate': '9.959e-05', 'epoch': '0.7059'}
+{'loss': '0.823', 'grad_norm': '1.333', 'learning_rate': '9.959e-05', 'epoch': '0.7068'}
+{'loss': '0.8413', 'grad_norm': '1.384', 'learning_rate': '9.958e-05', 'epoch': '0.7077'}
+{'loss': '0.7743', 'grad_norm': '1.314', 'learning_rate': '9.958e-05', 'epoch': '0.7085'}
+{'loss': '0.7564', 'grad_norm': '1.473', 'learning_rate': '9.958e-05', 'epoch': '0.7094'}
+{'loss': '0.7086', 'grad_norm': '1.189', 'learning_rate': '9.957e-05', 'epoch': '0.7102'}
+{'loss': '0.6838', 'grad_norm': '1.124', 'learning_rate': '9.957e-05', 'epoch': '0.7111'}
+{'loss': '0.7256', 'grad_norm': '1.222', 'learning_rate': '9.957e-05', 'epoch': '0.712'}
+{'loss': '0.785', 'grad_norm': '1.36', 'learning_rate': '9.956e-05', 'epoch': '0.7128'}
+{'loss': '0.767', 'grad_norm': '1.282', 'learning_rate': '9.956e-05', 'epoch': '0.7137'}
+{'loss': '0.7053', 'grad_norm': '1.242', 'learning_rate': '9.956e-05', 'epoch': '0.7146'}
+{'loss': '0.83', 'grad_norm': '1.398', 'learning_rate': '9.955e-05', 'epoch': '0.7154'}
+{'loss': '0.7743', 'grad_norm': '1.159', 'learning_rate': '9.955e-05', 'epoch': '0.7163'}
+{'loss': '0.8458', 'grad_norm': '1.25', 'learning_rate': '9.955e-05', 'epoch': '0.7172'}
+{'loss': '0.7045', 'grad_norm': '1.295', 'learning_rate': '9.954e-05', 'epoch': '0.718'}
+{'loss': '0.7917', 'grad_norm': '1.167', 'learning_rate': '9.954e-05', 'epoch': '0.7189'}
+{'loss': '0.7175', 'grad_norm': '1.147', 'learning_rate': '9.954e-05', 'epoch': '0.7197'}
+{'loss': '0.8926', 'grad_norm': '1.367', 'learning_rate': '9.953e-05', 'epoch': '0.7206'}
+{'loss': '0.7095', 'grad_norm': '1.049', 'learning_rate': '9.953e-05', 'epoch': '0.7215'}
+{'loss': '0.6732', 'grad_norm': '0.9995', 'learning_rate': '9.953e-05', 'epoch': '0.7223'}
+{'loss': '0.7316', 'grad_norm': '1.101', 'learning_rate': '9.952e-05', 'epoch': '0.7232'}
+{'loss': '0.738', 'grad_norm': '1.259', 'learning_rate': '9.952e-05', 'epoch': '0.7241'}
+{'loss': '0.7072', 'grad_norm': '1.312', 'learning_rate': '9.952e-05', 'epoch': '0.7249'}
+{'loss': '0.7457', 'grad_norm': '1.212', 'learning_rate': '9.951e-05', 'epoch': '0.7258'}
+{'loss': '0.7969', 'grad_norm': '1.391', 'learning_rate': '9.951e-05', 'epoch': '0.7266'}
+{'loss': '0.7009', 'grad_norm': '1.323', 'learning_rate': '9.951e-05', 'epoch': '0.7275'}
+{'loss': '0.7973', 'grad_norm': '1.466', 'learning_rate': '9.95e-05', 'epoch': '0.7284'}
+{'loss': '0.7811', 'grad_norm': '1.17', 'learning_rate': '9.95e-05', 'epoch': '0.7292'}
+{'loss': '0.753', 'grad_norm': '1.202', 'learning_rate': '9.95e-05', 'epoch': '0.7301'}
+{'loss': '0.7621', 'grad_norm': '1.347', 'learning_rate': '9.949e-05', 'epoch': '0.731'}
+{'loss': '0.748', 'grad_norm': '1.219', 'learning_rate': '9.949e-05', 'epoch': '0.7318'}
+{'loss': '0.7328', 'grad_norm': '1.141', 'learning_rate': '9.949e-05', 'epoch': '0.7327'}
+{'loss': '0.7615', 'grad_norm': '1.31', 'learning_rate': '9.948e-05', 'epoch': '0.7335'}
+{'eval_loss': '0.7649', 'eval_runtime': '828.4', 'eval_samples_per_second': '2.487', 'eval_steps_per_second': '2.487', 'epoch': '0.7335'}
+{'loss': '0.7564', 'grad_norm': '1.303', 'learning_rate': '9.948e-05', 'epoch': '0.7344'}
+{'loss': '0.769', 'grad_norm': '1.267', 'learning_rate': '9.948e-05', 'epoch': '0.7353'}
+{'loss': '0.7924', 'grad_norm': '1.116', 'learning_rate': '9.947e-05', 'epoch': '0.7361'}
+{'loss': '0.6541', 'grad_norm': '1.119', 'learning_rate': '9.947e-05', 'epoch': '0.737'}
+{'loss': '0.7267', 'grad_norm': '1.374', 'learning_rate': '9.947e-05', 'epoch': '0.7379'}
+{'loss': '0.7693', 'grad_norm': '1.235', 'learning_rate': '9.946e-05', 'epoch': '0.7387'}
+{'loss': '0.7751', 'grad_norm': '1.163', 'learning_rate': '9.946e-05', 'epoch': '0.7396'}
+{'loss': '0.7799', 'grad_norm': '1.227', 'learning_rate': '9.945e-05', 'epoch': '0.7405'}
+{'loss': '0.6129', 'grad_norm': '0.9969', 'learning_rate': '9.945e-05', 'epoch': '0.7413'}
+{'loss': '0.7537', 'grad_norm': '1.214', 'learning_rate': '9.945e-05', 'epoch': '0.7422'}
+{'loss': '0.7034', 'grad_norm': '1.136', 'learning_rate': '9.944e-05', 'epoch': '0.743'}
+{'loss': '0.7285', 'grad_norm': '1.073', 'learning_rate': '9.944e-05', 'epoch': '0.7439'}
+{'loss': '0.7865', 'grad_norm': '1.361', 'learning_rate': '9.944e-05', 'epoch': '0.7448'}
+{'loss': '0.7511', 'grad_norm': '1.099', 'learning_rate': '9.943e-05', 'epoch': '0.7456'}
+{'loss': '0.7325', 'grad_norm': '1.151', 'learning_rate': '9.943e-05', 'epoch': '0.7465'}
+{'loss': '0.7024', 'grad_norm': '1.1', 'learning_rate': '9.943e-05', 'epoch': '0.7474'}
+{'loss': '0.7516', 'grad_norm': '1.008', 'learning_rate': '9.942e-05', 'epoch': '0.7482'}
+{'loss': '0.691', 'grad_norm': '1.336', 'learning_rate': '9.942e-05', 'epoch': '0.7491'}
+{'loss': '0.7186', 'grad_norm': '1.017', 'learning_rate': '9.941e-05', 'epoch': '0.7499'}
+{'loss': '0.7137', 'grad_norm': '1.272', 'learning_rate': '9.941e-05', 'epoch': '0.7508'}
+{'loss': '0.7596', 'grad_norm': '1.078', 'learning_rate': '9.941e-05', 'epoch': '0.7517'}
+{'loss': '0.7603', 'grad_norm': '1.344', 'learning_rate': '9.94e-05', 'epoch': '0.7525'}
+{'loss': '0.7955', 'grad_norm': '1.261', 'learning_rate': '9.94e-05', 'epoch': '0.7534'}
+{'loss': '0.7661', 'grad_norm': '1.279', 'learning_rate': '9.94e-05', 'epoch': '0.7543'}
+{'loss': '0.6855', 'grad_norm': '1.304', 'learning_rate': '9.939e-05', 'epoch': '0.7551'}
+{'loss': '0.7971', 'grad_norm': '1.28', 'learning_rate': '9.939e-05', 'epoch': '0.756'}
+{'loss': '0.7073', 'grad_norm': '1.169', 'learning_rate': '9.938e-05', 'epoch': '0.7569'}
+{'loss': '0.7135', 'grad_norm': '1.383', 'learning_rate': '9.938e-05', 'epoch': '0.7577'}
+{'loss': '0.7873', 'grad_norm': '1.223', 'learning_rate': '9.938e-05', 'epoch': '0.7586'}
+{'loss': '0.7681', 'grad_norm': '1.274', 'learning_rate': '9.937e-05', 'epoch': '0.7594'}
+{'loss': '0.7341', 'grad_norm': '1.256', 'learning_rate': '9.937e-05', 'epoch': '0.7603'}
+{'loss': '0.7972', 'grad_norm': '1.386', 'learning_rate': '9.936e-05', 'epoch': '0.7612'}
+{'loss': '0.7904', 'grad_norm': '1.271', 'learning_rate': '9.936e-05', 'epoch': '0.762'}
+{'loss': '0.7542', 'grad_norm': '1.211', 'learning_rate': '9.936e-05', 'epoch': '0.7629'}
+{'loss': '0.7975', 'grad_norm': '1.437', 'learning_rate': '9.935e-05', 'epoch': '0.7638'}
+{'loss': '0.7538', 'grad_norm': '1.099', 'learning_rate': '9.935e-05', 'epoch': '0.7646'}
+{'loss': '0.7334', 'grad_norm': '1.112', 'learning_rate': '9.934e-05', 'epoch': '0.7655'}
+{'loss': '0.7473', 'grad_norm': '1.288', 'learning_rate': '9.934e-05', 'epoch': '0.7663'}
+{'loss': '0.7645', 'grad_norm': '1.362', 'learning_rate': '9.934e-05', 'epoch': '0.7672'}
+{'loss': '0.7872', 'grad_norm': '1.124', 'learning_rate': '9.933e-05', 'epoch': '0.7681'}
+{'loss': '0.7343', 'grad_norm': '1.374', 'learning_rate': '9.933e-05', 'epoch': '0.7689'}
+{'loss': '0.7486', 'grad_norm': '1.212', 'learning_rate': '9.932e-05', 'epoch': '0.7698'}
+{'loss': '0.7333', 'grad_norm': '1.183', 'learning_rate': '9.932e-05', 'epoch': '0.7707'}
+{'loss': '0.7241', 'grad_norm': '1.284', 'learning_rate': '9.932e-05', 'epoch': '0.7715'}
+{'loss': '0.751', 'grad_norm': '1.231', 'learning_rate': '9.931e-05', 'epoch': '0.7724'}
+{'loss': '0.7108', 'grad_norm': '1.145', 'learning_rate': '9.931e-05', 'epoch': '0.7732'}
+{'loss': '0.7688', 'grad_norm': '1.186', 'learning_rate': '9.93e-05', 'epoch': '0.7741'}
+{'loss': '0.7827', 'grad_norm': '1.283', 'learning_rate': '9.93e-05', 'epoch': '0.775'}
+{'loss': '0.697', 'grad_norm': '1.165', 'learning_rate': '9.93e-05', 'epoch': '0.7758'}
+{'loss': '0.7239', 'grad_norm': '1.044', 'learning_rate': '9.929e-05', 'epoch': '0.7767'}
+{'eval_loss': '0.7599', 'eval_runtime': '839', 'eval_samples_per_second': '2.455', 'eval_steps_per_second': '2.455', 'epoch': '0.7767'}
+{'loss': '0.7724', 'grad_norm': '1.246', 'learning_rate': '9.929e-05', 'epoch': '0.7776'}
+{'loss': '0.7155', 'grad_norm': '1.245', 'learning_rate': '9.928e-05', 'epoch': '0.7784'}
+{'loss': '0.8326', 'grad_norm': '1.224', 'learning_rate': '9.928e-05', 'epoch': '0.7793'}
+{'loss': '0.7093', 'grad_norm': '1.268', 'learning_rate': '9.928e-05', 'epoch': '0.7802'}
+{'loss': '0.7602', 'grad_norm': '1.173', 'learning_rate': '9.927e-05', 'epoch': '0.781'}
+{'loss': '0.6889', 'grad_norm': '1.069', 'learning_rate': '9.927e-05', 'epoch': '0.7819'}
+{'loss': '0.7397', 'grad_norm': '1.11', 'learning_rate': '9.926e-05', 'epoch': '0.7827'}
+{'loss': '0.718', 'grad_norm': '1.361', 'learning_rate': '9.926e-05', 'epoch': '0.7836'}
+{'loss': '0.7039', 'grad_norm': '1.143', 'learning_rate': '9.925e-05', 'epoch': '0.7845'}
+{'loss': '0.7419', 'grad_norm': '1.181', 'learning_rate': '9.925e-05', 'epoch': '0.7853'}
+{'loss': '0.758', 'grad_norm': '1.195', 'learning_rate': '9.925e-05', 'epoch': '0.7862'}
+{'loss': '0.7479', 'grad_norm': '1.305', 'learning_rate': '9.924e-05', 'epoch': '0.7871'}
+{'loss': '0.7382', 'grad_norm': '1.396', 'learning_rate': '9.924e-05', 'epoch': '0.7879'}
+{'loss': '0.6356', 'grad_norm': '1.152', 'learning_rate': '9.923e-05', 'epoch': '0.7888'}
+{'loss': '0.7674', 'grad_norm': '1.254', 'learning_rate': '9.923e-05', 'epoch': '0.7896'}
+{'loss': '0.7084', 'grad_norm': '1.136', 'learning_rate': '9.923e-05', 'epoch': '0.7905'}
+{'loss': '0.6273', 'grad_norm': '1.174', 'learning_rate': '9.922e-05', 'epoch': '0.7914'}
+{'loss': '0.8082', 'grad_norm': '1.205', 'learning_rate': '9.922e-05', 'epoch': '0.7922'}
+{'loss': '0.6632', 'grad_norm': '1.204', 'learning_rate': '9.921e-05', 'epoch': '0.7931'}
+{'loss': '0.7202', 'grad_norm': '1.059', 'learning_rate': '9.921e-05', 'epoch': '0.794'}
+{'loss': '0.724', 'grad_norm': '1.036', 'learning_rate': '9.92e-05', 'epoch': '0.7948'}
+{'loss': '0.7432', 'grad_norm': '1.185', 'learning_rate': '9.92e-05', 'epoch': '0.7957'}
+{'loss': '0.7064', 'grad_norm': '1.294', 'learning_rate': '9.919e-05', 'epoch': '0.7965'}
+{'loss': '0.8059', 'grad_norm': '1.274', 'learning_rate': '9.919e-05', 'epoch': '0.7974'}
+{'loss': '0.8051', 'grad_norm': '1.28', 'learning_rate': '9.919e-05', 'epoch': '0.7983'}
+{'loss': '0.7013', 'grad_norm': '1.223', 'learning_rate': '9.918e-05', 'epoch': '0.7991'}
+{'loss': '0.7415', 'grad_norm': '1.338', 'learning_rate': '9.918e-05', 'epoch': '0.8'}
+{'loss': '0.7265', 'grad_norm': '1.221', 'learning_rate': '9.917e-05', 'epoch': '0.8009'}
+{'loss': '0.7084', 'grad_norm': '1.116', 'learning_rate': '9.917e-05', 'epoch': '0.8017'}
+{'loss': '0.8204', 'grad_norm': '1.171', 'learning_rate': '9.916e-05', 'epoch': '0.8026'}
+{'loss': '0.6754', 'grad_norm': '0.9943', 'learning_rate': '9.916e-05', 'epoch': '0.8035'}
+{'loss': '0.8132', 'grad_norm': '1.24', 'learning_rate': '9.915e-05', 'epoch': '0.8043'}
+{'loss': '0.7456', 'grad_norm': '1.076', 'learning_rate': '9.915e-05', 'epoch': '0.8052'}
+{'loss': '0.7565', 'grad_norm': '1.213', 'learning_rate': '9.915e-05', 'epoch': '0.806'}
+{'loss': '0.6652', 'grad_norm': '1.177', 'learning_rate': '9.914e-05', 'epoch': '0.8069'}
+{'loss': '0.7598', 'grad_norm': '1.061', 'learning_rate': '9.914e-05', 'epoch': '0.8078'}
+{'loss': '0.7454', 'grad_norm': '1.397', 'learning_rate': '9.913e-05', 'epoch': '0.8086'}
+{'loss': '0.7158', 'grad_norm': '1.25', 'learning_rate': '9.913e-05', 'epoch': '0.8095'}
+{'loss': '0.7094', 'grad_norm': '1.234', 'learning_rate': '9.912e-05', 'epoch': '0.8104'}
+{'loss': '0.6653', 'grad_norm': '1.454', 'learning_rate': '9.912e-05', 'epoch': '0.8112'}
+{'loss': '0.6765', 'grad_norm': '1.031', 'learning_rate': '9.911e-05', 'epoch': '0.8121'}
+{'loss': '0.7043', 'grad_norm': '1.273', 'learning_rate': '9.911e-05', 'epoch': '0.8129'}
+{'loss': '0.6961', 'grad_norm': '1.161', 'learning_rate': '9.91e-05', 'epoch': '0.8138'}
+{'loss': '0.8029', 'grad_norm': '1.205', 'learning_rate': '9.91e-05', 'epoch': '0.8147'}
+{'loss': '0.6831', 'grad_norm': '1.167', 'learning_rate': '9.91e-05', 'epoch': '0.8155'}
+{'loss': '0.7194', 'grad_norm': '1.248', 'learning_rate': '9.909e-05', 'epoch': '0.8164'}
+{'loss': '0.7957', 'grad_norm': '1.138', 'learning_rate': '9.909e-05', 'epoch': '0.8173'}
+{'loss': '0.7168', 'grad_norm': '1.192', 'learning_rate': '9.908e-05', 'epoch': '0.8181'}
+{'loss': '0.7519', 'grad_norm': '1.125', 'learning_rate': '9.908e-05', 'epoch': '0.819'}
+{'loss': '0.7496', 'grad_norm': '1.069', 'learning_rate': '9.907e-05', 'epoch': '0.8198'}
+{'eval_loss': '0.7534', 'eval_runtime': '830.3', 'eval_samples_per_second': '2.481', 'eval_steps_per_second': '2.481', 'epoch': '0.8198'}
+{'loss': '0.6397', 'grad_norm': '0.9264', 'learning_rate': '9.907e-05', 'epoch': '0.8207'}
+{'loss': '0.7556', 'grad_norm': '1.223', 'learning_rate': '9.906e-05', 'epoch': '0.8216'}
+{'loss': '0.7114', 'grad_norm': '1.337', 'learning_rate': '9.906e-05', 'epoch': '0.8224'}
+{'loss': '0.6823', 'grad_norm': '1.217', 'learning_rate': '9.905e-05', 'epoch': '0.8233'}
+{'loss': '0.7347', 'grad_norm': '1.246', 'learning_rate': '9.905e-05', 'epoch': '0.8242'}
+{'loss': '0.7982', 'grad_norm': '1.251', 'learning_rate': '9.904e-05', 'epoch': '0.825'}
+{'loss': '0.7598', 'grad_norm': '1.225', 'learning_rate': '9.904e-05', 'epoch': '0.8259'}
+{'loss': '0.7411', 'grad_norm': '1.175', 'learning_rate': '9.903e-05', 'epoch': '0.8268'}
+{'loss': '0.7491', 'grad_norm': '1.191', 'learning_rate': '9.903e-05', 'epoch': '0.8276'}
+{'loss': '0.7656', 'grad_norm': '1.234', 'learning_rate': '9.902e-05', 'epoch': '0.8285'}
+{'loss': '0.72', 'grad_norm': '1.076', 'learning_rate': '9.902e-05', 'epoch': '0.8293'}
+{'loss': '0.8326', 'grad_norm': '1.152', 'learning_rate': '9.901e-05', 'epoch': '0.8302'}
+{'loss': '0.7674', 'grad_norm': '1.208', 'learning_rate': '9.901e-05', 'epoch': '0.8311'}
+{'loss': '0.7415', 'grad_norm': '1.335', 'learning_rate': '9.9e-05', 'epoch': '0.8319'}
+{'loss': '0.6723', 'grad_norm': '1.051', 'learning_rate': '9.9e-05', 'epoch': '0.8328'}
+{'loss': '0.8078', 'grad_norm': '1.416', 'learning_rate': '9.899e-05', 'epoch': '0.8337'}
+{'loss': '0.8053', 'grad_norm': '1.179', 'learning_rate': '9.899e-05', 'epoch': '0.8345'}
+{'loss': '0.7586', 'grad_norm': '1.196', 'learning_rate': '9.899e-05', 'epoch': '0.8354'}
+{'loss': '0.6911', 'grad_norm': '1.218', 'learning_rate': '9.898e-05', 'epoch': '0.8362'}
+{'loss': '0.7723', 'grad_norm': '1.293', 'learning_rate': '9.898e-05', 'epoch': '0.8371'}
+{'loss': '0.8237', 'grad_norm': '1.286', 'learning_rate': '9.897e-05', 'epoch': '0.838'}
+{'loss': '0.7316', 'grad_norm': '1.204', 'learning_rate': '9.897e-05', 'epoch': '0.8388'}
+{'loss': '0.6707', 'grad_norm': '1.206', 'learning_rate': '9.896e-05', 'epoch': '0.8397'}
+{'loss': '0.6921', 'grad_norm': '1.191', 'learning_rate': '9.896e-05', 'epoch': '0.8406'}
+{'loss': '0.7135', 'grad_norm': '1.123', 'learning_rate': '9.895e-05', 'epoch': '0.8414'}
+{'loss': '0.6334', 'grad_norm': '1.068', 'learning_rate': '9.895e-05', 'epoch': '0.8423'}
+{'loss': '0.7744', 'grad_norm': '1.522', 'learning_rate': '9.894e-05', 'epoch': '0.8431'}
+{'loss': '0.7678', 'grad_norm': '1.093', 'learning_rate': '9.894e-05', 'epoch': '0.844'}
+{'loss': '0.7345', 'grad_norm': '1.223', 'learning_rate': '9.893e-05', 'epoch': '0.8449'}
+{'loss': '0.7603', 'grad_norm': '1.112', 'learning_rate': '9.893e-05', 'epoch': '0.8457'}
+{'loss': '0.7176', 'grad_norm': '1.047', 'learning_rate': '9.892e-05', 'epoch': '0.8466'}
+{'loss': '0.722', 'grad_norm': '1.854', 'learning_rate': '9.892e-05', 'epoch': '0.8475'}
+{'loss': '0.74', 'grad_norm': '1.228', 'learning_rate': '9.891e-05', 'epoch': '0.8483'}
+{'loss': '0.77', 'grad_norm': '1.148', 'learning_rate': '9.89e-05', 'epoch': '0.8492'}
+{'loss': '0.6666', 'grad_norm': '1.188', 'learning_rate': '9.89e-05', 'epoch': '0.8501'}
+{'loss': '0.7883', 'grad_norm': '1.408', 'learning_rate': '9.889e-05', 'epoch': '0.8509'}
+{'loss': '0.7442', 'grad_norm': '1.143', 'learning_rate': '9.889e-05', 'epoch': '0.8518'}
+{'loss': '0.6464', 'grad_norm': '1.266', 'learning_rate': '9.888e-05', 'epoch': '0.8526'}
+{'loss': '0.7242', 'grad_norm': '1.164', 'learning_rate': '9.888e-05', 'epoch': '0.8535'}
+{'loss': '0.7432', 'grad_norm': '1.255', 'learning_rate': '9.887e-05', 'epoch': '0.8544'}
+{'loss': '0.7662', 'grad_norm': '1.106', 'learning_rate': '9.887e-05', 'epoch': '0.8552'}
+{'loss': '0.7121', 'grad_norm': '1.102', 'learning_rate': '9.886e-05', 'epoch': '0.8561'}
+{'loss': '0.7892', 'grad_norm': '1.355', 'learning_rate': '9.886e-05', 'epoch': '0.857'}
+{'loss': '0.76', 'grad_norm': '1.192', 'learning_rate': '9.885e-05', 'epoch': '0.8578'}
+{'loss': '0.8316', 'grad_norm': '1.171', 'learning_rate': '9.885e-05', 'epoch': '0.8587'}
+{'loss': '0.714', 'grad_norm': '1.253', 'learning_rate': '9.884e-05', 'epoch': '0.8595'}
+{'loss': '0.6657', 'grad_norm': '1.106', 'learning_rate': '9.884e-05', 'epoch': '0.8604'}
+{'loss': '0.7164', 'grad_norm': '1.169', 'learning_rate': '9.883e-05', 'epoch': '0.8613'}
+{'loss': '0.6886', 'grad_norm': '1.197', 'learning_rate': '9.883e-05', 'epoch': '0.8621'}
+{'loss': '0.7266', 'grad_norm': '1.17', 'learning_rate': '9.882e-05', 'epoch': '0.863'}
+{'eval_loss': '0.7487', 'eval_runtime': '825.1', 'eval_samples_per_second': '2.497', 'eval_steps_per_second': '2.497', 'epoch': '0.863'}
+{'loss': '0.7372', 'grad_norm': '1.312', 'learning_rate': '9.882e-05', 'epoch': '0.8639'}
+{'loss': '0.7453', 'grad_norm': '1.099', 'learning_rate': '9.881e-05', 'epoch': '0.8647'}
+{'loss': '0.7562', 'grad_norm': '1.196', 'learning_rate': '9.881e-05', 'epoch': '0.8656'}
+{'loss': '0.7655', 'grad_norm': '1.166', 'learning_rate': '9.88e-05', 'epoch': '0.8665'}
+{'loss': '0.6808', 'grad_norm': '1.098', 'learning_rate': '9.879e-05', 'epoch': '0.8673'}
+{'loss': '0.6887', 'grad_norm': '1.038', 'learning_rate': '9.879e-05', 'epoch': '0.8682'}
+{'loss': '0.7186', 'grad_norm': '2.023', 'learning_rate': '9.878e-05', 'epoch': '0.869'}
+{'loss': '0.7425', 'grad_norm': '1.161', 'learning_rate': '9.878e-05', 'epoch': '0.8699'}
+{'loss': '0.7451', 'grad_norm': '1.142', 'learning_rate': '9.877e-05', 'epoch': '0.8708'}
+{'loss': '0.7399', 'grad_norm': '1.101', 'learning_rate': '9.877e-05', 'epoch': '0.8716'}
+{'loss': '0.7122', 'grad_norm': '1.166', 'learning_rate': '9.876e-05', 'epoch': '0.8725'}
+{'loss': '0.7026', 'grad_norm': '1.184', 'learning_rate': '9.876e-05', 'epoch': '0.8734'}
+{'loss': '0.7211', 'grad_norm': '1.16', 'learning_rate': '9.875e-05', 'epoch': '0.8742'}
+{'loss': '0.7565', 'grad_norm': '1.223', 'learning_rate': '9.875e-05', 'epoch': '0.8751'}
+{'loss': '0.7173', 'grad_norm': '1.204', 'learning_rate': '9.874e-05', 'epoch': '0.8759'}
+{'loss': '0.8208', 'grad_norm': '1.332', 'learning_rate': '9.874e-05', 'epoch': '0.8768'}
+{'loss': '0.8689', 'grad_norm': '1.307', 'learning_rate': '9.873e-05', 'epoch': '0.8777'}
+{'loss': '0.709', 'grad_norm': '1.356', 'learning_rate': '9.872e-05', 'epoch': '0.8785'}
+{'loss': '0.6883', 'grad_norm': '1.223', 'learning_rate': '9.872e-05', 'epoch': '0.8794'}
+{'loss': '0.7134', 'grad_norm': '1.306', 'learning_rate': '9.871e-05', 'epoch': '0.8803'}
+{'loss': '0.7127', 'grad_norm': '1.25', 'learning_rate': '9.871e-05', 'epoch': '0.8811'}
+{'loss': '0.7194', 'grad_norm': '1.135', 'learning_rate': '9.87e-05', 'epoch': '0.882'}
+{'loss': '0.6933', 'grad_norm': '1.171', 'learning_rate': '9.87e-05', 'epoch': '0.8828'}
+{'loss': '0.6825', 'grad_norm': '1.023', 'learning_rate': '9.869e-05', 'epoch': '0.8837'}
+{'loss': '0.7306', 'grad_norm': '1.103', 'learning_rate': '9.869e-05', 'epoch': '0.8846'}
+{'loss': '0.6956', 'grad_norm': '1.094', 'learning_rate': '9.868e-05', 'epoch': '0.8854'}
+{'loss': '0.774', 'grad_norm': '1.116', 'learning_rate': '9.867e-05', 'epoch': '0.8863'}
+{'loss': '0.6996', 'grad_norm': '1.026', 'learning_rate': '9.867e-05', 'epoch': '0.8872'}
+{'loss': '0.7604', 'grad_norm': '1.252', 'learning_rate': '9.866e-05', 'epoch': '0.888'}
+{'loss': '0.639', 'grad_norm': '1.06', 'learning_rate': '9.866e-05', 'epoch': '0.8889'}
+{'loss': '0.7859', 'grad_norm': '1.199', 'learning_rate': '9.865e-05', 'epoch': '0.8898'}
+{'loss': '0.712', 'grad_norm': '1.027', 'learning_rate': '9.865e-05', 'epoch': '0.8906'}
+{'loss': '0.6752', 'grad_norm': '1.162', 'learning_rate': '9.864e-05', 'epoch': '0.8915'}
+{'loss': '0.6104', 'grad_norm': '1.281', 'learning_rate': '9.863e-05', 'epoch': '0.8923'}
+{'loss': '0.654', 'grad_norm': '1.113', 'learning_rate': '9.863e-05', 'epoch': '0.8932'}
+{'loss': '0.7263', 'grad_norm': '1.626', 'learning_rate': '9.862e-05', 'epoch': '0.8941'}
+{'loss': '0.7799', 'grad_norm': '1.227', 'learning_rate': '9.862e-05', 'epoch': '0.8949'}
+{'loss': '0.6371', 'grad_norm': '1.073', 'learning_rate': '9.861e-05', 'epoch': '0.8958'}
+{'loss': '0.7333', 'grad_norm': '1.22', 'learning_rate': '9.861e-05', 'epoch': '0.8967'}
+{'loss': '0.7337', 'grad_norm': '1.229', 'learning_rate': '9.86e-05', 'epoch': '0.8975'}
+{'loss': '0.7423', 'grad_norm': '1.07', 'learning_rate': '9.859e-05', 'epoch': '0.8984'}
+{'loss': '0.701', 'grad_norm': '1.158', 'learning_rate': '9.859e-05', 'epoch': '0.8992'}
+{'loss': '0.6211', 'grad_norm': '0.9645', 'learning_rate': '9.858e-05', 'epoch': '0.9001'}
+{'loss': '0.73', 'grad_norm': '1.245', 'learning_rate': '9.858e-05', 'epoch': '0.901'}
+{'loss': '0.7462', 'grad_norm': '1.274', 'learning_rate': '9.857e-05', 'epoch': '0.9018'}
+{'loss': '0.7218', 'grad_norm': '1.368', 'learning_rate': '9.857e-05', 'epoch': '0.9027'}
+{'loss': '0.7046', 'grad_norm': '1.147', 'learning_rate': '9.856e-05', 'epoch': '0.9036'}
+{'loss': '0.719', 'grad_norm': '1.058', 'learning_rate': '9.855e-05', 'epoch': '0.9044'}
+{'loss': '0.7408', 'grad_norm': '1.407', 'learning_rate': '9.855e-05', 'epoch': '0.9053'}
+{'loss': '0.8191', 'grad_norm': '1.218', 'learning_rate': '9.854e-05', 'epoch': '0.9061'}
+{'eval_loss': '0.7449', 'eval_runtime': '832.5', 'eval_samples_per_second': '2.474', 'eval_steps_per_second': '2.474', 'epoch': '0.9061'}
+{'loss': '0.7296', 'grad_norm': '1.14', 'learning_rate': '9.854e-05', 'epoch': '0.907'}
+{'loss': '0.7142', 'grad_norm': '1.18', 'learning_rate': '9.853e-05', 'epoch': '0.9079'}
+{'loss': '0.7558', 'grad_norm': '1.077', 'learning_rate': '9.852e-05', 'epoch': '0.9087'}
+{'loss': '0.7025', 'grad_norm': '1.1', 'learning_rate': '9.852e-05', 'epoch': '0.9096'}
+{'loss': '0.6968', 'grad_norm': '1.024', 'learning_rate': '9.851e-05', 'epoch': '0.9105'}
+{'loss': '0.789', 'grad_norm': '1.231', 'learning_rate': '9.851e-05', 'epoch': '0.9113'}
+{'loss': '0.6825', 'grad_norm': '1.174', 'learning_rate': '9.85e-05', 'epoch': '0.9122'}
+{'loss': '0.7158', 'grad_norm': '1.217', 'learning_rate': '9.849e-05', 'epoch': '0.9131'}
+{'loss': '0.7164', 'grad_norm': '1.252', 'learning_rate': '9.849e-05', 'epoch': '0.9139'}
+{'loss': '0.6831', 'grad_norm': '1.133', 'learning_rate': '9.848e-05', 'epoch': '0.9148'}
+{'loss': '0.7149', 'grad_norm': '1.105', 'learning_rate': '9.848e-05', 'epoch': '0.9156'}
+{'loss': '0.7213', 'grad_norm': '1.039', 'learning_rate': '9.847e-05', 'epoch': '0.9165'}
+{'loss': '0.7033', 'grad_norm': '1.152', 'learning_rate': '9.846e-05', 'epoch': '0.9174'}
+{'loss': '0.806', 'grad_norm': '1.162', 'learning_rate': '9.846e-05', 'epoch': '0.9182'}
+{'loss': '0.7355', 'grad_norm': '1.225', 'learning_rate': '9.845e-05', 'epoch': '0.9191'}
+{'loss': '0.8131', 'grad_norm': '1.165', 'learning_rate': '9.845e-05', 'epoch': '0.92'}
+{'loss': '0.6909', 'grad_norm': '1.192', 'learning_rate': '9.844e-05', 'epoch': '0.9208'}
+{'loss': '0.6522', 'grad_norm': '1.211', 'learning_rate': '9.843e-05', 'epoch': '0.9217'}
+{'loss': '0.6672', 'grad_norm': '1.267', 'learning_rate': '9.843e-05', 'epoch': '0.9225'}
+{'loss': '0.7274', 'grad_norm': '1.704', 'learning_rate': '9.842e-05', 'epoch': '0.9234'}
+{'loss': '0.708', 'grad_norm': '1.151', 'learning_rate': '9.842e-05', 'epoch': '0.9243'}
+{'loss': '0.664', 'grad_norm': '1.041', 'learning_rate': '9.841e-05', 'epoch': '0.9251'}
+{'loss': '0.7096', 'grad_norm': '1.105', 'learning_rate': '9.84e-05', 'epoch': '0.926'}
+{'loss': '0.7125', 'grad_norm': '1.119', 'learning_rate': '9.84e-05', 'epoch': '0.9269'}
+{'loss': '0.7351', 'grad_norm': '1.261', 'learning_rate': '9.839e-05', 'epoch': '0.9277'}
+{'loss': '0.7125', 'grad_norm': '1.113', 'learning_rate': '9.838e-05', 'epoch': '0.9286'}
+{'loss': '0.7891', 'grad_norm': '1.477', 'learning_rate': '9.838e-05', 'epoch': '0.9294'}
+{'loss': '0.7371', 'grad_norm': '1.07', 'learning_rate': '9.837e-05', 'epoch': '0.9303'}
+{'loss': '0.7543', 'grad_norm': '1.205', 'learning_rate': '9.837e-05', 'epoch': '0.9312'}
+{'loss': '0.6904', 'grad_norm': '1.194', 'learning_rate': '9.836e-05', 'epoch': '0.932'}
+{'loss': '0.6923', 'grad_norm': '1.104', 'learning_rate': '9.835e-05', 'epoch': '0.9329'}
+{'loss': '0.81', 'grad_norm': '1.377', 'learning_rate': '9.835e-05', 'epoch': '0.9338'}
+{'loss': '0.7896', 'grad_norm': '1.087', 'learning_rate': '9.834e-05', 'epoch': '0.9346'}
+{'loss': '0.7659', 'grad_norm': '1.163', 'learning_rate': '9.833e-05', 'epoch': '0.9355'}
+{'loss': '0.7475', 'grad_norm': '1.405', 'learning_rate': '9.833e-05', 'epoch': '0.9364'}
+{'loss': '0.7019', 'grad_norm': '1.179', 'learning_rate': '9.832e-05', 'epoch': '0.9372'}
+{'loss': '0.6477', 'grad_norm': '1.082', 'learning_rate': '9.832e-05', 'epoch': '0.9381'}
+{'loss': '0.6516', 'grad_norm': '0.983', 'learning_rate': '9.831e-05', 'epoch': '0.9389'}
+{'loss': '0.6793', 'grad_norm': '1.112', 'learning_rate': '9.83e-05', 'epoch': '0.9398'}
+{'loss': '0.7221', 'grad_norm': '1.121', 'learning_rate': '9.83e-05', 'epoch': '0.9407'}
+{'loss': '0.7622', 'grad_norm': '1.246', 'learning_rate': '9.829e-05', 'epoch': '0.9415'}
+{'loss': '0.6988', 'grad_norm': '1.173', 'learning_rate': '9.828e-05', 'epoch': '0.9424'}
+{'loss': '0.7909', 'grad_norm': '1.177', 'learning_rate': '9.828e-05', 'epoch': '0.9433'}
+{'loss': '0.6651', 'grad_norm': '1.007', 'learning_rate': '9.827e-05', 'epoch': '0.9441'}
+{'loss': '0.7742', 'grad_norm': '1.184', 'learning_rate': '9.827e-05', 'epoch': '0.945'}
+{'loss': '0.7614', 'grad_norm': '1.123', 'learning_rate': '9.826e-05', 'epoch': '0.9458'}
+{'loss': '0.7302', 'grad_norm': '1.103', 'learning_rate': '9.825e-05', 'epoch': '0.9467'}
+{'loss': '0.7409', 'grad_norm': '1.521', 'learning_rate': '9.825e-05', 'epoch': '0.9476'}
+{'loss': '0.805', 'grad_norm': '1.185', 'learning_rate': '9.824e-05', 'epoch': '0.9484'}
+{'loss': '0.7267', 'grad_norm': '1.286', 'learning_rate': '9.823e-05', 'epoch': '0.9493'}
+{'eval_loss': '0.74', 'eval_runtime': '838.8', 'eval_samples_per_second': '2.456', 'eval_steps_per_second': '2.456', 'epoch': '0.9493'}
+{'loss': '0.7062', 'grad_norm': '1.216', 'learning_rate': '9.823e-05', 'epoch': '0.9502'}
+{'loss': '0.8243', 'grad_norm': '1.119', 'learning_rate': '9.822e-05', 'epoch': '0.951'}
+{'loss': '0.7231', 'grad_norm': '0.9969', 'learning_rate': '9.821e-05', 'epoch': '0.9519'}
+{'loss': '0.6634', 'grad_norm': '1.101', 'learning_rate': '9.821e-05', 'epoch': '0.9528'}
+{'loss': '0.6548', 'grad_norm': '1.084', 'learning_rate': '9.82e-05', 'epoch': '0.9536'}
+{'loss': '0.6434', 'grad_norm': '1.129', 'learning_rate': '9.819e-05', 'epoch': '0.9545'}
+{'loss': '0.8122', 'grad_norm': '2.303', 'learning_rate': '9.819e-05', 'epoch': '0.9553'}
+{'loss': '0.6121', 'grad_norm': '1.245', 'learning_rate': '9.818e-05', 'epoch': '0.9562'}
+{'loss': '0.757', 'grad_norm': '1.105', 'learning_rate': '9.817e-05', 'epoch': '0.9571'}
+{'loss': '0.7589', 'grad_norm': '1.316', 'learning_rate': '9.817e-05', 'epoch': '0.9579'}
+{'loss': '0.7434', 'grad_norm': '1.253', 'learning_rate': '9.816e-05', 'epoch': '0.9588'}
+{'loss': '0.8052', 'grad_norm': '1.196', 'learning_rate': '9.815e-05', 'epoch': '0.9597'}
+{'loss': '0.7332', 'grad_norm': '1.32', 'learning_rate': '9.815e-05', 'epoch': '0.9605'}
+{'loss': '0.681', 'grad_norm': '1.244', 'learning_rate': '9.814e-05', 'epoch': '0.9614'}
+{'loss': '0.727', 'grad_norm': '1.174', 'learning_rate': '9.813e-05', 'epoch': '0.9622'}
+{'loss': '0.7159', 'grad_norm': '1.102', 'learning_rate': '9.813e-05', 'epoch': '0.9631'}
+{'loss': '0.7146', 'grad_norm': '1.165', 'learning_rate': '9.812e-05', 'epoch': '0.964'}
+{'loss': '0.6874', 'grad_norm': '1.052', 'learning_rate': '9.811e-05', 'epoch': '0.9648'}
+{'loss': '0.6683', 'grad_norm': '1.229', 'learning_rate': '9.811e-05', 'epoch': '0.9657'}
+{'loss': '0.6631', 'grad_norm': '0.9826', 'learning_rate': '9.81e-05', 'epoch': '0.9666'}
+{'loss': '0.6689', 'grad_norm': '1.01', 'learning_rate': '9.809e-05', 'epoch': '0.9674'}
+{'loss': '0.6712', 'grad_norm': '0.9782', 'learning_rate': '9.809e-05', 'epoch': '0.9683'}
+{'loss': '0.7673', 'grad_norm': '1.137', 'learning_rate': '9.808e-05', 'epoch': '0.9691'}
+{'loss': '0.8086', 'grad_norm': '1.336', 'learning_rate': '9.807e-05', 'epoch': '0.97'}
+{'loss': '0.675', 'grad_norm': '1.151', 'learning_rate': '9.807e-05', 'epoch': '0.9709'}
+{'loss': '0.5928', 'grad_norm': '1.08', 'learning_rate': '9.806e-05', 'epoch': '0.9717'}
+{'loss': '0.6947', 'grad_norm': '1.049', 'learning_rate': '9.805e-05', 'epoch': '0.9726'}
+{'loss': '0.7274', 'grad_norm': '1.199', 'learning_rate': '9.805e-05', 'epoch': '0.9735'}
+{'loss': '0.6686', 'grad_norm': '0.9831', 'learning_rate': '9.804e-05', 'epoch': '0.9743'}
+{'loss': '0.7256', 'grad_norm': '1.166', 'learning_rate': '9.803e-05', 'epoch': '0.9752'}
+{'loss': '0.6866', 'grad_norm': '1.029', 'learning_rate': '9.803e-05', 'epoch': '0.9761'}
+{'loss': '0.6681', 'grad_norm': '1.102', 'learning_rate': '9.802e-05', 'epoch': '0.9769'}
+{'loss': '0.7476', 'grad_norm': '1.085', 'learning_rate': '9.801e-05', 'epoch': '0.9778'}
+{'loss': '0.7173', 'grad_norm': '1.183', 'learning_rate': '9.801e-05', 'epoch': '0.9786'}
+{'loss': '0.6769', 'grad_norm': '1.008', 'learning_rate': '9.8e-05', 'epoch': '0.9795'}
+{'loss': '0.7877', 'grad_norm': '1.157', 'learning_rate': '9.799e-05', 'epoch': '0.9804'}
+{'loss': '0.6844', 'grad_norm': '1.203', 'learning_rate': '9.799e-05', 'epoch': '0.9812'}
+{'loss': '0.6084', 'grad_norm': '1.081', 'learning_rate': '9.798e-05', 'epoch': '0.9821'}
+{'loss': '0.6785', 'grad_norm': '1.149', 'learning_rate': '9.797e-05', 'epoch': '0.983'}
+{'loss': '0.6112', 'grad_norm': '1.145', 'learning_rate': '9.796e-05', 'epoch': '0.9838'}
+{'loss': '0.7635', 'grad_norm': '1.136', 'learning_rate': '9.796e-05', 'epoch': '0.9847'}
+{'loss': '0.6471', 'grad_norm': '1.139', 'learning_rate': '9.795e-05', 'epoch': '0.9855'}
+{'loss': '0.7147', 'grad_norm': '1.156', 'learning_rate': '9.794e-05', 'epoch': '0.9864'}
+{'loss': '0.6414', 'grad_norm': '1.061', 'learning_rate': '9.794e-05', 'epoch': '0.9873'}
+{'loss': '0.6596', 'grad_norm': '1.166', 'learning_rate': '9.793e-05', 'epoch': '0.9881'}
+{'loss': '0.7144', 'grad_norm': '1.166', 'learning_rate': '9.792e-05', 'epoch': '0.989'}
+{'loss': '0.7956', 'grad_norm': '1.225', 'learning_rate': '9.792e-05', 'epoch': '0.9899'}
+{'loss': '0.5786', 'grad_norm': '0.9594', 'learning_rate': '9.791e-05', 'epoch': '0.9907'}
+{'loss': '0.6817', 'grad_norm': '1.07', 'learning_rate': '9.79e-05', 'epoch': '0.9916'}
+{'loss': '0.6845', 'grad_norm': '1.266', 'learning_rate': '9.789e-05', 'epoch': '0.9924'}
+{'eval_loss': '0.736', 'eval_runtime': '825.9', 'eval_samples_per_second': '2.494', 'eval_steps_per_second': '2.494', 'epoch': '0.9924'}
+{'loss': '0.669', 'grad_norm': '1.087', 'learning_rate': '9.789e-05', 'epoch': '0.9933'}
+{'loss': '0.6471', 'grad_norm': '1.005', 'learning_rate': '9.788e-05', 'epoch': '0.9942'}
+{'loss': '0.8018', 'grad_norm': '1.447', 'learning_rate': '9.787e-05', 'epoch': '0.995'}
+{'loss': '0.8078', 'grad_norm': '1.375', 'learning_rate': '9.787e-05', 'epoch': '0.9959'}
+{'loss': '0.6741', 'grad_norm': '1.098', 'learning_rate': '9.786e-05', 'epoch': '0.9968'}
+{'loss': '0.6844', 'grad_norm': '1.031', 'learning_rate': '9.785e-05', 'epoch': '0.9976'}
+{'loss': '0.7547', 'grad_norm': '1.448', 'learning_rate': '9.785e-05', 'epoch': '0.9985'}
+{'loss': '0.7714', 'grad_norm': '1.09', 'learning_rate': '9.784e-05', 'epoch': '0.9994'}
+{'loss': '0.6868', 'grad_norm': '1.789', 'learning_rate': '9.783e-05', 'epoch': '1'}
+{'loss': '0.6964', 'grad_norm': '1.002', 'learning_rate': '9.782e-05', 'epoch': '1.001'}
+{'loss': '0.7316', 'grad_norm': '1.365', 'learning_rate': '9.782e-05', 'epoch': '1.002'}
+{'loss': '0.7668', 'grad_norm': '1.225', 'learning_rate': '9.781e-05', 'epoch': '1.003'}
+{'loss': '0.7381', 'grad_norm': '1.192', 'learning_rate': '9.78e-05', 'epoch': '1.003'}
+{'loss': '0.6781', 'grad_norm': '1.146', 'learning_rate': '9.78e-05', 'epoch': '1.004'}
+{'loss': '0.8116', 'grad_norm': '1.488', 'learning_rate': '9.779e-05', 'epoch': '1.005'}
+{'loss': '0.6525', 'grad_norm': '1.349', 'learning_rate': '9.778e-05', 'epoch': '1.006'}
+{'loss': '0.5843', 'grad_norm': '1.097', 'learning_rate': '9.777e-05', 'epoch': '1.007'}
+{'loss': '0.6595', 'grad_norm': '1.064', 'learning_rate': '9.777e-05', 'epoch': '1.008'}
+{'loss': '0.6963', 'grad_norm': '1.281', 'learning_rate': '9.776e-05', 'epoch': '1.009'}
+{'loss': '0.6246', 'grad_norm': '0.9492', 'learning_rate': '9.775e-05', 'epoch': '1.009'}
+{'loss': '0.6352', 'grad_norm': '1.199', 'learning_rate': '9.774e-05', 'epoch': '1.01'}
+{'loss': '0.6885', 'grad_norm': '1.218', 'learning_rate': '9.774e-05', 'epoch': '1.011'}
+{'loss': '0.5945', 'grad_norm': '1.191', 'learning_rate': '9.773e-05', 'epoch': '1.012'}
+{'loss': '0.6534', 'grad_norm': '1.21', 'learning_rate': '9.772e-05', 'epoch': '1.013'}
+{'loss': '0.6619', 'grad_norm': '1.155', 'learning_rate': '9.772e-05', 'epoch': '1.014'}
+{'loss': '0.6581', 'grad_norm': '1.083', 'learning_rate': '9.771e-05', 'epoch': '1.015'}
+{'loss': '0.6671', 'grad_norm': '1.033', 'learning_rate': '9.77e-05', 'epoch': '1.016'}
+{'loss': '0.6802', 'grad_norm': '1.147', 'learning_rate': '9.769e-05', 'epoch': '1.016'}
+{'loss': '0.7206', 'grad_norm': '1.22', 'learning_rate': '9.769e-05', 'epoch': '1.017'}
+{'loss': '0.6885', 'grad_norm': '1.234', 'learning_rate': '9.768e-05', 'epoch': '1.018'}
+{'loss': '0.6551', 'grad_norm': '1.086', 'learning_rate': '9.767e-05', 'epoch': '1.019'}
+{'loss': '0.6479', 'grad_norm': '1.118', 'learning_rate': '9.766e-05', 'epoch': '1.02'}
+{'loss': '0.6311', 'grad_norm': '1.2', 'learning_rate': '9.766e-05', 'epoch': '1.021'}
+{'loss': '0.6373', 'grad_norm': '1.355', 'learning_rate': '9.765e-05', 'epoch': '1.022'}
+{'loss': '0.6596', 'grad_norm': '1.219', 'learning_rate': '9.764e-05', 'epoch': '1.022'}
+{'loss': '0.6299', 'grad_norm': '1.149', 'learning_rate': '9.763e-05', 'epoch': '1.023'}
+{'loss': '0.6789', 'grad_norm': '1.115', 'learning_rate': '9.763e-05', 'epoch': '1.024'}
+{'loss': '0.5591', 'grad_norm': '1.069', 'learning_rate': '9.762e-05', 'epoch': '1.025'}
+{'loss': '0.7165', 'grad_norm': '1.19', 'learning_rate': '9.761e-05', 'epoch': '1.026'}
+{'loss': '0.6791', 'grad_norm': '1.179', 'learning_rate': '9.76e-05', 'epoch': '1.027'}
+{'loss': '0.7422', 'grad_norm': '1.362', 'learning_rate': '9.76e-05', 'epoch': '1.028'}
+{'loss': '0.7033', 'grad_norm': '1.196', 'learning_rate': '9.759e-05', 'epoch': '1.028'}
+{'loss': '0.6953', 'grad_norm': '1.323', 'learning_rate': '9.758e-05', 'epoch': '1.029'}
+{'loss': '0.7288', 'grad_norm': '1.234', 'learning_rate': '9.757e-05', 'epoch': '1.03'}
+{'loss': '0.6813', 'grad_norm': '1.071', 'learning_rate': '9.757e-05', 'epoch': '1.031'}
+{'loss': '0.7174', 'grad_norm': '1.265', 'learning_rate': '9.756e-05', 'epoch': '1.032'}
+{'loss': '0.651', 'grad_norm': '1.156', 'learning_rate': '9.755e-05', 'epoch': '1.033'}
+{'loss': '0.7184', 'grad_norm': '1.326', 'learning_rate': '9.754e-05', 'epoch': '1.034'}
+{'loss': '0.6645', 'grad_norm': '1.226', 'learning_rate': '9.754e-05', 'epoch': '1.035'}
+{'loss': '0.6755', 'grad_norm': '1.287', 'learning_rate': '9.753e-05', 'epoch': '1.035'}
+{'eval_loss': '0.7347', 'eval_runtime': '825.8', 'eval_samples_per_second': '2.494', 'eval_steps_per_second': '2.494', 'epoch': '1.035'}
+{'loss': '0.7082', 'grad_norm': '1.305', 'learning_rate': '9.752e-05', 'epoch': '1.036'}
+{'loss': '0.7213', 'grad_norm': '1.218', 'learning_rate': '9.751e-05', 'epoch': '1.037'}
+{'loss': '0.6739', 'grad_norm': '1.295', 'learning_rate': '9.751e-05', 'epoch': '1.038'}
+{'loss': '0.7671', 'grad_norm': '1.138', 'learning_rate': '9.75e-05', 'epoch': '1.039'}
+{'loss': '0.6593', 'grad_norm': '1.126', 'learning_rate': '9.749e-05', 'epoch': '1.04'}
+{'loss': '0.7131', 'grad_norm': '1.236', 'learning_rate': '9.748e-05', 'epoch': '1.041'}
+{'loss': '0.6989', 'grad_norm': '1.202', 'learning_rate': '9.747e-05', 'epoch': '1.041'}
+{'loss': '0.623', 'grad_norm': '1.127', 'learning_rate': '9.747e-05', 'epoch': '1.042'}
+{'loss': '0.649', 'grad_norm': '1.125', 'learning_rate': '9.746e-05', 'epoch': '1.043'}
+{'loss': '0.6211', 'grad_norm': '1.214', 'learning_rate': '9.745e-05', 'epoch': '1.044'}
+{'loss': '0.6243', 'grad_norm': '1.131', 'learning_rate': '9.744e-05', 'epoch': '1.045'}
+{'loss': '0.6313', 'grad_norm': '1.203', 'learning_rate': '9.744e-05', 'epoch': '1.046'}
+{'loss': '0.6191', 'grad_norm': '1.056', 'learning_rate': '9.743e-05', 'epoch': '1.047'}
+{'loss': '0.7497', 'grad_norm': '1.507', 'learning_rate': '9.742e-05', 'epoch': '1.047'}
+{'loss': '0.6289', 'grad_norm': '1.145', 'learning_rate': '9.741e-05', 'epoch': '1.048'}
+{'loss': '0.6938', 'grad_norm': '1.007', 'learning_rate': '9.74e-05', 'epoch': '1.049'}
+{'loss': '0.753', 'grad_norm': '1.199', 'learning_rate': '9.74e-05', 'epoch': '1.05'}
+{'loss': '0.7194', 'grad_norm': '1.217', 'learning_rate': '9.739e-05', 'epoch': '1.051'}
+{'loss': '0.6488', 'grad_norm': '1.038', 'learning_rate': '9.738e-05', 'epoch': '1.052'}
+{'loss': '0.6672', 'grad_norm': '1.156', 'learning_rate': '9.737e-05', 'epoch': '1.053'}
+{'loss': '0.651', 'grad_norm': '1.155', 'learning_rate': '9.737e-05', 'epoch': '1.054'}
+{'loss': '0.6812', 'grad_norm': '1.137', 'learning_rate': '9.736e-05', 'epoch': '1.054'}
+{'loss': '0.7331', 'grad_norm': '1.178', 'learning_rate': '9.735e-05', 'epoch': '1.055'}
+{'loss': '0.7317', 'grad_norm': '1.298', 'learning_rate': '9.734e-05', 'epoch': '1.056'}
+{'loss': '0.6514', 'grad_norm': '1.119', 'learning_rate': '9.733e-05', 'epoch': '1.057'}
+{'loss': '0.6469', 'grad_norm': '1.326', 'learning_rate': '9.733e-05', 'epoch': '1.058'}
+{'loss': '0.586', 'grad_norm': '0.9767', 'learning_rate': '9.732e-05', 'epoch': '1.059'}
+{'loss': '0.6784', 'grad_norm': '1.136', 'learning_rate': '9.731e-05', 'epoch': '1.06'}
+{'loss': '0.7894', 'grad_norm': '1.348', 'learning_rate': '9.73e-05', 'epoch': '1.06'}
+{'loss': '0.7098', 'grad_norm': '1.141', 'learning_rate': '9.729e-05', 'epoch': '1.061'}
+{'loss': '0.7138', 'grad_norm': '1.267', 'learning_rate': '9.729e-05', 'epoch': '1.062'}
+{'loss': '0.7136', 'grad_norm': '1.258', 'learning_rate': '9.728e-05', 'epoch': '1.063'}
+{'loss': '0.6273', 'grad_norm': '1.099', 'learning_rate': '9.727e-05', 'epoch': '1.064'}
+{'loss': '0.7076', 'grad_norm': '1.155', 'learning_rate': '9.726e-05', 'epoch': '1.065'}
+{'loss': '0.6857', 'grad_norm': '1.086', 'learning_rate': '9.725e-05', 'epoch': '1.066'}
+{'loss': '0.7311', 'grad_norm': '1.274', 'learning_rate': '9.725e-05', 'epoch': '1.066'}
+{'loss': '0.7728', 'grad_norm': '1.504', 'learning_rate': '9.724e-05', 'epoch': '1.067'}
+{'loss': '0.6259', 'grad_norm': '1.223', 'learning_rate': '9.723e-05', 'epoch': '1.068'}
+{'loss': '0.7133', 'grad_norm': '1.324', 'learning_rate': '9.722e-05', 'epoch': '1.069'}
+{'loss': '0.6068', 'grad_norm': '1.229', 'learning_rate': '9.721e-05', 'epoch': '1.07'}
+{'loss': '0.6885', 'grad_norm': '1.31', 'learning_rate': '9.721e-05', 'epoch': '1.071'}
+{'loss': '0.7165', 'grad_norm': '1.438', 'learning_rate': '9.72e-05', 'epoch': '1.072'}
+{'loss': '0.6327', 'grad_norm': '1.098', 'learning_rate': '9.719e-05', 'epoch': '1.072'}
+{'loss': '0.5713', 'grad_norm': '0.992', 'learning_rate': '9.718e-05', 'epoch': '1.073'}
+{'loss': '0.6923', 'grad_norm': '1.077', 'learning_rate': '9.717e-05', 'epoch': '1.074'}
+{'loss': '0.6737', 'grad_norm': '1.155', 'learning_rate': '9.717e-05', 'epoch': '1.075'}
+{'loss': '0.6422', 'grad_norm': '1.113', 'learning_rate': '9.716e-05', 'epoch': '1.076'}
+{'loss': '0.64', 'grad_norm': '1.065', 'learning_rate': '9.715e-05', 'epoch': '1.077'}
+{'loss': '0.6218', 'grad_norm': '1.139', 'learning_rate': '9.714e-05', 'epoch': '1.078'}
+{'loss': '0.6595', 'grad_norm': '1.332', 'learning_rate': '9.713e-05', 'epoch': '1.079'}
+{'eval_loss': '0.7295', 'eval_runtime': '836.5', 'eval_samples_per_second': '2.463', 'eval_steps_per_second': '2.463', 'epoch': '1.079'}
+{'loss': '0.6668', 'grad_norm': '1.312', 'learning_rate': '9.712e-05', 'epoch': '1.079'}
+{'loss': '0.6927', 'grad_norm': '1.363', 'learning_rate': '9.712e-05', 'epoch': '1.08'}
+{'loss': '0.6707', 'grad_norm': '1.238', 'learning_rate': '9.711e-05', 'epoch': '1.081'}
+{'loss': '0.6011', 'grad_norm': '1.166', 'learning_rate': '9.71e-05', 'epoch': '1.082'}
+{'loss': '0.689', 'grad_norm': '1.302', 'learning_rate': '9.709e-05', 'epoch': '1.083'}
+{'loss': '0.6217', 'grad_norm': '1.055', 'learning_rate': '9.708e-05', 'epoch': '1.084'}
+{'loss': '0.644', 'grad_norm': '1.186', 'learning_rate': '9.707e-05', 'epoch': '1.085'}
+{'loss': '0.6825', 'grad_norm': '1.117', 'learning_rate': '9.707e-05', 'epoch': '1.085'}
+{'loss': '0.6471', 'grad_norm': '1.129', 'learning_rate': '9.706e-05', 'epoch': '1.086'}
+{'loss': '0.6536', 'grad_norm': '1.325', 'learning_rate': '9.705e-05', 'epoch': '1.087'}
+{'loss': '0.678', 'grad_norm': '1.359', 'learning_rate': '9.704e-05', 'epoch': '1.088'}
+{'loss': '0.5904', 'grad_norm': '1.133', 'learning_rate': '9.703e-05', 'epoch': '1.089'}
+{'loss': '0.6073', 'grad_norm': '1.182', 'learning_rate': '9.703e-05', 'epoch': '1.09'}
+{'loss': '0.6752', 'grad_norm': '1.253', 'learning_rate': '9.702e-05', 'epoch': '1.091'}
+{'loss': '0.7193', 'grad_norm': '1.281', 'learning_rate': '9.701e-05', 'epoch': '1.091'}
+{'loss': '0.6992', 'grad_norm': '1.143', 'learning_rate': '9.7e-05', 'epoch': '1.092'}
+{'loss': '0.701', 'grad_norm': '1.267', 'learning_rate': '9.699e-05', 'epoch': '1.093'}
+{'loss': '0.6946', 'grad_norm': '1.208', 'learning_rate': '9.698e-05', 'epoch': '1.094'}
+{'loss': '0.6895', 'grad_norm': '1.206', 'learning_rate': '9.697e-05', 'epoch': '1.095'}
+{'loss': '0.6967', 'grad_norm': '1.171', 'learning_rate': '9.697e-05', 'epoch': '1.096'}
+{'loss': '0.6525', 'grad_norm': '1.155', 'learning_rate': '9.696e-05', 'epoch': '1.097'}
+{'loss': '0.577', 'grad_norm': '1.097', 'learning_rate': '9.695e-05', 'epoch': '1.098'}
+{'loss': '0.6986', 'grad_norm': '1.321', 'learning_rate': '9.694e-05', 'epoch': '1.098'}
+{'loss': '0.7014', 'grad_norm': '1.168', 'learning_rate': '9.693e-05', 'epoch': '1.099'}
+{'loss': '0.5787', 'grad_norm': '1.153', 'learning_rate': '9.692e-05', 'epoch': '1.1'}
+{'loss': '0.7559', 'grad_norm': '1.251', 'learning_rate': '9.692e-05', 'epoch': '1.101'}
+{'loss': '0.6488', 'grad_norm': '1.091', 'learning_rate': '9.691e-05', 'epoch': '1.102'}
+{'loss': '0.6347', 'grad_norm': '1.368', 'learning_rate': '9.69e-05', 'epoch': '1.103'}
+{'loss': '0.6471', 'grad_norm': '1.216', 'learning_rate': '9.689e-05', 'epoch': '1.104'}
+{'loss': '0.6462', 'grad_norm': '1.256', 'learning_rate': '9.688e-05', 'epoch': '1.104'}
+{'loss': '0.6889', 'grad_norm': '1.126', 'learning_rate': '9.687e-05', 'epoch': '1.105'}
+{'loss': '0.6318', 'grad_norm': '1.15', 'learning_rate': '9.686e-05', 'epoch': '1.106'}
+{'loss': '0.6538', 'grad_norm': '1.178', 'learning_rate': '9.686e-05', 'epoch': '1.107'}
+{'loss': '0.6707', 'grad_norm': '1.265', 'learning_rate': '9.685e-05', 'epoch': '1.108'}
+{'loss': '0.7573', 'grad_norm': '1.113', 'learning_rate': '9.684e-05', 'epoch': '1.109'}
+{'loss': '0.6603', 'grad_norm': '1.278', 'learning_rate': '9.683e-05', 'epoch': '1.11'}
+{'loss': '0.6524', 'grad_norm': '1.251', 'learning_rate': '9.682e-05', 'epoch': '1.11'}
+{'loss': '0.6776', 'grad_norm': '1.263', 'learning_rate': '9.681e-05', 'epoch': '1.111'}
+{'loss': '0.6665', 'grad_norm': '1.021', 'learning_rate': '9.68e-05', 'epoch': '1.112'}
+{'loss': '0.7467', 'grad_norm': '1.222', 'learning_rate': '9.68e-05', 'epoch': '1.113'}
+{'loss': '0.6735', 'grad_norm': '1.118', 'learning_rate': '9.679e-05', 'epoch': '1.114'}
+{'loss': '0.5725', 'grad_norm': '1.127', 'learning_rate': '9.678e-05', 'epoch': '1.115'}
+{'loss': '0.635', 'grad_norm': '1.298', 'learning_rate': '9.677e-05', 'epoch': '1.116'}
+{'loss': '0.6814', 'grad_norm': '1.191', 'learning_rate': '9.676e-05', 'epoch': '1.117'}
+{'loss': '0.6738', 'grad_norm': '1.179', 'learning_rate': '9.675e-05', 'epoch': '1.117'}
+{'loss': '0.6663', 'grad_norm': '1.133', 'learning_rate': '9.674e-05', 'epoch': '1.118'}
+{'loss': '0.6774', 'grad_norm': '1.125', 'learning_rate': '9.673e-05', 'epoch': '1.119'}
+{'loss': '0.6257', 'grad_norm': '1.141', 'learning_rate': '9.673e-05', 'epoch': '1.12'}
+{'loss': '0.7029', 'grad_norm': '1.161', 'learning_rate': '9.672e-05', 'epoch': '1.121'}
+{'loss': '0.6041', 'grad_norm': '1.048', 'learning_rate': '9.671e-05', 'epoch': '1.122'}
+{'eval_loss': '0.7287', 'eval_runtime': '831.8', 'eval_samples_per_second': '2.477', 'eval_steps_per_second': '2.477', 'epoch': '1.122'}
+{'loss': '0.6873', 'grad_norm': '1.057', 'learning_rate': '9.67e-05', 'epoch': '1.123'}
+{'loss': '0.6173', 'grad_norm': '1.094', 'learning_rate': '9.669e-05', 'epoch': '1.123'}
+{'loss': '0.6143', 'grad_norm': '1.118', 'learning_rate': '9.668e-05', 'epoch': '1.124'}
+{'loss': '0.7226', 'grad_norm': '1.251', 'learning_rate': '9.667e-05', 'epoch': '1.125'}
+{'loss': '0.683', 'grad_norm': '1.233', 'learning_rate': '9.666e-05', 'epoch': '1.126'}
+{'loss': '0.6835', 'grad_norm': '1.29', 'learning_rate': '9.666e-05', 'epoch': '1.127'}
+{'loss': '0.6994', 'grad_norm': '1.478', 'learning_rate': '9.665e-05', 'epoch': '1.128'}
+{'loss': '0.6819', 'grad_norm': '1.267', 'learning_rate': '9.664e-05', 'epoch': '1.129'}
+{'loss': '0.62', 'grad_norm': '1.076', 'learning_rate': '9.663e-05', 'epoch': '1.129'}
+{'loss': '0.7139', 'grad_norm': '1.204', 'learning_rate': '9.662e-05', 'epoch': '1.13'}
+{'loss': '0.7238', 'grad_norm': '1.264', 'learning_rate': '9.661e-05', 'epoch': '1.131'}
+{'loss': '0.6401', 'grad_norm': '1.042', 'learning_rate': '9.66e-05', 'epoch': '1.132'}
+{'loss': '0.6306', 'grad_norm': '1.23', 'learning_rate': '9.659e-05', 'epoch': '1.133'}
+{'loss': '0.5752', 'grad_norm': '1.255', 'learning_rate': '9.658e-05', 'epoch': '1.134'}
+{'loss': '0.663', 'grad_norm': '1.157', 'learning_rate': '9.658e-05', 'epoch': '1.135'}
+{'loss': '0.6832', 'grad_norm': '1.194', 'learning_rate': '9.657e-05', 'epoch': '1.135'}
+{'loss': '0.6824', 'grad_norm': '1.233', 'learning_rate': '9.656e-05', 'epoch': '1.136'}
+{'loss': '0.6671', 'grad_norm': '1.393', 'learning_rate': '9.655e-05', 'epoch': '1.137'}
+{'loss': '0.6622', 'grad_norm': '1.3', 'learning_rate': '9.654e-05', 'epoch': '1.138'}
+{'loss': '0.6419', 'grad_norm': '1.158', 'learning_rate': '9.653e-05', 'epoch': '1.139'}
+{'loss': '0.6844', 'grad_norm': '1.358', 'learning_rate': '9.652e-05', 'epoch': '1.14'}
+{'loss': '0.6762', 'grad_norm': '1.262', 'learning_rate': '9.651e-05', 'epoch': '1.141'}
+{'loss': '0.7258', 'grad_norm': '1.22', 'learning_rate': '9.65e-05', 'epoch': '1.142'}
+{'loss': '0.571', 'grad_norm': '1.023', 'learning_rate': '9.65e-05', 'epoch': '1.142'}
+{'loss': '0.7354', 'grad_norm': '1.096', 'learning_rate': '9.649e-05', 'epoch': '1.143'}
+{'loss': '0.7084', 'grad_norm': '0.9393', 'learning_rate': '9.648e-05', 'epoch': '1.144'}
+{'loss': '0.6405', 'grad_norm': '1.09', 'learning_rate': '9.647e-05', 'epoch': '1.145'}
+{'loss': '0.6177', 'grad_norm': '1.158', 'learning_rate': '9.646e-05', 'epoch': '1.146'}
+{'loss': '0.6948', 'grad_norm': '1.134', 'learning_rate': '9.645e-05', 'epoch': '1.147'}
+{'loss': '0.6393', 'grad_norm': '1.029', 'learning_rate': '9.644e-05', 'epoch': '1.148'}
+{'loss': '0.6385', 'grad_norm': '1.209', 'learning_rate': '9.643e-05', 'epoch': '1.148'}
+{'loss': '0.5909', 'grad_norm': '0.9906', 'learning_rate': '9.642e-05', 'epoch': '1.149'}
+{'loss': '0.7009', 'grad_norm': '1.167', 'learning_rate': '9.641e-05', 'epoch': '1.15'}
+{'loss': '0.6037', 'grad_norm': '1.165', 'learning_rate': '9.64e-05', 'epoch': '1.151'}
+{'loss': '0.6159', 'grad_norm': '1.009', 'learning_rate': '9.64e-05', 'epoch': '1.152'}
+{'loss': '0.728', 'grad_norm': '1.224', 'learning_rate': '9.639e-05', 'epoch': '1.153'}
+{'loss': '0.6434', 'grad_norm': '1.093', 'learning_rate': '9.638e-05', 'epoch': '1.154'}
+{'loss': '0.6771', 'grad_norm': '1.366', 'learning_rate': '9.637e-05', 'epoch': '1.154'}
+{'loss': '0.6382', 'grad_norm': '2.25', 'learning_rate': '9.636e-05', 'epoch': '1.155'}
+{'loss': '0.6734', 'grad_norm': '1.274', 'learning_rate': '9.635e-05', 'epoch': '1.156'}
+{'loss': '0.7601', 'grad_norm': '1.372', 'learning_rate': '9.634e-05', 'epoch': '1.157'}
+{'loss': '0.5959', 'grad_norm': '1.131', 'learning_rate': '9.633e-05', 'epoch': '1.158'}
+{'loss': '0.6151', 'grad_norm': '1.063', 'learning_rate': '9.632e-05', 'epoch': '1.159'}
+{'loss': '0.6554', 'grad_norm': '1.075', 'learning_rate': '9.631e-05', 'epoch': '1.16'}
+{'loss': '0.6963', 'grad_norm': '1.18', 'learning_rate': '9.63e-05', 'epoch': '1.161'}
+{'loss': '0.6634', 'grad_norm': '1.255', 'learning_rate': '9.629e-05', 'epoch': '1.161'}
+{'loss': '0.6728', 'grad_norm': '1.237', 'learning_rate': '9.628e-05', 'epoch': '1.162'}
+{'loss': '0.7018', 'grad_norm': '1.405', 'learning_rate': '9.628e-05', 'epoch': '1.163'}
+{'loss': '0.6567', 'grad_norm': '1.154', 'learning_rate': '9.627e-05', 'epoch': '1.164'}
+{'loss': '0.7018', 'grad_norm': '1.194', 'learning_rate': '9.626e-05', 'epoch': '1.165'}
+{'eval_loss': '0.724', 'eval_runtime': '824.3', 'eval_samples_per_second': '2.499', 'eval_steps_per_second': '2.499', 'epoch': '1.165'}
+{'loss': '0.7059', 'grad_norm': '1.151', 'learning_rate': '9.625e-05', 'epoch': '1.166'}
+{'loss': '0.7458', 'grad_norm': '1.474', 'learning_rate': '9.624e-05', 'epoch': '1.167'}
+{'loss': '0.6526', 'grad_norm': '1.053', 'learning_rate': '9.623e-05', 'epoch': '1.167'}
+{'loss': '0.6055', 'grad_norm': '1.041', 'learning_rate': '9.622e-05', 'epoch': '1.168'}
+{'loss': '0.6665', 'grad_norm': '1.214', 'learning_rate': '9.621e-05', 'epoch': '1.169'}
+{'loss': '0.7209', 'grad_norm': '1.253', 'learning_rate': '9.62e-05', 'epoch': '1.17'}
+{'loss': '0.6269', 'grad_norm': '1.21', 'learning_rate': '9.619e-05', 'epoch': '1.171'}
+{'loss': '0.6063', 'grad_norm': '1.036', 'learning_rate': '9.618e-05', 'epoch': '1.172'}
+{'loss': '0.7085', 'grad_norm': '1.288', 'learning_rate': '9.617e-05', 'epoch': '1.173'}
+{'loss': '0.6516', 'grad_norm': '1.281', 'learning_rate': '9.616e-05', 'epoch': '1.173'}
+{'loss': '0.6746', 'grad_norm': '1.117', 'learning_rate': '9.615e-05', 'epoch': '1.174'}
+{'loss': '0.6825', 'grad_norm': '1.122', 'learning_rate': '9.614e-05', 'epoch': '1.175'}
+{'loss': '0.6021', 'grad_norm': '1.156', 'learning_rate': '9.613e-05', 'epoch': '1.176'}
+{'loss': '0.6431', 'grad_norm': '1.25', 'learning_rate': '9.612e-05', 'epoch': '1.177'}
+{'loss': '0.7195', 'grad_norm': '1.616', 'learning_rate': '9.612e-05', 'epoch': '1.178'}
+{'loss': '0.7104', 'grad_norm': '1.32', 'learning_rate': '9.611e-05', 'epoch': '1.179'}
+{'loss': '0.6501', 'grad_norm': '1.212', 'learning_rate': '9.61e-05', 'epoch': '1.18'}
+{'loss': '0.715', 'grad_norm': '2.272', 'learning_rate': '9.609e-05', 'epoch': '1.18'}
+{'loss': '0.6272', 'grad_norm': '1.052', 'learning_rate': '9.608e-05', 'epoch': '1.181'}
+{'loss': '0.6598', 'grad_norm': '1.257', 'learning_rate': '9.607e-05', 'epoch': '1.182'}
+{'loss': '0.7097', 'grad_norm': '1.145', 'learning_rate': '9.606e-05', 'epoch': '1.183'}
+{'loss': '0.6693', 'grad_norm': '1.37', 'learning_rate': '9.605e-05', 'epoch': '1.184'}
+{'loss': '0.6453', 'grad_norm': '1.142', 'learning_rate': '9.604e-05', 'epoch': '1.185'}
+{'loss': '0.6197', 'grad_norm': '1.27', 'learning_rate': '9.603e-05', 'epoch': '1.186'}
+{'loss': '0.6525', 'grad_norm': '1.117', 'learning_rate': '9.602e-05', 'epoch': '1.186'}
+{'loss': '0.646', 'grad_norm': '1.133', 'learning_rate': '9.601e-05', 'epoch': '1.187'}
+{'loss': '0.6176', 'grad_norm': '1.136', 'learning_rate': '9.6e-05', 'epoch': '1.188'}
+{'loss': '0.6782', 'grad_norm': '1.122', 'learning_rate': '9.599e-05', 'epoch': '1.189'}
+{'loss': '0.6387', 'grad_norm': '1.025', 'learning_rate': '9.598e-05', 'epoch': '1.19'}
+{'loss': '0.6974', 'grad_norm': '1.354', 'learning_rate': '9.597e-05', 'epoch': '1.191'}
+{'loss': '0.604', 'grad_norm': '1.002', 'learning_rate': '9.596e-05', 'epoch': '1.192'}
+{'loss': '0.6544', 'grad_norm': '1.151', 'learning_rate': '9.595e-05', 'epoch': '1.192'}
+{'loss': '0.7389', 'grad_norm': '1.413', 'learning_rate': '9.594e-05', 'epoch': '1.193'}
+{'loss': '0.7128', 'grad_norm': '1.204', 'learning_rate': '9.593e-05', 'epoch': '1.194'}
+{'loss': '0.6594', 'grad_norm': '1.215', 'learning_rate': '9.592e-05', 'epoch': '1.195'}
+{'loss': '0.6128', 'grad_norm': '1.623', 'learning_rate': '9.591e-05', 'epoch': '1.196'}
+{'loss': '0.6538', 'grad_norm': '1.19', 'learning_rate': '9.59e-05', 'epoch': '1.197'}
+{'loss': '0.6326', 'grad_norm': '1.16', 'learning_rate': '9.589e-05', 'epoch': '1.198'}
+{'loss': '0.6439', 'grad_norm': '1.131', 'learning_rate': '9.588e-05', 'epoch': '1.198'}
+{'loss': '0.6703', 'grad_norm': '1.018', 'learning_rate': '9.587e-05', 'epoch': '1.199'}
+{'loss': '0.7694', 'grad_norm': '1.551', 'learning_rate': '9.587e-05', 'epoch': '1.2'}
+{'loss': '0.6118', 'grad_norm': '1.077', 'learning_rate': '9.586e-05', 'epoch': '1.201'}
+{'loss': '0.6376', 'grad_norm': '1.207', 'learning_rate': '9.585e-05', 'epoch': '1.202'}
+{'loss': '0.6527', 'grad_norm': '1.129', 'learning_rate': '9.584e-05', 'epoch': '1.203'}
+{'loss': '0.708', 'grad_norm': '1.308', 'learning_rate': '9.583e-05', 'epoch': '1.204'}
+{'loss': '0.6416', 'grad_norm': '1.167', 'learning_rate': '9.582e-05', 'epoch': '1.205'}
+{'loss': '0.6539', 'grad_norm': '1.125', 'learning_rate': '9.581e-05', 'epoch': '1.205'}
+{'loss': '0.6945', 'grad_norm': '1.407', 'learning_rate': '9.58e-05', 'epoch': '1.206'}
+{'loss': '0.7065', 'grad_norm': '1.186', 'learning_rate': '9.579e-05', 'epoch': '1.207'}
+{'loss': '0.579', 'grad_norm': '1.086', 'learning_rate': '9.578e-05', 'epoch': '1.208'}
+{'eval_loss': '0.722', 'eval_runtime': '831.6', 'eval_samples_per_second': '2.477', 'eval_steps_per_second': '2.477', 'epoch': '1.208'}
+{'loss': '0.6645', 'grad_norm': '1.276', 'learning_rate': '9.577e-05', 'epoch': '1.209'}
+{'loss': '0.632', 'grad_norm': '1.35', 'learning_rate': '9.576e-05', 'epoch': '1.21'}
+{'loss': '0.6297', 'grad_norm': '1.264', 'learning_rate': '9.575e-05', 'epoch': '1.211'}
+{'loss': '0.757', 'grad_norm': '1.079', 'learning_rate': '9.574e-05', 'epoch': '1.211'}
+{'loss': '0.6713', 'grad_norm': '1.097', 'learning_rate': '9.573e-05', 'epoch': '1.212'}
+{'loss': '0.6775', 'grad_norm': '1.1', 'learning_rate': '9.572e-05', 'epoch': '1.213'}
+{'loss': '0.6691', 'grad_norm': '1.739', 'learning_rate': '9.571e-05', 'epoch': '1.214'}
+{'loss': '0.6406', 'grad_norm': '1.228', 'learning_rate': '9.57e-05', 'epoch': '1.215'}
+{'loss': '0.7009', 'grad_norm': '1.133', 'learning_rate': '9.569e-05', 'epoch': '1.216'}
+{'loss': '0.7438', 'grad_norm': '1.304', 'learning_rate': '9.568e-05', 'epoch': '1.217'}
+{'loss': '0.6938', 'grad_norm': '1.219', 'learning_rate': '9.567e-05', 'epoch': '1.217'}
+{'loss': '0.7586', 'grad_norm': '1.321', 'learning_rate': '9.566e-05', 'epoch': '1.218'}
+{'loss': '0.6356', 'grad_norm': '1.354', 'learning_rate': '9.565e-05', 'epoch': '1.219'}
+{'loss': '0.6305', 'grad_norm': '1.102', 'learning_rate': '9.564e-05', 'epoch': '1.22'}
+{'loss': '0.6459', 'grad_norm': '0.9658', 'learning_rate': '9.563e-05', 'epoch': '1.221'}
+{'loss': '0.7034', 'grad_norm': '1.398', 'learning_rate': '9.562e-05', 'epoch': '1.222'}
+{'loss': '0.663', 'grad_norm': '1.174', 'learning_rate': '9.561e-05', 'epoch': '1.223'}
+{'loss': '0.7167', 'grad_norm': '1.207', 'learning_rate': '9.56e-05', 'epoch': '1.224'}
+{'loss': '0.6712', 'grad_norm': '1.252', 'learning_rate': '9.559e-05', 'epoch': '1.224'}
+{'loss': '0.7135', 'grad_norm': '1.23', 'learning_rate': '9.558e-05', 'epoch': '1.225'}
+{'loss': '0.6279', 'grad_norm': '1.151', 'learning_rate': '9.557e-05', 'epoch': '1.226'}
+{'loss': '0.6849', 'grad_norm': '1.202', 'learning_rate': '9.556e-05', 'epoch': '1.227'}
+{'loss': '0.7109', 'grad_norm': '1.36', 'learning_rate': '9.555e-05', 'epoch': '1.228'}
+{'loss': '0.5983', 'grad_norm': '1.156', 'learning_rate': '9.554e-05', 'epoch': '1.229'}
+{'loss': '0.6387', 'grad_norm': '1.131', 'learning_rate': '9.553e-05', 'epoch': '1.23'}
+{'loss': '0.6585', 'grad_norm': '1.185', 'learning_rate': '9.552e-05', 'epoch': '1.23'}
+{'loss': '0.6284', 'grad_norm': '1.181', 'learning_rate': '9.551e-05', 'epoch': '1.231'}
+{'loss': '0.6183', 'grad_norm': '1.104', 'learning_rate': '9.55e-05', 'epoch': '1.232'}
+{'loss': '0.7267', 'grad_norm': '1.365', 'learning_rate': '9.549e-05', 'epoch': '1.233'}
+{'loss': '0.6581', 'grad_norm': '1.19', 'learning_rate': '9.548e-05', 'epoch': '1.234'}
+{'loss': '0.6698', 'grad_norm': '1.135', 'learning_rate': '9.547e-05', 'epoch': '1.235'}
+{'loss': '0.7303', 'grad_norm': '1.194', 'learning_rate': '9.545e-05', 'epoch': '1.236'}
+{'loss': '0.6867', 'grad_norm': '1.094', 'learning_rate': '9.544e-05', 'epoch': '1.236'}
+{'loss': '0.7355', 'grad_norm': '1.451', 'learning_rate': '9.543e-05', 'epoch': '1.237'}
+{'loss': '0.6316', 'grad_norm': '1.12', 'learning_rate': '9.542e-05', 'epoch': '1.238'}
+{'loss': '0.6638', 'grad_norm': '1.023', 'learning_rate': '9.541e-05', 'epoch': '1.239'}
+{'loss': '0.658', 'grad_norm': '1.187', 'learning_rate': '9.54e-05', 'epoch': '1.24'}
+{'loss': '0.7563', 'grad_norm': '1.327', 'learning_rate': '9.539e-05', 'epoch': '1.241'}
+{'loss': '0.6373', 'grad_norm': '1.007', 'learning_rate': '9.538e-05', 'epoch': '1.242'}
+{'loss': '0.6989', 'grad_norm': '1.421', 'learning_rate': '9.537e-05', 'epoch': '1.243'}
+{'loss': '0.759', 'grad_norm': '1.256', 'learning_rate': '9.536e-05', 'epoch': '1.243'}
+{'loss': '0.6652', 'grad_norm': '1.151', 'learning_rate': '9.535e-05', 'epoch': '1.244'}
+{'loss': '0.6511', 'grad_norm': '1.213', 'learning_rate': '9.534e-05', 'epoch': '1.245'}
+{'loss': '0.6406', 'grad_norm': '1.139', 'learning_rate': '9.533e-05', 'epoch': '1.246'}
+{'loss': '0.6819', 'grad_norm': '1.147', 'learning_rate': '9.532e-05', 'epoch': '1.247'}
+{'loss': '0.6334', 'grad_norm': '1.152', 'learning_rate': '9.531e-05', 'epoch': '1.248'}
+{'loss': '0.726', 'grad_norm': '1.463', 'learning_rate': '9.53e-05', 'epoch': '1.249'}
+{'loss': '0.6617', 'grad_norm': '1.045', 'learning_rate': '9.529e-05', 'epoch': '1.249'}
+{'loss': '0.5889', 'grad_norm': '1.053', 'learning_rate': '9.528e-05', 'epoch': '1.25'}
+{'loss': '0.6381', 'grad_norm': '1.13', 'learning_rate': '9.527e-05', 'epoch': '1.251'}
+{'eval_loss': '0.7182', 'eval_runtime': '840.8', 'eval_samples_per_second': '2.45', 'eval_steps_per_second': '2.45', 'epoch': '1.251'}
+{'loss': '0.6397', 'grad_norm': '1.019', 'learning_rate': '9.526e-05', 'epoch': '1.252'}
+{'loss': '0.7932', 'grad_norm': '1.39', 'learning_rate': '9.525e-05', 'epoch': '1.253'}
+{'loss': '0.6997', 'grad_norm': '1.258', 'learning_rate': '9.524e-05', 'epoch': '1.254'}
+{'loss': '0.6947', 'grad_norm': '1.34', 'learning_rate': '9.523e-05', 'epoch': '1.255'}
+{'loss': '0.6309', 'grad_norm': '1.172', 'learning_rate': '9.522e-05', 'epoch': '1.255'}
+{'loss': '0.6454', 'grad_norm': '1.21', 'learning_rate': '9.521e-05', 'epoch': '1.256'}
+{'loss': '0.6748', 'grad_norm': '1.233', 'learning_rate': '9.52e-05', 'epoch': '1.257'}
+{'loss': '0.6729', 'grad_norm': '1.179', 'learning_rate': '9.519e-05', 'epoch': '1.258'}
+{'loss': '0.6581', 'grad_norm': '1.078', 'learning_rate': '9.517e-05', 'epoch': '1.259'}
+{'loss': '0.6576', 'grad_norm': '1.112', 'learning_rate': '9.516e-05', 'epoch': '1.26'}
+{'loss': '0.7026', 'grad_norm': '1.338', 'learning_rate': '9.515e-05', 'epoch': '1.261'}
+{'loss': '0.6112', 'grad_norm': '1.128', 'learning_rate': '9.514e-05', 'epoch': '1.261'}
+{'loss': '0.6699', 'grad_norm': '1.206', 'learning_rate': '9.513e-05', 'epoch': '1.262'}
+{'loss': '0.6794', 'grad_norm': '1.221', 'learning_rate': '9.512e-05', 'epoch': '1.263'}
+{'loss': '0.7094', 'grad_norm': '1.167', 'learning_rate': '9.511e-05', 'epoch': '1.264'}
+{'loss': '0.6856', 'grad_norm': '1.335', 'learning_rate': '9.51e-05', 'epoch': '1.265'}
+{'loss': '0.6775', 'grad_norm': '1.186', 'learning_rate': '9.509e-05', 'epoch': '1.266'}
+{'loss': '0.6588', 'grad_norm': '1.393', 'learning_rate': '9.508e-05', 'epoch': '1.267'}
+{'loss': '0.6927', 'grad_norm': '1.377', 'learning_rate': '9.507e-05', 'epoch': '1.268'}
+{'loss': '0.6482', 'grad_norm': '1.035', 'learning_rate': '9.506e-05', 'epoch': '1.268'}
+{'loss': '0.5868', 'grad_norm': '1.113', 'learning_rate': '9.505e-05', 'epoch': '1.269'}
+{'loss': '0.6147', 'grad_norm': '1.077', 'learning_rate': '9.504e-05', 'epoch': '1.27'}
+{'loss': '0.6503', 'grad_norm': '1.101', 'learning_rate': '9.503e-05', 'epoch': '1.271'}
+{'loss': '0.6896', 'grad_norm': '1.198', 'learning_rate': '9.502e-05', 'epoch': '1.272'}
+{'loss': '0.678', 'grad_norm': '1.277', 'learning_rate': '9.5e-05', 'epoch': '1.273'}
+{'loss': '0.6906', 'grad_norm': '1.131', 'learning_rate': '9.499e-05', 'epoch': '1.274'}
+{'loss': '0.6419', 'grad_norm': '1.125', 'learning_rate': '9.498e-05', 'epoch': '1.274'}
+{'loss': '0.6908', 'grad_norm': '1.251', 'learning_rate': '9.497e-05', 'epoch': '1.275'}
+{'loss': '0.6457', 'grad_norm': '0.9559', 'learning_rate': '9.496e-05', 'epoch': '1.276'}
+{'loss': '0.6314', 'grad_norm': '1.178', 'learning_rate': '9.495e-05', 'epoch': '1.277'}
+{'loss': '0.5854', 'grad_norm': '0.9714', 'learning_rate': '9.494e-05', 'epoch': '1.278'}
+{'loss': '0.6966', 'grad_norm': '1.379', 'learning_rate': '9.493e-05', 'epoch': '1.279'}
+{'loss': '0.7282', 'grad_norm': '1.251', 'learning_rate': '9.492e-05', 'epoch': '1.28'}
+{'loss': '0.7323', 'grad_norm': '1.327', 'learning_rate': '9.491e-05', 'epoch': '1.28'}
+{'loss': '0.7875', 'grad_norm': '1.388', 'learning_rate': '9.49e-05', 'epoch': '1.281'}
+{'loss': '0.6017', 'grad_norm': '1.02', 'learning_rate': '9.489e-05', 'epoch': '1.282'}
+{'loss': '0.6256', 'grad_norm': '1.087', 'learning_rate': '9.488e-05', 'epoch': '1.283'}
+{'loss': '0.5967', 'grad_norm': '1.006', 'learning_rate': '9.486e-05', 'epoch': '1.284'}
+{'loss': '0.7196', 'grad_norm': '1.307', 'learning_rate': '9.485e-05', 'epoch': '1.285'}
+{'loss': '0.6655', 'grad_norm': '1.248', 'learning_rate': '9.484e-05', 'epoch': '1.286'}
+{'loss': '0.6799', 'grad_norm': '1.122', 'learning_rate': '9.483e-05', 'epoch': '1.287'}
+{'loss': '0.672', 'grad_norm': '1.137', 'learning_rate': '9.482e-05', 'epoch': '1.287'}
+{'loss': '0.6756', 'grad_norm': '1.208', 'learning_rate': '9.481e-05', 'epoch': '1.288'}
+{'loss': '0.7504', 'grad_norm': '1.33', 'learning_rate': '9.48e-05', 'epoch': '1.289'}
+{'loss': '0.6837', 'grad_norm': '1.09', 'learning_rate': '9.479e-05', 'epoch': '1.29'}
+{'loss': '0.6553', 'grad_norm': '1.102', 'learning_rate': '9.478e-05', 'epoch': '1.291'}
+{'loss': '0.616', 'grad_norm': '1.179', 'learning_rate': '9.477e-05', 'epoch': '1.292'}
+{'loss': '0.7406', 'grad_norm': '1.478', 'learning_rate': '9.476e-05', 'epoch': '1.293'}
+{'loss': '0.5951', 'grad_norm': '1.239', 'learning_rate': '9.475e-05', 'epoch': '1.293'}
+{'loss': '0.6587', 'grad_norm': '1.167', 'learning_rate': '9.473e-05', 'epoch': '1.294'}
+{'eval_loss': '0.7161', 'eval_runtime': '830.8', 'eval_samples_per_second': '2.479', 'eval_steps_per_second': '2.479', 'epoch': '1.294'}
+{'loss': '0.5803', 'grad_norm': '0.9834', 'learning_rate': '9.472e-05', 'epoch': '1.295'}
+{'loss': '0.6255', 'grad_norm': '1.047', 'learning_rate': '9.471e-05', 'epoch': '1.296'}
+{'loss': '0.7038', 'grad_norm': '1.276', 'learning_rate': '9.47e-05', 'epoch': '1.297'}
+{'loss': '0.7064', 'grad_norm': '1.257', 'learning_rate': '9.469e-05', 'epoch': '1.298'}
+{'loss': '0.6931', 'grad_norm': '1.234', 'learning_rate': '9.468e-05', 'epoch': '1.299'}
+{'loss': '0.666', 'grad_norm': '1.156', 'learning_rate': '9.467e-05', 'epoch': '1.299'}
+{'loss': '0.6523', 'grad_norm': '1.234', 'learning_rate': '9.466e-05', 'epoch': '1.3'}
+{'loss': '0.6305', 'grad_norm': '1.05', 'learning_rate': '9.465e-05', 'epoch': '1.301'}
+{'loss': '0.7271', 'grad_norm': '1.192', 'learning_rate': '9.464e-05', 'epoch': '1.302'}
+{'loss': '0.5725', 'grad_norm': '1.07', 'learning_rate': '9.462e-05', 'epoch': '1.303'}
+{'loss': '0.5726', 'grad_norm': '1.056', 'learning_rate': '9.461e-05', 'epoch': '1.304'}
+{'loss': '0.6499', 'grad_norm': '1.131', 'learning_rate': '9.46e-05', 'epoch': '1.305'}
+{'loss': '0.5473', 'grad_norm': '1.062', 'learning_rate': '9.459e-05', 'epoch': '1.306'}
+{'loss': '0.6977', 'grad_norm': '1.246', 'learning_rate': '9.458e-05', 'epoch': '1.306'}
+{'loss': '0.647', 'grad_norm': '1.097', 'learning_rate': '9.457e-05', 'epoch': '1.307'}
+{'loss': '0.6713', 'grad_norm': '1.156', 'learning_rate': '9.456e-05', 'epoch': '1.308'}
+{'loss': '0.6504', 'grad_norm': '1.037', 'learning_rate': '9.455e-05', 'epoch': '1.309'}
+{'loss': '0.7001', 'grad_norm': '1.349', 'learning_rate': '9.453e-05', 'epoch': '1.31'}
+{'loss': '0.6947', 'grad_norm': '1.244', 'learning_rate': '9.452e-05', 'epoch': '1.311'}
+{'loss': '0.5928', 'grad_norm': '1.073', 'learning_rate': '9.451e-05', 'epoch': '1.312'}
+{'loss': '0.659', 'grad_norm': '1.232', 'learning_rate': '9.45e-05', 'epoch': '1.312'}
+{'loss': '0.6516', 'grad_norm': '1.144', 'learning_rate': '9.449e-05', 'epoch': '1.313'}
+{'loss': '0.689', 'grad_norm': '1.342', 'learning_rate': '9.448e-05', 'epoch': '1.314'}
+{'loss': '0.7203', 'grad_norm': '1.413', 'learning_rate': '9.447e-05', 'epoch': '1.315'}
+{'loss': '0.6565', 'grad_norm': '1.141', 'learning_rate': '9.446e-05', 'epoch': '1.316'}
+{'loss': '0.6888', 'grad_norm': '1.09', 'learning_rate': '9.445e-05', 'epoch': '1.317'}
+{'loss': '0.615', 'grad_norm': '1.153', 'learning_rate': '9.443e-05', 'epoch': '1.318'}
+{'loss': '0.6446', 'grad_norm': '1.078', 'learning_rate': '9.442e-05', 'epoch': '1.318'}
+{'loss': '0.6724', 'grad_norm': '1.113', 'learning_rate': '9.441e-05', 'epoch': '1.319'}
+{'loss': '0.5901', 'grad_norm': '1.158', 'learning_rate': '9.44e-05', 'epoch': '1.32'}
+{'loss': '0.6817', 'grad_norm': '1.095', 'learning_rate': '9.439e-05', 'epoch': '1.321'}
+{'loss': '0.6884', 'grad_norm': '1.217', 'learning_rate': '9.438e-05', 'epoch': '1.322'}
+{'loss': '0.6829', 'grad_norm': '1.154', 'learning_rate': '9.437e-05', 'epoch': '1.323'}
+{'loss': '0.6087', 'grad_norm': '1.219', 'learning_rate': '9.436e-05', 'epoch': '1.324'}
+{'loss': '0.6001', 'grad_norm': '1.035', 'learning_rate': '9.434e-05', 'epoch': '1.324'}
+{'loss': '0.6812', 'grad_norm': '1.136', 'learning_rate': '9.433e-05', 'epoch': '1.325'}
+{'loss': '0.6667', 'grad_norm': '1.191', 'learning_rate': '9.432e-05', 'epoch': '1.326'}
+{'loss': '0.6456', 'grad_norm': '1.088', 'learning_rate': '9.431e-05', 'epoch': '1.327'}
+{'loss': '0.5861', 'grad_norm': '1.186', 'learning_rate': '9.43e-05', 'epoch': '1.328'}
+{'loss': '0.6174', 'grad_norm': '1.084', 'learning_rate': '9.429e-05', 'epoch': '1.329'}
+{'loss': '0.8176', 'grad_norm': '1.282', 'learning_rate': '9.428e-05', 'epoch': '1.33'}
+{'loss': '0.7455', 'grad_norm': '1.127', 'learning_rate': '9.426e-05', 'epoch': '1.331'}
+{'loss': '0.6253', 'grad_norm': '1.01', 'learning_rate': '9.425e-05', 'epoch': '1.331'}
+{'loss': '0.5977', 'grad_norm': '1.183', 'learning_rate': '9.424e-05', 'epoch': '1.332'}
+{'loss': '0.6251', 'grad_norm': '1.131', 'learning_rate': '9.423e-05', 'epoch': '1.333'}
+{'loss': '0.7469', 'grad_norm': '1.323', 'learning_rate': '9.422e-05', 'epoch': '1.334'}
+{'loss': '0.7013', 'grad_norm': '1.139', 'learning_rate': '9.421e-05', 'epoch': '1.335'}
+{'loss': '0.7113', 'grad_norm': '1.207', 'learning_rate': '9.42e-05', 'epoch': '1.336'}
+{'loss': '0.6219', 'grad_norm': '1.253', 'learning_rate': '9.418e-05', 'epoch': '1.337'}
+{'loss': '0.6905', 'grad_norm': '1.186', 'learning_rate': '9.417e-05', 'epoch': '1.337'}
+{'eval_loss': '0.7128', 'eval_runtime': '821.1', 'eval_samples_per_second': '2.509', 'eval_steps_per_second': '2.509', 'epoch': '1.337'}
+{'loss': '0.6473', 'grad_norm': '1.173', 'learning_rate': '9.416e-05', 'epoch': '1.338'}
+{'loss': '0.7377', 'grad_norm': '1.297', 'learning_rate': '9.415e-05', 'epoch': '1.339'}
+{'loss': '0.6842', 'grad_norm': '1.084', 'learning_rate': '9.414e-05', 'epoch': '1.34'}
+{'loss': '0.6623', 'grad_norm': '0.9301', 'learning_rate': '9.413e-05', 'epoch': '1.341'}
+{'loss': '0.6979', 'grad_norm': '1.138', 'learning_rate': '9.411e-05', 'epoch': '1.342'}
+{'loss': '0.7032', 'grad_norm': '1.156', 'learning_rate': '9.41e-05', 'epoch': '1.343'}
+{'loss': '0.6283', 'grad_norm': '1.148', 'learning_rate': '9.409e-05', 'epoch': '1.343'}
+{'loss': '0.6573', 'grad_norm': '1.093', 'learning_rate': '9.408e-05', 'epoch': '1.344'}
+{'loss': '0.7234', 'grad_norm': '1.363', 'learning_rate': '9.407e-05', 'epoch': '1.345'}
+{'loss': '0.6465', 'grad_norm': '1.399', 'learning_rate': '9.406e-05', 'epoch': '1.346'}
+{'loss': '0.6348', 'grad_norm': '1.186', 'learning_rate': '9.405e-05', 'epoch': '1.347'}
+{'loss': '0.6559', 'grad_norm': '1.172', 'learning_rate': '9.403e-05', 'epoch': '1.348'}
+{'loss': '0.6694', 'grad_norm': '1.176', 'learning_rate': '9.402e-05', 'epoch': '1.349'}
+{'loss': '0.6766', 'grad_norm': '1.127', 'learning_rate': '9.401e-05', 'epoch': '1.35'}
+{'loss': '0.7134', 'grad_norm': '1.252', 'learning_rate': '9.4e-05', 'epoch': '1.35'}
+{'loss': '0.6572', 'grad_norm': '1.158', 'learning_rate': '9.399e-05', 'epoch': '1.351'}
+{'loss': '0.639', 'grad_norm': '1.235', 'learning_rate': '9.398e-05', 'epoch': '1.352'}
+{'loss': '0.581', 'grad_norm': '0.9648', 'learning_rate': '9.396e-05', 'epoch': '1.353'}
+{'loss': '0.7157', 'grad_norm': '1.147', 'learning_rate': '9.395e-05', 'epoch': '1.354'}
+{'loss': '0.6749', 'grad_norm': '1.195', 'learning_rate': '9.394e-05', 'epoch': '1.355'}
+{'loss': '0.637', 'grad_norm': '1.095', 'learning_rate': '9.393e-05', 'epoch': '1.356'}
+{'loss': '0.6804', 'grad_norm': '1.065', 'learning_rate': '9.392e-05', 'epoch': '1.356'}
+{'loss': '0.7056', 'grad_norm': '1.398', 'learning_rate': '9.39e-05', 'epoch': '1.357'}
+{'loss': '0.7384', 'grad_norm': '1.161', 'learning_rate': '9.389e-05', 'epoch': '1.358'}
+{'loss': '0.734', 'grad_norm': '1.192', 'learning_rate': '9.388e-05', 'epoch': '1.359'}
+{'loss': '0.6564', 'grad_norm': '1.16', 'learning_rate': '9.387e-05', 'epoch': '1.36'}
+{'loss': '0.7393', 'grad_norm': '1.237', 'learning_rate': '9.386e-05', 'epoch': '1.361'}
+{'loss': '0.6795', 'grad_norm': '1.198', 'learning_rate': '9.385e-05', 'epoch': '1.362'}
+{'loss': '0.5906', 'grad_norm': '1.158', 'learning_rate': '9.383e-05', 'epoch': '1.362'}
+{'loss': '0.6814', 'grad_norm': '1.061', 'learning_rate': '9.382e-05', 'epoch': '1.363'}
+{'loss': '0.6393', 'grad_norm': '0.9798', 'learning_rate': '9.381e-05', 'epoch': '1.364'}
+{'loss': '0.6674', 'grad_norm': '1.102', 'learning_rate': '9.38e-05', 'epoch': '1.365'}
+{'loss': '0.6716', 'grad_norm': '1.111', 'learning_rate': '9.379e-05', 'epoch': '1.366'}
+{'loss': '0.6234', 'grad_norm': '0.9442', 'learning_rate': '9.378e-05', 'epoch': '1.367'}
+{'loss': '0.6559', 'grad_norm': '1.309', 'learning_rate': '9.376e-05', 'epoch': '1.368'}
+{'loss': '0.6432', 'grad_norm': '1.066', 'learning_rate': '9.375e-05', 'epoch': '1.369'}
+{'loss': '0.6542', 'grad_norm': '1.028', 'learning_rate': '9.374e-05', 'epoch': '1.369'}
+{'loss': '0.671', 'grad_norm': '1.165', 'learning_rate': '9.373e-05', 'epoch': '1.37'}
+{'loss': '0.6525', 'grad_norm': '1.107', 'learning_rate': '9.372e-05', 'epoch': '1.371'}
+{'loss': '0.6362', 'grad_norm': '1.094', 'learning_rate': '9.37e-05', 'epoch': '1.372'}
+{'loss': '0.6017', 'grad_norm': '1.175', 'learning_rate': '9.369e-05', 'epoch': '1.373'}
+{'loss': '0.6854', 'grad_norm': '1.361', 'learning_rate': '9.368e-05', 'epoch': '1.374'}
+{'loss': '0.6434', 'grad_norm': '1.225', 'learning_rate': '9.367e-05', 'epoch': '1.375'}
+{'loss': '0.5966', 'grad_norm': '1.115', 'learning_rate': '9.366e-05', 'epoch': '1.375'}
+{'loss': '0.6717', 'grad_norm': '1.267', 'learning_rate': '9.364e-05', 'epoch': '1.376'}
+{'loss': '0.6231', 'grad_norm': '1.322', 'learning_rate': '9.363e-05', 'epoch': '1.377'}
+{'loss': '0.6495', 'grad_norm': '1.298', 'learning_rate': '9.362e-05', 'epoch': '1.378'}
+{'loss': '0.6535', 'grad_norm': '1.154', 'learning_rate': '9.361e-05', 'epoch': '1.379'}
+{'loss': '0.5707', 'grad_norm': '1.129', 'learning_rate': '9.36e-05', 'epoch': '1.38'}
+{'loss': '0.6194', 'grad_norm': '1.062', 'learning_rate': '9.358e-05', 'epoch': '1.381'}
+{'eval_loss': '0.7104', 'eval_runtime': '831.4', 'eval_samples_per_second': '2.478', 'eval_steps_per_second': '2.478', 'epoch': '1.381'}
+{'loss': '0.6021', 'grad_norm': '1.172', 'learning_rate': '9.357e-05', 'epoch': '1.381'}
+{'loss': '0.7195', 'grad_norm': '1.273', 'learning_rate': '9.356e-05', 'epoch': '1.382'}
+{'loss': '0.6315', 'grad_norm': '1.02', 'learning_rate': '9.355e-05', 'epoch': '1.383'}
+{'loss': '0.7229', 'grad_norm': '1.362', 'learning_rate': '9.354e-05', 'epoch': '1.384'}
+{'loss': '0.7058', 'grad_norm': '1.129', 'learning_rate': '9.352e-05', 'epoch': '1.385'}
+{'loss': '0.6187', 'grad_norm': '1.216', 'learning_rate': '9.351e-05', 'epoch': '1.386'}
+{'loss': '0.5912', 'grad_norm': '1.166', 'learning_rate': '9.35e-05', 'epoch': '1.387'}
+{'loss': '0.6494', 'grad_norm': '1.215', 'learning_rate': '9.349e-05', 'epoch': '1.387'}
+{'loss': '0.6258', 'grad_norm': '1.018', 'learning_rate': '9.348e-05', 'epoch': '1.388'}
+{'loss': '0.7077', 'grad_norm': '1.202', 'learning_rate': '9.346e-05', 'epoch': '1.389'}
+{'loss': '0.5959', 'grad_norm': '1.073', 'learning_rate': '9.345e-05', 'epoch': '1.39'}
+{'loss': '0.5659', 'grad_norm': '1.088', 'learning_rate': '9.344e-05', 'epoch': '1.391'}
+{'loss': '0.637', 'grad_norm': '1.263', 'learning_rate': '9.343e-05', 'epoch': '1.392'}
+{'loss': '0.6405', 'grad_norm': '1.158', 'learning_rate': '9.341e-05', 'epoch': '1.393'}
+{'loss': '0.7039', 'grad_norm': '1.223', 'learning_rate': '9.34e-05', 'epoch': '1.394'}
+{'loss': '0.6461', 'grad_norm': '1.193', 'learning_rate': '9.339e-05', 'epoch': '1.394'}
+{'loss': '0.6539', 'grad_norm': '1.434', 'learning_rate': '9.338e-05', 'epoch': '1.395'}
+{'loss': '0.6789', 'grad_norm': '1.23', 'learning_rate': '9.337e-05', 'epoch': '1.396'}
+{'loss': '0.618', 'grad_norm': '1.103', 'learning_rate': '9.335e-05', 'epoch': '1.397'}
+{'loss': '0.7098', 'grad_norm': '1.188', 'learning_rate': '9.334e-05', 'epoch': '1.398'}
+{'loss': '0.6521', 'grad_norm': '1.151', 'learning_rate': '9.333e-05', 'epoch': '1.399'}
+{'loss': '0.6739', 'grad_norm': '1.272', 'learning_rate': '9.332e-05', 'epoch': '1.4'}
+{'loss': '0.6261', 'grad_norm': '1.153', 'learning_rate': '9.33e-05', 'epoch': '1.4'}
+{'loss': '0.7071', 'grad_norm': '1.172', 'learning_rate': '9.329e-05', 'epoch': '1.401'}
+{'loss': '0.634', 'grad_norm': '1.153', 'learning_rate': '9.328e-05', 'epoch': '1.402'}
+{'loss': '0.6572', 'grad_norm': '1.155', 'learning_rate': '9.327e-05', 'epoch': '1.403'}
+{'loss': '0.6182', 'grad_norm': '1.109', 'learning_rate': '9.326e-05', 'epoch': '1.404'}
+{'loss': '0.6559', 'grad_norm': '1.235', 'learning_rate': '9.324e-05', 'epoch': '1.405'}
+{'loss': '0.6407', 'grad_norm': '1.187', 'learning_rate': '9.323e-05', 'epoch': '1.406'}
+{'loss': '0.6869', 'grad_norm': '1.533', 'learning_rate': '9.322e-05', 'epoch': '1.406'}
+{'loss': '0.6806', 'grad_norm': '1.214', 'learning_rate': '9.321e-05', 'epoch': '1.407'}
+{'loss': '0.6193', 'grad_norm': '1.233', 'learning_rate': '9.319e-05', 'epoch': '1.408'}
+{'loss': '0.5869', 'grad_norm': '1.062', 'learning_rate': '9.318e-05', 'epoch': '1.409'}
+{'loss': '0.636', 'grad_norm': '1.092', 'learning_rate': '9.317e-05', 'epoch': '1.41'}
+{'loss': '0.6083', 'grad_norm': '1.025', 'learning_rate': '9.316e-05', 'epoch': '1.411'}
+{'loss': '0.6341', 'grad_norm': '1.253', 'learning_rate': '9.314e-05', 'epoch': '1.412'}
+{'loss': '0.6333', 'grad_norm': '0.9817', 'learning_rate': '9.313e-05', 'epoch': '1.413'}
+{'loss': '0.7257', 'grad_norm': '1.214', 'learning_rate': '9.312e-05', 'epoch': '1.413'}
+{'loss': '0.7063', 'grad_norm': '1.217', 'learning_rate': '9.311e-05', 'epoch': '1.414'}
+{'loss': '0.6192', 'grad_norm': '1.022', 'learning_rate': '9.309e-05', 'epoch': '1.415'}
+{'loss': '0.6522', 'grad_norm': '1.19', 'learning_rate': '9.308e-05', 'epoch': '1.416'}
+{'loss': '0.6472', 'grad_norm': '1.173', 'learning_rate': '9.307e-05', 'epoch': '1.417'}
+{'loss': '0.699', 'grad_norm': '1.235', 'learning_rate': '9.306e-05', 'epoch': '1.418'}
+{'loss': '0.6182', 'grad_norm': '1.173', 'learning_rate': '9.304e-05', 'epoch': '1.419'}
+{'loss': '0.7472', 'grad_norm': '1.366', 'learning_rate': '9.303e-05', 'epoch': '1.419'}
+{'loss': '0.5542', 'grad_norm': '1.087', 'learning_rate': '9.302e-05', 'epoch': '1.42'}
+{'loss': '0.6935', 'grad_norm': '1.211', 'learning_rate': '9.301e-05', 'epoch': '1.421'}
+{'loss': '0.6888', 'grad_norm': '1.178', 'learning_rate': '9.299e-05', 'epoch': '1.422'}
+{'loss': '0.6956', 'grad_norm': '1.3', 'learning_rate': '9.298e-05', 'epoch': '1.423'}
+{'loss': '0.6901', 'grad_norm': '1.166', 'learning_rate': '9.297e-05', 'epoch': '1.424'}
+{'eval_loss': '0.7086', 'eval_runtime': '838.9', 'eval_samples_per_second': '2.456', 'eval_steps_per_second': '2.456', 'epoch': '1.424'}
+{'loss': '0.6801', 'grad_norm': '1.163', 'learning_rate': '9.296e-05', 'epoch': '1.425'}
+{'loss': '0.6802', 'grad_norm': '1.336', 'learning_rate': '9.294e-05', 'epoch': '1.425'}
+{'loss': '0.6335', 'grad_norm': '1.091', 'learning_rate': '9.293e-05', 'epoch': '1.426'}
+{'loss': '0.6677', 'grad_norm': '1.187', 'learning_rate': '9.292e-05', 'epoch': '1.427'}
+{'loss': '0.6742', 'grad_norm': '1.212', 'learning_rate': '9.291e-05', 'epoch': '1.428'}
+{'loss': '0.7169', 'grad_norm': '1.239', 'learning_rate': '9.289e-05', 'epoch': '1.429'}
+{'loss': '0.7112', 'grad_norm': '1.209', 'learning_rate': '9.288e-05', 'epoch': '1.43'}
+{'loss': '0.691', 'grad_norm': '1.128', 'learning_rate': '9.287e-05', 'epoch': '1.431'}
+{'loss': '0.7288', 'grad_norm': '1.181', 'learning_rate': '9.286e-05', 'epoch': '1.431'}
+{'loss': '0.6918', 'grad_norm': '1.552', 'learning_rate': '9.284e-05', 'epoch': '1.432'}
+{'loss': '0.6757', 'grad_norm': '1.283', 'learning_rate': '9.283e-05', 'epoch': '1.433'}
+{'loss': '0.6005', 'grad_norm': '1.096', 'learning_rate': '9.282e-05', 'epoch': '1.434'}
+{'loss': '0.6577', 'grad_norm': '1.25', 'learning_rate': '9.281e-05', 'epoch': '1.435'}
+{'loss': '0.6687', 'grad_norm': '1.259', 'learning_rate': '9.279e-05', 'epoch': '1.436'}
+{'loss': '0.5858', 'grad_norm': '1.096', 'learning_rate': '9.278e-05', 'epoch': '1.437'}
+{'loss': '0.6662', 'grad_norm': '1.2', 'learning_rate': '9.277e-05', 'epoch': '1.438'}
+{'loss': '0.6922', 'grad_norm': '1.244', 'learning_rate': '9.275e-05', 'epoch': '1.438'}
+{'loss': '0.7506', 'grad_norm': '1.32', 'learning_rate': '9.274e-05', 'epoch': '1.439'}
+{'loss': '0.6817', 'grad_norm': '1.078', 'learning_rate': '9.273e-05', 'epoch': '1.44'}
+{'loss': '0.6164', 'grad_norm': '1.146', 'learning_rate': '9.272e-05', 'epoch': '1.441'}
+{'loss': '0.6625', 'grad_norm': '1.062', 'learning_rate': '9.27e-05', 'epoch': '1.442'}
+{'loss': '0.6803', 'grad_norm': '1.139', 'learning_rate': '9.269e-05', 'epoch': '1.443'}
+{'loss': '0.6429', 'grad_norm': '1.082', 'learning_rate': '9.268e-05', 'epoch': '1.444'}
+{'loss': '0.6854', 'grad_norm': '1.174', 'learning_rate': '9.266e-05', 'epoch': '1.444'}
+{'loss': '0.689', 'grad_norm': '1.133', 'learning_rate': '9.265e-05', 'epoch': '1.445'}
+{'loss': '0.6038', 'grad_norm': '1.061', 'learning_rate': '9.264e-05', 'epoch': '1.446'}
+{'loss': '0.6672', 'grad_norm': '1.158', 'learning_rate': '9.263e-05', 'epoch': '1.447'}
+{'loss': '0.7165', 'grad_norm': '1.188', 'learning_rate': '9.261e-05', 'epoch': '1.448'}
+{'loss': '0.7108', 'grad_norm': '1.209', 'learning_rate': '9.26e-05', 'epoch': '1.449'}
+{'loss': '0.7182', 'grad_norm': '1.111', 'learning_rate': '9.259e-05', 'epoch': '1.45'}
+{'loss': '0.645', 'grad_norm': '1.028', 'learning_rate': '9.257e-05', 'epoch': '1.45'}
+{'loss': '0.6895', 'grad_norm': '1.236', 'learning_rate': '9.256e-05', 'epoch': '1.451'}
+{'loss': '0.7243', 'grad_norm': '1.369', 'learning_rate': '9.255e-05', 'epoch': '1.452'}
+{'loss': '0.6178', 'grad_norm': '1.18', 'learning_rate': '9.254e-05', 'epoch': '1.453'}
+{'loss': '0.6604', 'grad_norm': '1.076', 'learning_rate': '9.252e-05', 'epoch': '1.454'}
+{'loss': '0.5989', 'grad_norm': '0.9496', 'learning_rate': '9.251e-05', 'epoch': '1.455'}
+{'loss': '0.5744', 'grad_norm': '1.133', 'learning_rate': '9.25e-05', 'epoch': '1.456'}
+{'loss': '0.5905', 'grad_norm': '1.114', 'learning_rate': '9.248e-05', 'epoch': '1.457'}
+{'loss': '0.6751', 'grad_norm': '1.24', 'learning_rate': '9.247e-05', 'epoch': '1.457'}
+{'loss': '0.6873', 'grad_norm': '1.285', 'learning_rate': '9.246e-05', 'epoch': '1.458'}
+{'loss': '0.6468', 'grad_norm': '1.054', 'learning_rate': '9.245e-05', 'epoch': '1.459'}
+{'loss': '0.6486', 'grad_norm': '1.096', 'learning_rate': '9.243e-05', 'epoch': '1.46'}
+{'loss': '0.7024', 'grad_norm': '1.649', 'learning_rate': '9.242e-05', 'epoch': '1.461'}
+{'loss': '0.7082', 'grad_norm': '1.302', 'learning_rate': '9.241e-05', 'epoch': '1.462'}
+{'loss': '0.5843', 'grad_norm': '0.9295', 'learning_rate': '9.239e-05', 'epoch': '1.463'}
+{'loss': '0.6355', 'grad_norm': '1.056', 'learning_rate': '9.238e-05', 'epoch': '1.463'}
+{'loss': '0.6341', 'grad_norm': '1.225', 'learning_rate': '9.237e-05', 'epoch': '1.464'}
+{'loss': '0.6406', 'grad_norm': '1.163', 'learning_rate': '9.235e-05', 'epoch': '1.465'}
+{'loss': '0.5885', 'grad_norm': '1.009', 'learning_rate': '9.234e-05', 'epoch': '1.466'}
+{'loss': '0.6533', 'grad_norm': '1.175', 'learning_rate': '9.233e-05', 'epoch': '1.467'}
+{'eval_loss': '0.7067', 'eval_runtime': '829.9', 'eval_samples_per_second': '2.482', 'eval_steps_per_second': '2.482', 'epoch': '1.467'}
+{'loss': '0.6505', 'grad_norm': '0.9843', 'learning_rate': '9.232e-05', 'epoch': '1.468'}
+{'loss': '0.5809', 'grad_norm': '1.126', 'learning_rate': '9.23e-05', 'epoch': '1.469'}
+{'loss': '0.6432', 'grad_norm': '1.399', 'learning_rate': '9.229e-05', 'epoch': '1.469'}
+{'loss': '0.6126', 'grad_norm': '1.145', 'learning_rate': '9.228e-05', 'epoch': '1.47'}
+{'loss': '0.6272', 'grad_norm': '1.097', 'learning_rate': '9.226e-05', 'epoch': '1.471'}
+{'loss': '0.5895', 'grad_norm': '1.055', 'learning_rate': '9.225e-05', 'epoch': '1.472'}
+{'loss': '0.6621', 'grad_norm': '1.329', 'learning_rate': '9.224e-05', 'epoch': '1.473'}
+{'loss': '0.6911', 'grad_norm': '1.031', 'learning_rate': '9.222e-05', 'epoch': '1.474'}
+{'loss': '0.5945', 'grad_norm': '1.104', 'learning_rate': '9.221e-05', 'epoch': '1.475'}
+{'loss': '0.6231', 'grad_norm': '1.156', 'learning_rate': '9.22e-05', 'epoch': '1.476'}
+{'loss': '0.6413', 'grad_norm': '1.05', 'learning_rate': '9.218e-05', 'epoch': '1.476'}
+{'loss': '0.6639', 'grad_norm': '1.086', 'learning_rate': '9.217e-05', 'epoch': '1.477'}
+{'loss': '0.6161', 'grad_norm': '1.068', 'learning_rate': '9.216e-05', 'epoch': '1.478'}
+{'loss': '0.6693', 'grad_norm': '1.096', 'learning_rate': '9.214e-05', 'epoch': '1.479'}
+{'loss': '0.6363', 'grad_norm': '1.071', 'learning_rate': '9.213e-05', 'epoch': '1.48'}
+{'loss': '0.7379', 'grad_norm': '1.201', 'learning_rate': '9.212e-05', 'epoch': '1.481'}
+{'loss': '0.6866', 'grad_norm': '1.199', 'learning_rate': '9.21e-05', 'epoch': '1.482'}
+{'loss': '0.6873', 'grad_norm': '1.289', 'learning_rate': '9.209e-05', 'epoch': '1.482'}
+{'loss': '0.625', 'grad_norm': '1.003', 'learning_rate': '9.208e-05', 'epoch': '1.483'}
+{'loss': '0.6065', 'grad_norm': '1.139', 'learning_rate': '9.207e-05', 'epoch': '1.484'}
+{'loss': '0.614', 'grad_norm': '1.037', 'learning_rate': '9.205e-05', 'epoch': '1.485'}
+{'loss': '0.6176', 'grad_norm': '1.121', 'learning_rate': '9.204e-05', 'epoch': '1.486'}
+{'loss': '0.6921', 'grad_norm': '1.244', 'learning_rate': '9.203e-05', 'epoch': '1.487'}
+{'loss': '0.686', 'grad_norm': '1.212', 'learning_rate': '9.201e-05', 'epoch': '1.488'}
+{'loss': '0.7358', 'grad_norm': '1.227', 'learning_rate': '9.2e-05', 'epoch': '1.488'}
+{'loss': '0.6879', 'grad_norm': '1.281', 'learning_rate': '9.199e-05', 'epoch': '1.489'}
+{'loss': '0.6996', 'grad_norm': '1.234', 'learning_rate': '9.197e-05', 'epoch': '1.49'}
+{'loss': '0.6423', 'grad_norm': '1.234', 'learning_rate': '9.196e-05', 'epoch': '1.491'}
+{'loss': '0.7513', 'grad_norm': '1.349', 'learning_rate': '9.195e-05', 'epoch': '1.492'}
+{'loss': '0.6667', 'grad_norm': '0.9914', 'learning_rate': '9.193e-05', 'epoch': '1.493'}
+{'loss': '0.6436', 'grad_norm': '1.146', 'learning_rate': '9.192e-05', 'epoch': '1.494'}
+{'loss': '0.6552', 'grad_norm': '1.03', 'learning_rate': '9.191e-05', 'epoch': '1.494'}
+{'loss': '0.6792', 'grad_norm': '1.285', 'learning_rate': '9.189e-05', 'epoch': '1.495'}
+{'loss': '0.7081', 'grad_norm': '1.146', 'learning_rate': '9.188e-05', 'epoch': '1.496'}
+{'loss': '0.6613', 'grad_norm': '1.256', 'learning_rate': '9.186e-05', 'epoch': '1.497'}
+{'loss': '0.6299', 'grad_norm': '1.239', 'learning_rate': '9.185e-05', 'epoch': '1.498'}
+{'loss': '0.5911', 'grad_norm': '1.135', 'learning_rate': '9.184e-05', 'epoch': '1.499'}
+{'loss': '0.7077', 'grad_norm': '1.174', 'learning_rate': '9.182e-05', 'epoch': '1.5'}
+{'loss': '0.6169', 'grad_norm': '1.114', 'learning_rate': '9.181e-05', 'epoch': '1.501'}
+{'loss': '0.6588', 'grad_norm': '1.146', 'learning_rate': '9.18e-05', 'epoch': '1.501'}
+{'loss': '0.7696', 'grad_norm': '1.395', 'learning_rate': '9.178e-05', 'epoch': '1.502'}
+{'loss': '0.7143', 'grad_norm': '1.327', 'learning_rate': '9.177e-05', 'epoch': '1.503'}
+{'loss': '0.7085', 'grad_norm': '1.186', 'learning_rate': '9.176e-05', 'epoch': '1.504'}
+{'loss': '0.6643', 'grad_norm': '1.107', 'learning_rate': '9.174e-05', 'epoch': '1.505'}
+{'loss': '0.6283', 'grad_norm': '1.088', 'learning_rate': '9.173e-05', 'epoch': '1.506'}
+{'loss': '0.6812', 'grad_norm': '1.265', 'learning_rate': '9.172e-05', 'epoch': '1.507'}
+{'loss': '0.6443', 'grad_norm': '1.154', 'learning_rate': '9.17e-05', 'epoch': '1.507'}
+{'loss': '0.7175', 'grad_norm': '1.118', 'learning_rate': '9.169e-05', 'epoch': '1.508'}
+{'loss': '0.6537', 'grad_norm': '1.192', 'learning_rate': '9.168e-05', 'epoch': '1.509'}
+{'loss': '0.7314', 'grad_norm': '1.33', 'learning_rate': '9.166e-05', 'epoch': '1.51'}
+{'eval_loss': '0.7048', 'eval_runtime': '821.9', 'eval_samples_per_second': '2.506', 'eval_steps_per_second': '2.506', 'epoch': '1.51'}
+{'loss': '0.6363', 'grad_norm': '1.148', 'learning_rate': '9.165e-05', 'epoch': '1.511'}
+{'loss': '0.602', 'grad_norm': '1.107', 'learning_rate': '9.164e-05', 'epoch': '1.512'}
+{'loss': '0.7054', 'grad_norm': '1.213', 'learning_rate': '9.162e-05', 'epoch': '1.513'}
+{'loss': '0.6994', 'grad_norm': '1.188', 'learning_rate': '9.161e-05', 'epoch': '1.513'}
+{'loss': '0.7001', 'grad_norm': '1.302', 'learning_rate': '9.159e-05', 'epoch': '1.514'}
+{'loss': '0.6847', 'grad_norm': '1.111', 'learning_rate': '9.158e-05', 'epoch': '1.515'}
+{'loss': '0.712', 'grad_norm': '1.235', 'learning_rate': '9.157e-05', 'epoch': '1.516'}
+{'loss': '0.6407', 'grad_norm': '1.079', 'learning_rate': '9.155e-05', 'epoch': '1.517'}
+{'loss': '0.6569', 'grad_norm': '1.081', 'learning_rate': '9.154e-05', 'epoch': '1.518'}
+{'loss': '0.6841', 'grad_norm': '1.245', 'learning_rate': '9.153e-05', 'epoch': '1.519'}
+{'loss': '0.6192', 'grad_norm': '1.16', 'learning_rate': '9.151e-05', 'epoch': '1.52'}
+{'loss': '0.606', 'grad_norm': '1.24', 'learning_rate': '9.15e-05', 'epoch': '1.52'}
+{'loss': '0.6064', 'grad_norm': '1.092', 'learning_rate': '9.149e-05', 'epoch': '1.521'}
+{'loss': '0.6508', 'grad_norm': '1.114', 'learning_rate': '9.147e-05', 'epoch': '1.522'}
+{'loss': '0.7071', 'grad_norm': '1.31', 'learning_rate': '9.146e-05', 'epoch': '1.523'}
+{'loss': '0.6298', 'grad_norm': '1.181', 'learning_rate': '9.144e-05', 'epoch': '1.524'}
+{'loss': '0.6589', 'grad_norm': '1.159', 'learning_rate': '9.143e-05', 'epoch': '1.525'}
+{'loss': '0.6461', 'grad_norm': '1.24', 'learning_rate': '9.142e-05', 'epoch': '1.526'}
+{'loss': '0.6757', 'grad_norm': '1.071', 'learning_rate': '9.14e-05', 'epoch': '1.526'}
+{'loss': '0.637', 'grad_norm': '1.243', 'learning_rate': '9.139e-05', 'epoch': '1.527'}
+{'loss': '0.6708', 'grad_norm': '1.376', 'learning_rate': '9.138e-05', 'epoch': '1.528'}
+{'loss': '0.6109', 'grad_norm': '1.119', 'learning_rate': '9.136e-05', 'epoch': '1.529'}
+{'loss': '0.5866', 'grad_norm': '1.051', 'learning_rate': '9.135e-05', 'epoch': '1.53'}
+{'loss': '0.6567', 'grad_norm': '1.113', 'learning_rate': '9.133e-05', 'epoch': '1.531'}
+{'loss': '0.7188', 'grad_norm': '1.236', 'learning_rate': '9.132e-05', 'epoch': '1.532'}
+{'loss': '0.7416', 'grad_norm': '1.152', 'learning_rate': '9.131e-05', 'epoch': '1.532'}
+{'loss': '0.6286', 'grad_norm': '1.149', 'learning_rate': '9.129e-05', 'epoch': '1.533'}
+{'loss': '0.692', 'grad_norm': '1.158', 'learning_rate': '9.128e-05', 'epoch': '1.534'}
+{'loss': '0.6787', 'grad_norm': '1.145', 'learning_rate': '9.126e-05', 'epoch': '1.535'}
+{'loss': '0.6719', 'grad_norm': '1.095', 'learning_rate': '9.125e-05', 'epoch': '1.536'}
+{'loss': '0.736', 'grad_norm': '1.237', 'learning_rate': '9.124e-05', 'epoch': '1.537'}
+{'loss': '0.6546', 'grad_norm': '1.142', 'learning_rate': '9.122e-05', 'epoch': '1.538'}
+{'loss': '0.7129', 'grad_norm': '1.007', 'learning_rate': '9.121e-05', 'epoch': '1.539'}
+{'loss': '0.7289', 'grad_norm': '1.289', 'learning_rate': '9.12e-05', 'epoch': '1.539'}
+{'loss': '0.6207', 'grad_norm': '1.195', 'learning_rate': '9.118e-05', 'epoch': '1.54'}
+{'loss': '0.6508', 'grad_norm': '1.191', 'learning_rate': '9.117e-05', 'epoch': '1.541'}
+{'loss': '0.6737', 'grad_norm': '1.17', 'learning_rate': '9.115e-05', 'epoch': '1.542'}
+{'loss': '0.6999', 'grad_norm': '1.287', 'learning_rate': '9.114e-05', 'epoch': '1.543'}
+{'loss': '0.6986', 'grad_norm': '1.176', 'learning_rate': '9.113e-05', 'epoch': '1.544'}
+{'loss': '0.696', 'grad_norm': '1.209', 'learning_rate': '9.111e-05', 'epoch': '1.545'}
+{'loss': '0.6336', 'grad_norm': '1.221', 'learning_rate': '9.11e-05', 'epoch': '1.545'}
+{'loss': '0.653', 'grad_norm': '1.161', 'learning_rate': '9.108e-05', 'epoch': '1.546'}
+{'loss': '0.5593', 'grad_norm': '1.148', 'learning_rate': '9.107e-05', 'epoch': '1.547'}
+{'loss': '0.6488', 'grad_norm': '1.085', 'learning_rate': '9.106e-05', 'epoch': '1.548'}
+{'loss': '0.5878', 'grad_norm': '1.18', 'learning_rate': '9.104e-05', 'epoch': '1.549'}
+{'loss': '0.5924', 'grad_norm': '1.03', 'learning_rate': '9.103e-05', 'epoch': '1.55'}
+{'loss': '0.6225', 'grad_norm': '1.525', 'learning_rate': '9.101e-05', 'epoch': '1.551'}
+{'loss': '0.5919', 'grad_norm': '1.321', 'learning_rate': '9.1e-05', 'epoch': '1.551'}
+{'loss': '0.6757', 'grad_norm': '1.252', 'learning_rate': '9.099e-05', 'epoch': '1.552'}
+{'loss': '0.7416', 'grad_norm': '1.243', 'learning_rate': '9.097e-05', 'epoch': '1.553'}
+{'eval_loss': '0.7018', 'eval_runtime': '836', 'eval_samples_per_second': '2.464', 'eval_steps_per_second': '2.464', 'epoch': '1.553'}
+{'loss': '0.5911', 'grad_norm': '1.053', 'learning_rate': '9.096e-05', 'epoch': '1.554'}
+{'loss': '0.5926', 'grad_norm': '1.117', 'learning_rate': '9.094e-05', 'epoch': '1.555'}
+{'loss': '0.6038', 'grad_norm': '1.113', 'learning_rate': '9.093e-05', 'epoch': '1.556'}
+{'loss': '0.6781', 'grad_norm': '1.196', 'learning_rate': '9.091e-05', 'epoch': '1.557'}
+{'loss': '0.7172', 'grad_norm': '1.167', 'learning_rate': '9.09e-05', 'epoch': '1.557'}
+{'loss': '0.6465', 'grad_norm': '1.281', 'learning_rate': '9.089e-05', 'epoch': '1.558'}
+{'loss': '0.7105', 'grad_norm': '1.224', 'learning_rate': '9.087e-05', 'epoch': '1.559'}
+{'loss': '0.6721', 'grad_norm': '1.187', 'learning_rate': '9.086e-05', 'epoch': '1.56'}
+{'loss': '0.6685', 'grad_norm': '1.136', 'learning_rate': '9.084e-05', 'epoch': '1.561'}
+{'loss': '0.6535', 'grad_norm': '1.11', 'learning_rate': '9.083e-05', 'epoch': '1.562'}
+{'loss': '0.6731', 'grad_norm': '1.334', 'learning_rate': '9.082e-05', 'epoch': '1.563'}
+{'loss': '0.643', 'grad_norm': '1.068', 'learning_rate': '9.08e-05', 'epoch': '1.564'}
+{'loss': '0.6946', 'grad_norm': '1.003', 'learning_rate': '9.079e-05', 'epoch': '1.564'}
+{'loss': '0.6931', 'grad_norm': '1.1', 'learning_rate': '9.077e-05', 'epoch': '1.565'}
+{'loss': '0.6183', 'grad_norm': '1.172', 'learning_rate': '9.076e-05', 'epoch': '1.566'}
+{'loss': '0.6307', 'grad_norm': '1.014', 'learning_rate': '9.074e-05', 'epoch': '1.567'}
+{'loss': '0.6049', 'grad_norm': '1.1', 'learning_rate': '9.073e-05', 'epoch': '1.568'}
+{'loss': '0.6583', 'grad_norm': '1.169', 'learning_rate': '9.072e-05', 'epoch': '1.569'}
+{'loss': '0.6347', 'grad_norm': '1.256', 'learning_rate': '9.07e-05', 'epoch': '1.57'}
+{'loss': '0.6214', 'grad_norm': '1.038', 'learning_rate': '9.069e-05', 'epoch': '1.57'}
+{'loss': '0.6746', 'grad_norm': '1.251', 'learning_rate': '9.067e-05', 'epoch': '1.571'}
+{'loss': '0.6409', 'grad_norm': '1.272', 'learning_rate': '9.066e-05', 'epoch': '1.572'}
+{'loss': '0.5886', 'grad_norm': '1.175', 'learning_rate': '9.064e-05', 'epoch': '1.573'}
+{'loss': '0.6473', 'grad_norm': '1.302', 'learning_rate': '9.063e-05', 'epoch': '1.574'}
+{'loss': '0.6977', 'grad_norm': '1.249', 'learning_rate': '9.062e-05', 'epoch': '1.575'}
+{'loss': '0.6403', 'grad_norm': '1.043', 'learning_rate': '9.06e-05', 'epoch': '1.576'}
+{'loss': '0.6726', 'grad_norm': '1.09', 'learning_rate': '9.059e-05', 'epoch': '1.576'}
+{'loss': '0.6861', 'grad_norm': '1.189', 'learning_rate': '9.057e-05', 'epoch': '1.577'}
+{'loss': '0.6786', 'grad_norm': '1.08', 'learning_rate': '9.056e-05', 'epoch': '1.578'}
+{'loss': '0.6757', 'grad_norm': '1.07', 'learning_rate': '9.054e-05', 'epoch': '1.579'}
+{'loss': '0.6596', 'grad_norm': '0.9469', 'learning_rate': '9.053e-05', 'epoch': '1.58'}
+{'loss': '0.6254', 'grad_norm': '0.9797', 'learning_rate': '9.052e-05', 'epoch': '1.581'}
+{'loss': '0.6464', 'grad_norm': '1.295', 'learning_rate': '9.05e-05', 'epoch': '1.582'}
+{'loss': '0.6497', 'grad_norm': '1.157', 'learning_rate': '9.049e-05', 'epoch': '1.583'}
+{'loss': '0.5988', 'grad_norm': '1.021', 'learning_rate': '9.047e-05', 'epoch': '1.583'}
+{'loss': '0.6392', 'grad_norm': '1.153', 'learning_rate': '9.046e-05', 'epoch': '1.584'}
+{'loss': '0.6197', 'grad_norm': '1.043', 'learning_rate': '9.044e-05', 'epoch': '1.585'}
+{'loss': '0.6203', 'grad_norm': '1.205', 'learning_rate': '9.043e-05', 'epoch': '1.586'}
+{'loss': '0.6752', 'grad_norm': '1.233', 'learning_rate': '9.041e-05', 'epoch': '1.587'}
+{'loss': '0.6619', 'grad_norm': '1.166', 'learning_rate': '9.04e-05', 'epoch': '1.588'}
+{'loss': '0.6121', 'grad_norm': '1.018', 'learning_rate': '9.039e-05', 'epoch': '1.589'}
+{'loss': '0.6162', 'grad_norm': '1.194', 'learning_rate': '9.037e-05', 'epoch': '1.589'}
+{'loss': '0.723', 'grad_norm': '1.5', 'learning_rate': '9.036e-05', 'epoch': '1.59'}
+{'loss': '0.6194', 'grad_norm': '1.048', 'learning_rate': '9.034e-05', 'epoch': '1.591'}
+{'loss': '0.7277', 'grad_norm': '1.385', 'learning_rate': '9.033e-05', 'epoch': '1.592'}
+{'loss': '0.6636', 'grad_norm': '1.265', 'learning_rate': '9.031e-05', 'epoch': '1.593'}
+{'loss': '0.6905', 'grad_norm': '1.12', 'learning_rate': '9.03e-05', 'epoch': '1.594'}
+{'loss': '0.6475', 'grad_norm': '1.082', 'learning_rate': '9.028e-05', 'epoch': '1.595'}
+{'loss': '0.6741', 'grad_norm': '1.174', 'learning_rate': '9.027e-05', 'epoch': '1.595'}
+{'loss': '0.5662', 'grad_norm': '1.197', 'learning_rate': '9.026e-05', 'epoch': '1.596'}
+{'eval_loss': '0.6999', 'eval_runtime': '836.5', 'eval_samples_per_second': '2.463', 'eval_steps_per_second': '2.463', 'epoch': '1.596'}
+{'loss': '0.6141', 'grad_norm': '0.9948', 'learning_rate': '9.024e-05', 'epoch': '1.597'}
+{'loss': '0.6011', 'grad_norm': '1.031', 'learning_rate': '9.023e-05', 'epoch': '1.598'}
+{'loss': '0.7672', 'grad_norm': '1.352', 'learning_rate': '9.021e-05', 'epoch': '1.599'}
+{'loss': '0.563', 'grad_norm': '1.032', 'learning_rate': '9.02e-05', 'epoch': '1.6'}
+{'loss': '0.7143', 'grad_norm': '1.348', 'learning_rate': '9.018e-05', 'epoch': '1.601'}
+{'loss': '0.6625', 'grad_norm': '1.062', 'learning_rate': '9.017e-05', 'epoch': '1.602'}
+{'loss': '0.6314', 'grad_norm': '0.9874', 'learning_rate': '9.015e-05', 'epoch': '1.602'}
+{'loss': '0.6468', 'grad_norm': '1.087', 'learning_rate': '9.014e-05', 'epoch': '1.603'}
+{'loss': '0.6354', 'grad_norm': '0.9799', 'learning_rate': '9.012e-05', 'epoch': '1.604'}
+{'loss': '0.5802', 'grad_norm': '1.384', 'learning_rate': '9.011e-05', 'epoch': '1.605'}
+{'loss': '0.614', 'grad_norm': '1.087', 'learning_rate': '9.009e-05', 'epoch': '1.606'}
+{'loss': '0.6562', 'grad_norm': '1.058', 'learning_rate': '9.008e-05', 'epoch': '1.607'}
+{'loss': '0.5996', 'grad_norm': '1.049', 'learning_rate': '9.006e-05', 'epoch': '1.608'}
+{'loss': '0.6296', 'grad_norm': '1.115', 'learning_rate': '9.005e-05', 'epoch': '1.608'}
+{'loss': '0.7051', 'grad_norm': '1.211', 'learning_rate': '9.004e-05', 'epoch': '1.609'}
+{'loss': '0.6324', 'grad_norm': '1.111', 'learning_rate': '9.002e-05', 'epoch': '1.61'}
+{'loss': '0.7022', 'grad_norm': '1.103', 'learning_rate': '9.001e-05', 'epoch': '1.611'}
+{'loss': '0.5095', 'grad_norm': '0.8861', 'learning_rate': '8.999e-05', 'epoch': '1.612'}
+{'loss': '0.6482', 'grad_norm': '1.187', 'learning_rate': '8.998e-05', 'epoch': '1.613'}
+{'loss': '0.6846', 'grad_norm': '0.9258', 'learning_rate': '8.996e-05', 'epoch': '1.614'}
+{'loss': '0.7081', 'grad_norm': '1.111', 'learning_rate': '8.995e-05', 'epoch': '1.614'}
+{'loss': '0.683', 'grad_norm': '1.019', 'learning_rate': '8.993e-05', 'epoch': '1.615'}
+{'loss': '0.6626', 'grad_norm': '1.079', 'learning_rate': '8.992e-05', 'epoch': '1.616'}
+{'loss': '0.6715', 'grad_norm': '1.07', 'learning_rate': '8.99e-05', 'epoch': '1.617'}
+{'loss': '0.6702', 'grad_norm': '1.118', 'learning_rate': '8.989e-05', 'epoch': '1.618'}
+{'loss': '0.6783', 'grad_norm': '1.354', 'learning_rate': '8.987e-05', 'epoch': '1.619'}
+{'loss': '0.6775', 'grad_norm': '1.202', 'learning_rate': '8.986e-05', 'epoch': '1.62'}
+{'loss': '0.6123', 'grad_norm': '1.222', 'learning_rate': '8.984e-05', 'epoch': '1.62'}
+{'loss': '0.6659', 'grad_norm': '1.096', 'learning_rate': '8.983e-05', 'epoch': '1.621'}
+{'loss': '0.6943', 'grad_norm': '1.04', 'learning_rate': '8.981e-05', 'epoch': '1.622'}
+{'loss': '0.6101', 'grad_norm': '1.111', 'learning_rate': '8.98e-05', 'epoch': '1.623'}
+{'loss': '0.7149', 'grad_norm': '1.248', 'learning_rate': '8.978e-05', 'epoch': '1.624'}
+{'loss': '0.6626', 'grad_norm': '1.333', 'learning_rate': '8.977e-05', 'epoch': '1.625'}
+{'loss': '0.6074', 'grad_norm': '1.2', 'learning_rate': '8.975e-05', 'epoch': '1.626'}
+{'loss': '0.5862', 'grad_norm': '0.8741', 'learning_rate': '8.974e-05', 'epoch': '1.627'}
+{'loss': '0.7104', 'grad_norm': '1.07', 'learning_rate': '8.972e-05', 'epoch': '1.627'}
+{'loss': '0.6528', 'grad_norm': '1.27', 'learning_rate': '8.971e-05', 'epoch': '1.628'}
+{'loss': '0.6303', 'grad_norm': '1.205', 'learning_rate': '8.969e-05', 'epoch': '1.629'}
+{'loss': '0.5943', 'grad_norm': '1.218', 'learning_rate': '8.968e-05', 'epoch': '1.63'}
+{'loss': '0.5667', 'grad_norm': '0.9873', 'learning_rate': '8.966e-05', 'epoch': '1.631'}
+{'loss': '0.7656', 'grad_norm': '1.43', 'learning_rate': '8.965e-05', 'epoch': '1.632'}
+{'loss': '0.7027', 'grad_norm': '1.123', 'learning_rate': '8.963e-05', 'epoch': '1.633'}
+{'loss': '0.6778', 'grad_norm': '1.252', 'learning_rate': '8.962e-05', 'epoch': '1.633'}
+{'loss': '0.5866', 'grad_norm': '0.9574', 'learning_rate': '8.96e-05', 'epoch': '1.634'}
+{'loss': '0.6245', 'grad_norm': '1.047', 'learning_rate': '8.959e-05', 'epoch': '1.635'}
+{'loss': '0.638', 'grad_norm': '1.072', 'learning_rate': '8.957e-05', 'epoch': '1.636'}
+{'loss': '0.6252', 'grad_norm': '1.159', 'learning_rate': '8.956e-05', 'epoch': '1.637'}
+{'loss': '0.7554', 'grad_norm': '1.341', 'learning_rate': '8.954e-05', 'epoch': '1.638'}
+{'loss': '0.6711', 'grad_norm': '1.27', 'learning_rate': '8.953e-05', 'epoch': '1.639'}
+{'loss': '0.7102', 'grad_norm': '1.433', 'learning_rate': '8.951e-05', 'epoch': '1.639'}
+{'eval_loss': '0.6975', 'eval_runtime': '823.7', 'eval_samples_per_second': '2.501', 'eval_steps_per_second': '2.501', 'epoch': '1.639'}
+{'loss': '0.6029', 'grad_norm': '0.9819', 'learning_rate': '8.95e-05', 'epoch': '1.64'}
+{'loss': '0.6493', 'grad_norm': '1.078', 'learning_rate': '8.948e-05', 'epoch': '1.641'}
+{'loss': '0.6424', 'grad_norm': '1.21', 'learning_rate': '8.947e-05', 'epoch': '1.642'}
+{'loss': '0.6752', 'grad_norm': '1.17', 'learning_rate': '8.945e-05', 'epoch': '1.643'}
+{'loss': '0.617', 'grad_norm': '1.652', 'learning_rate': '8.944e-05', 'epoch': '1.644'}
+{'loss': '0.6578', 'grad_norm': '1.115', 'learning_rate': '8.942e-05', 'epoch': '1.645'}
+{'loss': '0.6534', 'grad_norm': '1.175', 'learning_rate': '8.941e-05', 'epoch': '1.646'}
+{'loss': '0.6356', 'grad_norm': '1.19', 'learning_rate': '8.939e-05', 'epoch': '1.646'}
+{'loss': '0.6097', 'grad_norm': '1.084', 'learning_rate': '8.938e-05', 'epoch': '1.647'}
+{'loss': '0.6339', 'grad_norm': '1.193', 'learning_rate': '8.936e-05', 'epoch': '1.648'}
+{'loss': '0.6589', 'grad_norm': '1.498', 'learning_rate': '8.935e-05', 'epoch': '1.649'}
+{'loss': '0.6418', 'grad_norm': '1.251', 'learning_rate': '8.933e-05', 'epoch': '1.65'}
+{'loss': '0.6476', 'grad_norm': '1.3', 'learning_rate': '8.932e-05', 'epoch': '1.651'}
+{'loss': '0.7352', 'grad_norm': '1.202', 'learning_rate': '8.93e-05', 'epoch': '1.652'}
+{'loss': '0.6268', 'grad_norm': '1.071', 'learning_rate': '8.929e-05', 'epoch': '1.652'}
+{'loss': '0.7097', 'grad_norm': '1.354', 'learning_rate': '8.927e-05', 'epoch': '1.653'}
+{'loss': '0.6013', 'grad_norm': '1.181', 'learning_rate': '8.926e-05', 'epoch': '1.654'}
+{'loss': '0.6258', 'grad_norm': '1.065', 'learning_rate': '8.924e-05', 'epoch': '1.655'}
+{'loss': '0.7109', 'grad_norm': '1.126', 'learning_rate': '8.923e-05', 'epoch': '1.656'}
+{'loss': '0.6432', 'grad_norm': '1.452', 'learning_rate': '8.921e-05', 'epoch': '1.657'}
+{'loss': '0.5963', 'grad_norm': '1.04', 'learning_rate': '8.92e-05', 'epoch': '1.658'}
+{'loss': '0.6397', 'grad_norm': '1.009', 'learning_rate': '8.918e-05', 'epoch': '1.658'}
+{'loss': '0.7058', 'grad_norm': '1.223', 'learning_rate': '8.917e-05', 'epoch': '1.659'}
+{'loss': '0.6138', 'grad_norm': '1.118', 'learning_rate': '8.915e-05', 'epoch': '1.66'}
+{'loss': '0.639', 'grad_norm': '1.124', 'learning_rate': '8.914e-05', 'epoch': '1.661'}
+{'loss': '0.7189', 'grad_norm': '1.148', 'learning_rate': '8.912e-05', 'epoch': '1.662'}
+{'loss': '0.6716', 'grad_norm': '1.312', 'learning_rate': '8.911e-05', 'epoch': '1.663'}
+{'loss': '0.6695', 'grad_norm': '1.003', 'learning_rate': '8.909e-05', 'epoch': '1.664'}
+{'loss': '0.6557', 'grad_norm': '1.243', 'learning_rate': '8.907e-05', 'epoch': '1.665'}
+{'loss': '0.6126', 'grad_norm': '1.147', 'learning_rate': '8.906e-05', 'epoch': '1.665'}
+{'loss': '0.7428', 'grad_norm': '1.276', 'learning_rate': '8.904e-05', 'epoch': '1.666'}
+{'loss': '0.6507', 'grad_norm': '1.061', 'learning_rate': '8.903e-05', 'epoch': '1.667'}
+{'loss': '0.6939', 'grad_norm': '1.161', 'learning_rate': '8.901e-05', 'epoch': '1.668'}
+{'loss': '0.6331', 'grad_norm': '1.252', 'learning_rate': '8.9e-05', 'epoch': '1.669'}
+{'loss': '0.6425', 'grad_norm': '0.9492', 'learning_rate': '8.898e-05', 'epoch': '1.67'}
+{'loss': '0.5937', 'grad_norm': '0.9616', 'learning_rate': '8.897e-05', 'epoch': '1.671'}
+{'loss': '0.7523', 'grad_norm': '1.199', 'learning_rate': '8.895e-05', 'epoch': '1.671'}
+{'loss': '0.6036', 'grad_norm': '1.115', 'learning_rate': '8.894e-05', 'epoch': '1.672'}
+{'loss': '0.6141', 'grad_norm': '0.9545', 'learning_rate': '8.892e-05', 'epoch': '1.673'}
+{'loss': '0.6016', 'grad_norm': '1.294', 'learning_rate': '8.891e-05', 'epoch': '1.674'}
+{'loss': '0.6011', 'grad_norm': '1.061', 'learning_rate': '8.889e-05', 'epoch': '1.675'}
+{'loss': '0.6256', 'grad_norm': '1.027', 'learning_rate': '8.887e-05', 'epoch': '1.676'}
+{'loss': '0.6119', 'grad_norm': '1.096', 'learning_rate': '8.886e-05', 'epoch': '1.677'}
+{'loss': '0.6526', 'grad_norm': '1.186', 'learning_rate': '8.884e-05', 'epoch': '1.677'}
+{'loss': '0.5967', 'grad_norm': '1.126', 'learning_rate': '8.883e-05', 'epoch': '1.678'}
+{'loss': '0.6421', 'grad_norm': '1.033', 'learning_rate': '8.881e-05', 'epoch': '1.679'}
+{'loss': '0.6055', 'grad_norm': '1.193', 'learning_rate': '8.88e-05', 'epoch': '1.68'}
+{'loss': '0.7106', 'grad_norm': '1.289', 'learning_rate': '8.878e-05', 'epoch': '1.681'}
+{'loss': '0.7001', 'grad_norm': '1.148', 'learning_rate': '8.877e-05', 'epoch': '1.682'}
+{'loss': '0.6607', 'grad_norm': '1.291', 'learning_rate': '8.875e-05', 'epoch': '1.683'}
+{'eval_loss': '0.6949', 'eval_runtime': '817.4', 'eval_samples_per_second': '2.52', 'eval_steps_per_second': '2.52', 'epoch': '1.683'}
+{'loss': '0.6463', 'grad_norm': '1.002', 'learning_rate': '8.874e-05', 'epoch': '1.683'}
+{'loss': '0.6628', 'grad_norm': '1.222', 'learning_rate': '8.872e-05', 'epoch': '1.684'}
+{'loss': '0.6311', 'grad_norm': '1.048', 'learning_rate': '8.87e-05', 'epoch': '1.685'}
+{'loss': '0.6916', 'grad_norm': '1.178', 'learning_rate': '8.869e-05', 'epoch': '1.686'}
+{'loss': '0.6965', 'grad_norm': '6.57', 'learning_rate': '8.867e-05', 'epoch': '1.687'}
+{'loss': '0.6246', 'grad_norm': '1.1', 'learning_rate': '8.866e-05', 'epoch': '1.688'}
+{'loss': '0.6179', 'grad_norm': '1.223', 'learning_rate': '8.864e-05', 'epoch': '1.689'}
+{'loss': '0.6599', 'grad_norm': '1.235', 'learning_rate': '8.863e-05', 'epoch': '1.69'}
+{'loss': '0.6504', 'grad_norm': '1.254', 'learning_rate': '8.861e-05', 'epoch': '1.69'}
+{'loss': '0.6858', 'grad_norm': '1.177', 'learning_rate': '8.86e-05', 'epoch': '1.691'}
+{'loss': '0.6795', 'grad_norm': '1.346', 'learning_rate': '8.858e-05', 'epoch': '1.692'}
+{'loss': '0.6901', 'grad_norm': '1.301', 'learning_rate': '8.856e-05', 'epoch': '1.693'}
+{'loss': '0.6756', 'grad_norm': '0.9616', 'learning_rate': '8.855e-05', 'epoch': '1.694'}
+{'loss': '0.669', 'grad_norm': '0.9635', 'learning_rate': '8.853e-05', 'epoch': '1.695'}
+{'loss': '0.6413', 'grad_norm': '1.292', 'learning_rate': '8.852e-05', 'epoch': '1.696'}
+{'loss': '0.6251', 'grad_norm': '1.028', 'learning_rate': '8.85e-05', 'epoch': '1.696'}
+{'loss': '0.6546', 'grad_norm': '1.322', 'learning_rate': '8.849e-05', 'epoch': '1.697'}
+{'loss': '0.6467', 'grad_norm': '1.152', 'learning_rate': '8.847e-05', 'epoch': '1.698'}
+{'loss': '0.6798', 'grad_norm': '1.132', 'learning_rate': '8.845e-05', 'epoch': '1.699'}
+{'loss': '0.6547', 'grad_norm': '0.9735', 'learning_rate': '8.844e-05', 'epoch': '1.7'}
+{'loss': '0.6207', 'grad_norm': '1.099', 'learning_rate': '8.842e-05', 'epoch': '1.701'}
+{'loss': '0.6114', 'grad_norm': '1.111', 'learning_rate': '8.841e-05', 'epoch': '1.702'}
+{'loss': '0.6246', 'grad_norm': '1.172', 'learning_rate': '8.839e-05', 'epoch': '1.702'}
+{'loss': '0.7028', 'grad_norm': '1.25', 'learning_rate': '8.838e-05', 'epoch': '1.703'}
+{'loss': '0.7121', 'grad_norm': '1.248', 'learning_rate': '8.836e-05', 'epoch': '1.704'}
+{'loss': '0.6521', 'grad_norm': '1.049', 'learning_rate': '8.834e-05', 'epoch': '1.705'}
+{'loss': '0.6736', 'grad_norm': '1.169', 'learning_rate': '8.833e-05', 'epoch': '1.706'}
+{'loss': '0.653', 'grad_norm': '1.423', 'learning_rate': '8.831e-05', 'epoch': '1.707'}
+{'loss': '0.6858', 'grad_norm': '1.224', 'learning_rate': '8.83e-05', 'epoch': '1.708'}
+{'loss': '0.6567', 'grad_norm': '1.133', 'learning_rate': '8.828e-05', 'epoch': '1.709'}
+{'loss': '0.6207', 'grad_norm': '1.244', 'learning_rate': '8.827e-05', 'epoch': '1.709'}
+{'loss': '0.6192', 'grad_norm': '0.9953', 'learning_rate': '8.825e-05', 'epoch': '1.71'}
+{'loss': '0.6486', 'grad_norm': '1.065', 'learning_rate': '8.823e-05', 'epoch': '1.711'}
+{'loss': '0.6591', 'grad_norm': '1.204', 'learning_rate': '8.822e-05', 'epoch': '1.712'}
+{'loss': '0.6428', 'grad_norm': '1.235', 'learning_rate': '8.82e-05', 'epoch': '1.713'}
+{'loss': '0.5952', 'grad_norm': '1.072', 'learning_rate': '8.819e-05', 'epoch': '1.714'}
+{'loss': '0.6809', 'grad_norm': '1.181', 'learning_rate': '8.817e-05', 'epoch': '1.715'}
+{'loss': '0.6928', 'grad_norm': '1.225', 'learning_rate': '8.815e-05', 'epoch': '1.715'}
+{'loss': '0.6451', 'grad_norm': '1.141', 'learning_rate': '8.814e-05', 'epoch': '1.716'}
+{'loss': '0.6776', 'grad_norm': '1.075', 'learning_rate': '8.812e-05', 'epoch': '1.717'}
+{'loss': '0.6083', 'grad_norm': '1.14', 'learning_rate': '8.811e-05', 'epoch': '1.718'}
+{'loss': '0.6711', 'grad_norm': '1.316', 'learning_rate': '8.809e-05', 'epoch': '1.719'}
+{'loss': '0.6819', 'grad_norm': '1.037', 'learning_rate': '8.808e-05', 'epoch': '1.72'}
+{'loss': '0.6347', 'grad_norm': '1.25', 'learning_rate': '8.806e-05', 'epoch': '1.721'}
+{'loss': '0.655', 'grad_norm': '1.089', 'learning_rate': '8.804e-05', 'epoch': '1.721'}
+{'loss': '0.6257', 'grad_norm': '1.296', 'learning_rate': '8.803e-05', 'epoch': '1.722'}
+{'loss': '0.5878', 'grad_norm': '1.009', 'learning_rate': '8.801e-05', 'epoch': '1.723'}
+{'loss': '0.6441', 'grad_norm': '1.07', 'learning_rate': '8.8e-05', 'epoch': '1.724'}
+{'loss': '0.6271', 'grad_norm': '1.115', 'learning_rate': '8.798e-05', 'epoch': '1.725'}
+{'loss': '0.6463', 'grad_norm': '1.067', 'learning_rate': '8.796e-05', 'epoch': '1.726'}
+{'eval_loss': '0.6941', 'eval_runtime': '498.1', 'eval_samples_per_second': '4.135', 'eval_steps_per_second': '4.135', 'epoch': '1.726'}
+{'loss': '0.6483', 'grad_norm': '1.129', 'learning_rate': '8.795e-05', 'epoch': '1.727'}
+{'loss': '0.6335', 'grad_norm': '1.075', 'learning_rate': '8.793e-05', 'epoch': '1.728'}
+{'loss': '0.6605', 'grad_norm': '1.028', 'learning_rate': '8.792e-05', 'epoch': '1.728'}
+{'loss': '0.6033', 'grad_norm': '1.078', 'learning_rate': '8.79e-05', 'epoch': '1.729'}
+{'loss': '0.6361', 'grad_norm': '1.251', 'learning_rate': '8.788e-05', 'epoch': '1.73'}
+{'loss': '0.5815', 'grad_norm': '1.091', 'learning_rate': '8.787e-05', 'epoch': '1.731'}
+{'loss': '0.7229', 'grad_norm': '1.122', 'learning_rate': '8.785e-05', 'epoch': '1.732'}
+{'loss': '0.654', 'grad_norm': '1.244', 'learning_rate': '8.784e-05', 'epoch': '1.733'}
+{'loss': '0.6483', 'grad_norm': '1.158', 'learning_rate': '8.782e-05', 'epoch': '1.734'}
+{'loss': '0.6999', 'grad_norm': '1.288', 'learning_rate': '8.78e-05', 'epoch': '1.734'}
+{'loss': '0.6012', 'grad_norm': '1.068', 'learning_rate': '8.779e-05', 'epoch': '1.735'}
+{'loss': '0.7336', 'grad_norm': '1.362', 'learning_rate': '8.777e-05', 'epoch': '1.736'}
+{'loss': '0.6371', 'grad_norm': '1.172', 'learning_rate': '8.775e-05', 'epoch': '1.737'}
+{'loss': '0.6756', 'grad_norm': '1.18', 'learning_rate': '8.774e-05', 'epoch': '1.738'}
+{'loss': '0.6209', 'grad_norm': '1.126', 'learning_rate': '8.772e-05', 'epoch': '1.739'}
+{'loss': '0.6727', 'grad_norm': '1.289', 'learning_rate': '8.771e-05', 'epoch': '1.74'}
+{'loss': '0.6565', 'grad_norm': '1.141', 'learning_rate': '8.769e-05', 'epoch': '1.74'}
+{'loss': '0.7123', 'grad_norm': '1.076', 'learning_rate': '8.767e-05', 'epoch': '1.741'}
+{'loss': '0.6391', 'grad_norm': '1.194', 'learning_rate': '8.766e-05', 'epoch': '1.742'}
+{'loss': '0.7138', 'grad_norm': '1.23', 'learning_rate': '8.764e-05', 'epoch': '1.743'}
+{'loss': '0.6011', 'grad_norm': '1.32', 'learning_rate': '8.763e-05', 'epoch': '1.744'}
+{'loss': '0.6666', 'grad_norm': '1.232', 'learning_rate': '8.761e-05', 'epoch': '1.745'}
+{'loss': '0.6443', 'grad_norm': '1.084', 'learning_rate': '8.759e-05', 'epoch': '1.746'}
+{'loss': '0.5871', 'grad_norm': '1.009', 'learning_rate': '8.758e-05', 'epoch': '1.746'}
+{'loss': '0.5936', 'grad_norm': '0.9774', 'learning_rate': '8.756e-05', 'epoch': '1.747'}
+{'loss': '0.6012', 'grad_norm': '1.063', 'learning_rate': '8.754e-05', 'epoch': '1.748'}
+{'loss': '0.6435', 'grad_norm': '1.205', 'learning_rate': '8.753e-05', 'epoch': '1.749'}
+{'loss': '0.6645', 'grad_norm': '1.212', 'learning_rate': '8.751e-05', 'epoch': '1.75'}
+{'loss': '0.649', 'grad_norm': '1.186', 'learning_rate': '8.75e-05', 'epoch': '1.751'}
+{'loss': '0.6018', 'grad_norm': '1.075', 'learning_rate': '8.748e-05', 'epoch': '1.752'}
+{'loss': '0.6242', 'grad_norm': '1.077', 'learning_rate': '8.746e-05', 'epoch': '1.753'}
+{'loss': '0.6011', 'grad_norm': '1.071', 'learning_rate': '8.745e-05', 'epoch': '1.753'}
+{'loss': '0.666', 'grad_norm': '1.164', 'learning_rate': '8.743e-05', 'epoch': '1.754'}
+{'loss': '0.6796', 'grad_norm': '1.228', 'learning_rate': '8.741e-05', 'epoch': '1.755'}
+{'loss': '0.6356', 'grad_norm': '1.072', 'learning_rate': '8.74e-05', 'epoch': '1.756'}
+{'loss': '0.5904', 'grad_norm': '1.138', 'learning_rate': '8.738e-05', 'epoch': '1.757'}
+{'loss': '0.6036', 'grad_norm': '1.169', 'learning_rate': '8.737e-05', 'epoch': '1.758'}
+{'loss': '0.6732', 'grad_norm': '1.172', 'learning_rate': '8.735e-05', 'epoch': '1.759'}
+{'loss': '0.6789', 'grad_norm': '1.09', 'learning_rate': '8.733e-05', 'epoch': '1.759'}
+{'loss': '0.7186', 'grad_norm': '1.405', 'learning_rate': '8.732e-05', 'epoch': '1.76'}
+{'loss': '0.7305', 'grad_norm': '1.288', 'learning_rate': '8.73e-05', 'epoch': '1.761'}
+{'loss': '0.6772', 'grad_norm': '1.229', 'learning_rate': '8.728e-05', 'epoch': '1.762'}
+{'loss': '0.6883', 'grad_norm': '1.206', 'learning_rate': '8.727e-05', 'epoch': '1.763'}
+{'loss': '0.7277', 'grad_norm': '1.345', 'learning_rate': '8.725e-05', 'epoch': '1.764'}
+{'loss': '0.6123', 'grad_norm': '1.089', 'learning_rate': '8.724e-05', 'epoch': '1.765'}
+{'loss': '0.6186', 'grad_norm': '1.069', 'learning_rate': '8.722e-05', 'epoch': '1.765'}
+{'loss': '0.715', 'grad_norm': '1.337', 'learning_rate': '8.72e-05', 'epoch': '1.766'}
+{'loss': '0.7331', 'grad_norm': '1.254', 'learning_rate': '8.719e-05', 'epoch': '1.767'}
+{'loss': '0.6567', 'grad_norm': '1.165', 'learning_rate': '8.717e-05', 'epoch': '1.768'}
+{'loss': '0.6126', 'grad_norm': '1.033', 'learning_rate': '8.715e-05', 'epoch': '1.769'}
+{'eval_loss': '0.6929', 'eval_runtime': '497.8', 'eval_samples_per_second': '4.138', 'eval_steps_per_second': '4.138', 'epoch': '1.769'}
+{'loss': '0.6242', 'grad_norm': '1.115', 'learning_rate': '8.714e-05', 'epoch': '1.77'}
+{'loss': '0.6431', 'grad_norm': '1.257', 'learning_rate': '8.712e-05', 'epoch': '1.771'}
+{'loss': '0.6322', 'grad_norm': '1.407', 'learning_rate': '8.71e-05', 'epoch': '1.772'}
+{'loss': '0.6277', 'grad_norm': '1.172', 'learning_rate': '8.709e-05', 'epoch': '1.772'}
+{'loss': '0.6275', 'grad_norm': '1.097', 'learning_rate': '8.707e-05', 'epoch': '1.773'}
+{'loss': '0.5985', 'grad_norm': '1.029', 'learning_rate': '8.705e-05', 'epoch': '1.774'}
+{'loss': '0.6176', 'grad_norm': '1.168', 'learning_rate': '8.704e-05', 'epoch': '1.775'}
+{'loss': '0.6356', 'grad_norm': '1.373', 'learning_rate': '8.702e-05', 'epoch': '1.776'}
+{'loss': '0.5878', 'grad_norm': '1.129', 'learning_rate': '8.7e-05', 'epoch': '1.777'}
+{'loss': '0.6119', 'grad_norm': '1.087', 'learning_rate': '8.699e-05', 'epoch': '1.778'}
+{'loss': '0.7158', 'grad_norm': '1.043', 'learning_rate': '8.697e-05', 'epoch': '1.778'}
+{'loss': '0.6831', 'grad_norm': '1.032', 'learning_rate': '8.696e-05', 'epoch': '1.779'}
+{'loss': '0.7015', 'grad_norm': '1.348', 'learning_rate': '8.694e-05', 'epoch': '1.78'}
+{'loss': '0.5946', 'grad_norm': '0.9731', 'learning_rate': '8.692e-05', 'epoch': '1.781'}
+{'loss': '0.6878', 'grad_norm': '1.161', 'learning_rate': '8.691e-05', 'epoch': '1.782'}
+{'loss': '0.6316', 'grad_norm': '1.159', 'learning_rate': '8.689e-05', 'epoch': '1.783'}
+{'loss': '0.6761', 'grad_norm': '1.259', 'learning_rate': '8.687e-05', 'epoch': '1.784'}
+{'loss': '0.6714', 'grad_norm': '1.082', 'learning_rate': '8.686e-05', 'epoch': '1.784'}
+{'loss': '0.75', 'grad_norm': '1.365', 'learning_rate': '8.684e-05', 'epoch': '1.785'}
+{'loss': '0.6818', 'grad_norm': '1.125', 'learning_rate': '8.682e-05', 'epoch': '1.786'}
+{'loss': '0.5877', 'grad_norm': '1.079', 'learning_rate': '8.681e-05', 'epoch': '1.787'}
+{'loss': '0.7282', 'grad_norm': '1.092', 'learning_rate': '8.679e-05', 'epoch': '1.788'}
+{'loss': '0.7005', 'grad_norm': '1.066', 'learning_rate': '8.677e-05', 'epoch': '1.789'}
+{'loss': '0.6277', 'grad_norm': '1.214', 'learning_rate': '8.676e-05', 'epoch': '1.79'}
+{'loss': '0.5959', 'grad_norm': '1.104', 'learning_rate': '8.674e-05', 'epoch': '1.791'}
+{'loss': '0.6575', 'grad_norm': '1.035', 'learning_rate': '8.672e-05', 'epoch': '1.791'}
+{'loss': '0.6123', 'grad_norm': '0.9854', 'learning_rate': '8.671e-05', 'epoch': '1.792'}
+{'loss': '0.6271', 'grad_norm': '1.195', 'learning_rate': '8.669e-05', 'epoch': '1.793'}
+{'loss': '0.6668', 'grad_norm': '1.121', 'learning_rate': '8.667e-05', 'epoch': '1.794'}
+{'loss': '0.6663', 'grad_norm': '1.199', 'learning_rate': '8.666e-05', 'epoch': '1.795'}
+{'loss': '0.7308', 'grad_norm': '1.254', 'learning_rate': '8.664e-05', 'epoch': '1.796'}
+{'loss': '0.5975', 'grad_norm': '1.18', 'learning_rate': '8.662e-05', 'epoch': '1.797'}
+{'loss': '0.652', 'grad_norm': '1.016', 'learning_rate': '8.661e-05', 'epoch': '1.797'}
+{'loss': '0.5664', 'grad_norm': '0.9538', 'learning_rate': '8.659e-05', 'epoch': '1.798'}
+{'loss': '0.658', 'grad_norm': '1.238', 'learning_rate': '8.657e-05', 'epoch': '1.799'}
+{'loss': '0.5924', 'grad_norm': '1.119', 'learning_rate': '8.656e-05', 'epoch': '1.8'}
+{'loss': '0.6451', 'grad_norm': '1.178', 'learning_rate': '8.654e-05', 'epoch': '1.801'}
+{'loss': '0.6331', 'grad_norm': '1.157', 'learning_rate': '8.652e-05', 'epoch': '1.802'}
+{'loss': '0.6497', 'grad_norm': '1.255', 'learning_rate': '8.651e-05', 'epoch': '1.803'}
+{'loss': '0.633', 'grad_norm': '1.021', 'learning_rate': '8.649e-05', 'epoch': '1.803'}
+{'loss': '0.643', 'grad_norm': '1.094', 'learning_rate': '8.647e-05', 'epoch': '1.804'}
+{'loss': '0.7175', 'grad_norm': '1.217', 'learning_rate': '8.646e-05', 'epoch': '1.805'}
+{'loss': '0.6182', 'grad_norm': '1.263', 'learning_rate': '8.644e-05', 'epoch': '1.806'}
+{'loss': '0.6002', 'grad_norm': '1.209', 'learning_rate': '8.642e-05', 'epoch': '1.807'}
+{'loss': '0.6274', 'grad_norm': '1.156', 'learning_rate': '8.64e-05', 'epoch': '1.808'}
+{'loss': '0.5653', 'grad_norm': '1.005', 'learning_rate': '8.639e-05', 'epoch': '1.809'}
+{'loss': '0.745', 'grad_norm': '1.135', 'learning_rate': '8.637e-05', 'epoch': '1.809'}
+{'loss': '0.6311', 'grad_norm': '1.199', 'learning_rate': '8.635e-05', 'epoch': '1.81'}
+{'loss': '0.6941', 'grad_norm': '1.396', 'learning_rate': '8.634e-05', 'epoch': '1.811'}
+{'loss': '0.7299', 'grad_norm': '1.136', 'learning_rate': '8.632e-05', 'epoch': '1.812'}
+{'eval_loss': '0.6919', 'eval_runtime': '498.5', 'eval_samples_per_second': '4.132', 'eval_steps_per_second': '4.132', 'epoch': '1.812'}
+{'loss': '0.5869', 'grad_norm': '1.083', 'learning_rate': '8.63e-05', 'epoch': '1.813'}
+{'loss': '0.6635', 'grad_norm': '1.291', 'learning_rate': '8.629e-05', 'epoch': '1.814'}
+{'loss': '0.6379', 'grad_norm': '1.164', 'learning_rate': '8.627e-05', 'epoch': '1.815'}
+{'loss': '0.6388', 'grad_norm': '1.524', 'learning_rate': '8.625e-05', 'epoch': '1.816'}
+{'loss': '0.7544', 'grad_norm': '1.159', 'learning_rate': '8.624e-05', 'epoch': '1.816'}
+{'loss': '0.6456', 'grad_norm': '1.086', 'learning_rate': '8.622e-05', 'epoch': '1.817'}
+{'loss': '0.6785', 'grad_norm': '1.069', 'learning_rate': '8.62e-05', 'epoch': '1.818'}
+{'loss': '0.7251', 'grad_norm': '1.332', 'learning_rate': '8.619e-05', 'epoch': '1.819'}
+{'loss': '0.6594', 'grad_norm': '1.369', 'learning_rate': '8.617e-05', 'epoch': '1.82'}
+{'loss': '0.6335', 'grad_norm': '1.098', 'learning_rate': '8.615e-05', 'epoch': '1.821'}
+{'loss': '0.7029', 'grad_norm': '1.298', 'learning_rate': '8.613e-05', 'epoch': '1.822'}
+{'loss': '0.6502', 'grad_norm': '1.118', 'learning_rate': '8.612e-05', 'epoch': '1.822'}
+{'loss': '0.6504', 'grad_norm': '1.079', 'learning_rate': '8.61e-05', 'epoch': '1.823'}
+{'loss': '0.6614', 'grad_norm': '1.018', 'learning_rate': '8.608e-05', 'epoch': '1.824'}
+{'loss': '0.615', 'grad_norm': '1.013', 'learning_rate': '8.607e-05', 'epoch': '1.825'}
+{'loss': '0.5616', 'grad_norm': '1.049', 'learning_rate': '8.605e-05', 'epoch': '1.826'}
+{'loss': '0.6323', 'grad_norm': '1.21', 'learning_rate': '8.603e-05', 'epoch': '1.827'}
+{'loss': '0.6541', 'grad_norm': '1.264', 'learning_rate': '8.602e-05', 'epoch': '1.828'}
+{'loss': '0.672', 'grad_norm': '1.135', 'learning_rate': '8.6e-05', 'epoch': '1.828'}
+{'loss': '0.7372', 'grad_norm': '1.248', 'learning_rate': '8.598e-05', 'epoch': '1.829'}
+{'loss': '0.6711', 'grad_norm': '1.109', 'learning_rate': '8.596e-05', 'epoch': '1.83'}
+{'loss': '0.6282', 'grad_norm': '1.301', 'learning_rate': '8.595e-05', 'epoch': '1.831'}
+{'loss': '0.6156', 'grad_norm': '0.967', 'learning_rate': '8.593e-05', 'epoch': '1.832'}
+{'loss': '0.6416', 'grad_norm': '1.183', 'learning_rate': '8.591e-05', 'epoch': '1.833'}
+{'loss': '0.5512', 'grad_norm': '0.9577', 'learning_rate': '8.59e-05', 'epoch': '1.834'}
+{'loss': '0.6178', 'grad_norm': '1.163', 'learning_rate': '8.588e-05', 'epoch': '1.835'}
+{'loss': '0.5702', 'grad_norm': '0.9364', 'learning_rate': '8.586e-05', 'epoch': '1.835'}
+{'loss': '0.6038', 'grad_norm': '1.143', 'learning_rate': '8.584e-05', 'epoch': '1.836'}
+{'loss': '0.6888', 'grad_norm': '1.191', 'learning_rate': '8.583e-05', 'epoch': '1.837'}
+{'loss': '0.6213', 'grad_norm': '1.021', 'learning_rate': '8.581e-05', 'epoch': '1.838'}
+{'loss': '0.6181', 'grad_norm': '1.214', 'learning_rate': '8.579e-05', 'epoch': '1.839'}
+{'loss': '0.6168', 'grad_norm': '1.118', 'learning_rate': '8.578e-05', 'epoch': '1.84'}
+{'loss': '0.5986', 'grad_norm': '0.9956', 'learning_rate': '8.576e-05', 'epoch': '1.841'}
+{'loss': '0.6276', 'grad_norm': '1.074', 'learning_rate': '8.574e-05', 'epoch': '1.841'}
+{'loss': '0.6734', 'grad_norm': '1.166', 'learning_rate': '8.572e-05', 'epoch': '1.842'}
+{'loss': '0.6379', 'grad_norm': '1.15', 'learning_rate': '8.571e-05', 'epoch': '1.843'}
+{'loss': '0.6093', 'grad_norm': '1.023', 'learning_rate': '8.569e-05', 'epoch': '1.844'}
+{'loss': '0.6187', 'grad_norm': '1.117', 'learning_rate': '8.567e-05', 'epoch': '1.845'}
+{'loss': '0.6713', 'grad_norm': '1.22', 'learning_rate': '8.566e-05', 'epoch': '1.846'}
+{'loss': '0.6964', 'grad_norm': '1.179', 'learning_rate': '8.564e-05', 'epoch': '1.847'}
+{'loss': '0.5926', 'grad_norm': '0.9836', 'learning_rate': '8.562e-05', 'epoch': '1.847'}
+{'loss': '0.651', 'grad_norm': '1.027', 'learning_rate': '8.56e-05', 'epoch': '1.848'}
+{'loss': '0.6745', 'grad_norm': '1.015', 'learning_rate': '8.559e-05', 'epoch': '1.849'}
+{'loss': '0.5462', 'grad_norm': '1.007', 'learning_rate': '8.557e-05', 'epoch': '1.85'}
+{'loss': '0.6502', 'grad_norm': '1.174', 'learning_rate': '8.555e-05', 'epoch': '1.851'}
+{'loss': '0.6824', 'grad_norm': '1.192', 'learning_rate': '8.554e-05', 'epoch': '1.852'}
+{'loss': '0.6168', 'grad_norm': '1.128', 'learning_rate': '8.552e-05', 'epoch': '1.853'}
+{'loss': '0.6046', 'grad_norm': '1.327', 'learning_rate': '8.55e-05', 'epoch': '1.854'}
+{'loss': '0.6526', 'grad_norm': '1.102', 'learning_rate': '8.548e-05', 'epoch': '1.854'}
+{'loss': '0.6063', 'grad_norm': '1.152', 'learning_rate': '8.547e-05', 'epoch': '1.855'}
+{'eval_loss': '0.689', 'eval_runtime': '498.5', 'eval_samples_per_second': '4.132', 'eval_steps_per_second': '4.132', 'epoch': '1.855'}
+{'loss': '0.622', 'grad_norm': '1.132', 'learning_rate': '8.545e-05', 'epoch': '1.856'}
+{'loss': '0.6814', 'grad_norm': '1.269', 'learning_rate': '8.543e-05', 'epoch': '1.857'}
+{'loss': '0.7254', 'grad_norm': '1.269', 'learning_rate': '8.541e-05', 'epoch': '1.858'}
+{'loss': '0.6455', 'grad_norm': '1.198', 'learning_rate': '8.54e-05', 'epoch': '1.859'}
+{'loss': '0.6356', 'grad_norm': '1.024', 'learning_rate': '8.538e-05', 'epoch': '1.86'}
+{'loss': '0.6544', 'grad_norm': '1.193', 'learning_rate': '8.536e-05', 'epoch': '1.86'}
+{'loss': '0.7182', 'grad_norm': '1.19', 'learning_rate': '8.534e-05', 'epoch': '1.861'}
+{'loss': '0.6611', 'grad_norm': '1.297', 'learning_rate': '8.533e-05', 'epoch': '1.862'}
+{'loss': '0.684', 'grad_norm': '1.227', 'learning_rate': '8.531e-05', 'epoch': '1.863'}
+{'loss': '0.6289', 'grad_norm': '1.098', 'learning_rate': '8.529e-05', 'epoch': '1.864'}
+{'loss': '0.6778', 'grad_norm': '1.072', 'learning_rate': '8.527e-05', 'epoch': '1.865'}
+{'loss': '0.6152', 'grad_norm': '1.064', 'learning_rate': '8.526e-05', 'epoch': '1.866'}
+{'loss': '0.6196', 'grad_norm': '1.196', 'learning_rate': '8.524e-05', 'epoch': '1.866'}
+{'loss': '0.5703', 'grad_norm': '0.9389', 'learning_rate': '8.522e-05', 'epoch': '1.867'}
+{'loss': '0.6521', 'grad_norm': '1.019', 'learning_rate': '8.521e-05', 'epoch': '1.868'}
+{'loss': '0.6146', 'grad_norm': '1.239', 'learning_rate': '8.519e-05', 'epoch': '1.869'}
+{'loss': '0.6639', 'grad_norm': '1.521', 'learning_rate': '8.517e-05', 'epoch': '1.87'}
+{'loss': '0.6087', 'grad_norm': '1.064', 'learning_rate': '8.515e-05', 'epoch': '1.871'}
+{'loss': '0.7242', 'grad_norm': '1.18', 'learning_rate': '8.514e-05', 'epoch': '1.872'}
+{'loss': '0.6799', 'grad_norm': '1.261', 'learning_rate': '8.512e-05', 'epoch': '1.872'}
+{'loss': '0.6315', 'grad_norm': '1.106', 'learning_rate': '8.51e-05', 'epoch': '1.873'}
+{'loss': '0.6779', 'grad_norm': '1.227', 'learning_rate': '8.508e-05', 'epoch': '1.874'}
+{'loss': '0.6303', 'grad_norm': '1.255', 'learning_rate': '8.507e-05', 'epoch': '1.875'}
+{'loss': '0.6264', 'grad_norm': '1.034', 'learning_rate': '8.505e-05', 'epoch': '1.876'}
+{'loss': '0.6452', 'grad_norm': '1.148', 'learning_rate': '8.503e-05', 'epoch': '1.877'}
+{'loss': '0.7557', 'grad_norm': '1.176', 'learning_rate': '8.501e-05', 'epoch': '1.878'}
+{'loss': '0.6423', 'grad_norm': '1.217', 'learning_rate': '8.5e-05', 'epoch': '1.879'}
+{'loss': '0.6152', 'grad_norm': '1.114', 'learning_rate': '8.498e-05', 'epoch': '1.879'}
+{'loss': '0.5745', 'grad_norm': '1.061', 'learning_rate': '8.496e-05', 'epoch': '1.88'}
+{'loss': '0.6658', 'grad_norm': '1.263', 'learning_rate': '8.494e-05', 'epoch': '1.881'}
+{'loss': '0.6444', 'grad_norm': '1.097', 'learning_rate': '8.493e-05', 'epoch': '1.882'}
+{'loss': '0.6505', 'grad_norm': '1.168', 'learning_rate': '8.491e-05', 'epoch': '1.883'}
+{'loss': '0.6365', 'grad_norm': '1.207', 'learning_rate': '8.489e-05', 'epoch': '1.884'}
+{'loss': '0.5638', 'grad_norm': '1.092', 'learning_rate': '8.487e-05', 'epoch': '1.885'}
+{'loss': '0.569', 'grad_norm': '1.156', 'learning_rate': '8.485e-05', 'epoch': '1.885'}
+{'loss': '0.6627', 'grad_norm': '1.228', 'learning_rate': '8.484e-05', 'epoch': '1.886'}
+{'loss': '0.7027', 'grad_norm': '1.171', 'learning_rate': '8.482e-05', 'epoch': '1.887'}
+{'loss': '0.6484', 'grad_norm': '1.043', 'learning_rate': '8.48e-05', 'epoch': '1.888'}
+{'loss': '0.6339', 'grad_norm': '1.227', 'learning_rate': '8.478e-05', 'epoch': '1.889'}
+{'loss': '0.5816', 'grad_norm': '1.068', 'learning_rate': '8.477e-05', 'epoch': '1.89'}
+{'loss': '0.6339', 'grad_norm': '1.031', 'learning_rate': '8.475e-05', 'epoch': '1.891'}
+{'loss': '0.6637', 'grad_norm': '1.137', 'learning_rate': '8.473e-05', 'epoch': '1.891'}
+{'loss': '0.7456', 'grad_norm': '1.131', 'learning_rate': '8.471e-05', 'epoch': '1.892'}
+{'loss': '0.6591', 'grad_norm': '0.9601', 'learning_rate': '8.47e-05', 'epoch': '1.893'}
+{'loss': '0.6725', 'grad_norm': '1.158', 'learning_rate': '8.468e-05', 'epoch': '1.894'}
+{'loss': '0.6484', 'grad_norm': '1.21', 'learning_rate': '8.466e-05', 'epoch': '1.895'}
+{'loss': '0.6018', 'grad_norm': '1.187', 'learning_rate': '8.464e-05', 'epoch': '1.896'}
+{'loss': '0.6217', 'grad_norm': '1.249', 'learning_rate': '8.463e-05', 'epoch': '1.897'}
+{'loss': '0.6095', 'grad_norm': '1.13', 'learning_rate': '8.461e-05', 'epoch': '1.898'}
+{'loss': '0.59', 'grad_norm': '1.052', 'learning_rate': '8.459e-05', 'epoch': '1.898'}
+{'eval_loss': '0.6874', 'eval_runtime': '497.5', 'eval_samples_per_second': '4.14', 'eval_steps_per_second': '4.14', 'epoch': '1.898'}
+{'loss': '0.6216', 'grad_norm': '1.194', 'learning_rate': '8.457e-05', 'epoch': '1.899'}
+{'loss': '0.5427', 'grad_norm': '1.194', 'learning_rate': '8.455e-05', 'epoch': '1.9'}
+{'loss': '0.5974', 'grad_norm': '1.299', 'learning_rate': '8.454e-05', 'epoch': '1.901'}
+{'loss': '0.6071', 'grad_norm': '1.019', 'learning_rate': '8.452e-05', 'epoch': '1.902'}
+{'loss': '0.6673', 'grad_norm': '1.142', 'learning_rate': '8.45e-05', 'epoch': '1.903'}
+{'loss': '0.7247', 'grad_norm': '1.04', 'learning_rate': '8.448e-05', 'epoch': '1.904'}
+{'loss': '0.6222', 'grad_norm': '1.174', 'learning_rate': '8.447e-05', 'epoch': '1.904'}
+{'loss': '0.5822', 'grad_norm': '0.9496', 'learning_rate': '8.445e-05', 'epoch': '1.905'}
+{'loss': '0.6769', 'grad_norm': '1.192', 'learning_rate': '8.443e-05', 'epoch': '1.906'}
+{'loss': '0.5883', 'grad_norm': '1.084', 'learning_rate': '8.441e-05', 'epoch': '1.907'}
+{'loss': '0.6348', 'grad_norm': '1.128', 'learning_rate': '8.439e-05', 'epoch': '1.908'}
+{'loss': '0.6967', 'grad_norm': '1.2', 'learning_rate': '8.438e-05', 'epoch': '1.909'}
+{'loss': '0.6998', 'grad_norm': '1.075', 'learning_rate': '8.436e-05', 'epoch': '1.91'}
+{'loss': '0.6184', 'grad_norm': '0.9859', 'learning_rate': '8.434e-05', 'epoch': '1.91'}
+{'loss': '0.6684', 'grad_norm': '1.177', 'learning_rate': '8.432e-05', 'epoch': '1.911'}
+{'loss': '0.619', 'grad_norm': '1.181', 'learning_rate': '8.431e-05', 'epoch': '1.912'}
+{'loss': '0.6235', 'grad_norm': '1.098', 'learning_rate': '8.429e-05', 'epoch': '1.913'}
+{'loss': '0.6297', 'grad_norm': '1.007', 'learning_rate': '8.427e-05', 'epoch': '1.914'}
+{'loss': '0.67', 'grad_norm': '1.28', 'learning_rate': '8.425e-05', 'epoch': '1.915'}
+{'loss': '0.5961', 'grad_norm': '1.046', 'learning_rate': '8.423e-05', 'epoch': '1.916'}
+{'loss': '0.7021', 'grad_norm': '1.212', 'learning_rate': '8.422e-05', 'epoch': '1.917'}
+{'loss': '0.6828', 'grad_norm': '1.238', 'learning_rate': '8.42e-05', 'epoch': '1.917'}
+{'loss': '0.6749', 'grad_norm': '1.101', 'learning_rate': '8.418e-05', 'epoch': '1.918'}
+{'loss': '0.6203', 'grad_norm': '1.024', 'learning_rate': '8.416e-05', 'epoch': '1.919'}
+{'loss': '0.6312', 'grad_norm': '1.114', 'learning_rate': '8.414e-05', 'epoch': '1.92'}
+{'loss': '0.6777', 'grad_norm': '1.157', 'learning_rate': '8.413e-05', 'epoch': '1.921'}
+{'loss': '0.6252', 'grad_norm': '1.176', 'learning_rate': '8.411e-05', 'epoch': '1.922'}
+{'loss': '0.6271', 'grad_norm': '1.258', 'learning_rate': '8.409e-05', 'epoch': '1.923'}
+{'loss': '0.6023', 'grad_norm': '1.035', 'learning_rate': '8.407e-05', 'epoch': '1.923'}
+{'loss': '0.6764', 'grad_norm': '1.292', 'learning_rate': '8.405e-05', 'epoch': '1.924'}
+{'loss': '0.5721', 'grad_norm': '1.209', 'learning_rate': '8.404e-05', 'epoch': '1.925'}
+{'loss': '0.6361', 'grad_norm': '1.128', 'learning_rate': '8.402e-05', 'epoch': '1.926'}
+{'loss': '0.654', 'grad_norm': '1.162', 'learning_rate': '8.4e-05', 'epoch': '1.927'}
+{'loss': '0.6744', 'grad_norm': '1.154', 'learning_rate': '8.398e-05', 'epoch': '1.928'}
+{'loss': '0.6474', 'grad_norm': '0.9634', 'learning_rate': '8.396e-05', 'epoch': '1.929'}
+{'loss': '0.6781', 'grad_norm': '1.197', 'learning_rate': '8.395e-05', 'epoch': '1.929'}
+{'loss': '0.6782', 'grad_norm': '1.161', 'learning_rate': '8.393e-05', 'epoch': '1.93'}
+{'loss': '0.6876', 'grad_norm': '1.201', 'learning_rate': '8.391e-05', 'epoch': '1.931'}
+{'loss': '0.6446', 'grad_norm': '1.091', 'learning_rate': '8.389e-05', 'epoch': '1.932'}
+{'loss': '0.6835', 'grad_norm': '1.04', 'learning_rate': '8.387e-05', 'epoch': '1.933'}
+{'loss': '0.5878', 'grad_norm': '1.182', 'learning_rate': '8.386e-05', 'epoch': '1.934'}
+{'loss': '0.6302', 'grad_norm': '1.065', 'learning_rate': '8.384e-05', 'epoch': '1.935'}
+{'loss': '0.5749', 'grad_norm': '1.089', 'learning_rate': '8.382e-05', 'epoch': '1.935'}
+{'loss': '0.6426', 'grad_norm': '1.035', 'learning_rate': '8.38e-05', 'epoch': '1.936'}
+{'loss': '0.6552', 'grad_norm': '1.089', 'learning_rate': '8.378e-05', 'epoch': '1.937'}
+{'loss': '0.6078', 'grad_norm': '1.044', 'learning_rate': '8.377e-05', 'epoch': '1.938'}
+{'loss': '0.5918', 'grad_norm': '1.085', 'learning_rate': '8.375e-05', 'epoch': '1.939'}
+{'loss': '0.6819', 'grad_norm': '1.092', 'learning_rate': '8.373e-05', 'epoch': '1.94'}
+{'loss': '0.6909', 'grad_norm': '1.275', 'learning_rate': '8.371e-05', 'epoch': '1.941'}
+{'loss': '0.6681', 'grad_norm': '1.061', 'learning_rate': '8.369e-05', 'epoch': '1.942'}
+{'eval_loss': '0.6851', 'eval_runtime': '497.1', 'eval_samples_per_second': '4.144', 'eval_steps_per_second': '4.144', 'epoch': '1.942'}
+{'loss': '0.5933', 'grad_norm': '1.013', 'learning_rate': '8.367e-05', 'epoch': '1.942'}
+{'loss': '0.6292', 'grad_norm': '1.186', 'learning_rate': '8.366e-05', 'epoch': '1.943'}
+{'loss': '0.6341', 'grad_norm': '1.075', 'learning_rate': '8.364e-05', 'epoch': '1.944'}
+{'loss': '0.7104', 'grad_norm': '1.201', 'learning_rate': '8.362e-05', 'epoch': '1.945'}
+{'loss': '0.5895', 'grad_norm': '0.992', 'learning_rate': '8.36e-05', 'epoch': '1.946'}
+{'loss': '0.5887', 'grad_norm': '1.002', 'learning_rate': '8.358e-05', 'epoch': '1.947'}
+{'loss': '0.7066', 'grad_norm': '1.036', 'learning_rate': '8.357e-05', 'epoch': '1.948'}
+{'loss': '0.6521', 'grad_norm': '1.208', 'learning_rate': '8.355e-05', 'epoch': '1.948'}
+{'loss': '0.5829', 'grad_norm': '1.001', 'learning_rate': '8.353e-05', 'epoch': '1.949'}
+{'loss': '0.6763', 'grad_norm': '1.347', 'learning_rate': '8.351e-05', 'epoch': '1.95'}
+{'loss': '0.6664', 'grad_norm': '1.196', 'learning_rate': '8.349e-05', 'epoch': '1.951'}
+{'loss': '0.6009', 'grad_norm': '1.031', 'learning_rate': '8.347e-05', 'epoch': '1.952'}
+{'loss': '0.5942', 'grad_norm': '1.137', 'learning_rate': '8.346e-05', 'epoch': '1.953'}
+{'loss': '0.6537', 'grad_norm': '1.183', 'learning_rate': '8.344e-05', 'epoch': '1.954'}
+{'loss': '0.6444', 'grad_norm': '1.2', 'learning_rate': '8.342e-05', 'epoch': '1.954'}
+{'loss': '0.6323', 'grad_norm': '1.222', 'learning_rate': '8.34e-05', 'epoch': '1.955'}
+{'loss': '0.5936', 'grad_norm': '1.103', 'learning_rate': '8.338e-05', 'epoch': '1.956'}
+{'loss': '0.597', 'grad_norm': '0.8767', 'learning_rate': '8.337e-05', 'epoch': '1.957'}
+{'loss': '0.6459', 'grad_norm': '1.02', 'learning_rate': '8.335e-05', 'epoch': '1.958'}
+{'loss': '0.7253', 'grad_norm': '1.179', 'learning_rate': '8.333e-05', 'epoch': '1.959'}
+{'loss': '0.5783', 'grad_norm': '0.938', 'learning_rate': '8.331e-05', 'epoch': '1.96'}
+{'loss': '0.6981', 'grad_norm': '1.307', 'learning_rate': '8.329e-05', 'epoch': '1.961'}
+{'loss': '0.7058', 'grad_norm': '1.057', 'learning_rate': '8.327e-05', 'epoch': '1.961'}
+{'loss': '0.6462', 'grad_norm': '1.161', 'learning_rate': '8.326e-05', 'epoch': '1.962'}
+{'loss': '0.613', 'grad_norm': '1.089', 'learning_rate': '8.324e-05', 'epoch': '1.963'}
+{'loss': '0.5801', 'grad_norm': '1.128', 'learning_rate': '8.322e-05', 'epoch': '1.964'}
+{'loss': '0.5806', 'grad_norm': '1.027', 'learning_rate': '8.32e-05', 'epoch': '1.965'}
+{'loss': '0.6186', 'grad_norm': '0.9745', 'learning_rate': '8.318e-05', 'epoch': '1.966'}
+{'loss': '0.6853', 'grad_norm': '1.227', 'learning_rate': '8.316e-05', 'epoch': '1.967'}
+{'loss': '0.5712', 'grad_norm': '1.027', 'learning_rate': '8.315e-05', 'epoch': '1.967'}
+{'loss': '0.645', 'grad_norm': '1.126', 'learning_rate': '8.313e-05', 'epoch': '1.968'}
+{'loss': '0.6193', 'grad_norm': '1.08', 'learning_rate': '8.311e-05', 'epoch': '1.969'}
+{'loss': '0.6526', 'grad_norm': '1.148', 'learning_rate': '8.309e-05', 'epoch': '1.97'}
+{'loss': '0.63', 'grad_norm': '1.062', 'learning_rate': '8.307e-05', 'epoch': '1.971'}
+{'loss': '0.603', 'grad_norm': '1.063', 'learning_rate': '8.305e-05', 'epoch': '1.972'}
+{'loss': '0.6269', 'grad_norm': '1.181', 'learning_rate': '8.303e-05', 'epoch': '1.973'}
+{'loss': '0.6771', 'grad_norm': '1.414', 'learning_rate': '8.302e-05', 'epoch': '1.973'}
+{'loss': '0.6195', 'grad_norm': '1.151', 'learning_rate': '8.3e-05', 'epoch': '1.974'}
+{'loss': '0.6515', 'grad_norm': '0.9357', 'learning_rate': '8.298e-05', 'epoch': '1.975'}
+{'loss': '0.5889', 'grad_norm': '1.054', 'learning_rate': '8.296e-05', 'epoch': '1.976'}
+{'loss': '0.6179', 'grad_norm': '1.395', 'learning_rate': '8.294e-05', 'epoch': '1.977'}
+{'loss': '0.6646', 'grad_norm': '1.012', 'learning_rate': '8.292e-05', 'epoch': '1.978'}
+{'loss': '0.6752', 'grad_norm': '1.144', 'learning_rate': '8.291e-05', 'epoch': '1.979'}
+{'loss': '0.6158', 'grad_norm': '1.415', 'learning_rate': '8.289e-05', 'epoch': '1.98'}
+{'loss': '0.7818', 'grad_norm': '1.25', 'learning_rate': '8.287e-05', 'epoch': '1.98'}
+{'loss': '0.6193', 'grad_norm': '1.039', 'learning_rate': '8.285e-05', 'epoch': '1.981'}
+{'loss': '0.6016', 'grad_norm': '1.258', 'learning_rate': '8.283e-05', 'epoch': '1.982'}
+{'loss': '0.628', 'grad_norm': '1.175', 'learning_rate': '8.281e-05', 'epoch': '1.983'}
+{'loss': '0.5715', 'grad_norm': '1.119', 'learning_rate': '8.279e-05', 'epoch': '1.984'}
+{'loss': '0.5901', 'grad_norm': '1.2', 'learning_rate': '8.278e-05', 'epoch': '1.985'}
+{'eval_loss': '0.6843', 'eval_runtime': '498.3', 'eval_samples_per_second': '4.134', 'eval_steps_per_second': '4.134', 'epoch': '1.985'}
+{'loss': '0.5733', 'grad_norm': '1.012', 'learning_rate': '8.276e-05', 'epoch': '1.986'}
+{'loss': '0.7013', 'grad_norm': '1.336', 'learning_rate': '8.274e-05', 'epoch': '1.986'}
+{'loss': '0.5518', 'grad_norm': '1.291', 'learning_rate': '8.272e-05', 'epoch': '1.987'}
+{'loss': '0.5563', 'grad_norm': '0.9338', 'learning_rate': '8.27e-05', 'epoch': '1.988'}
+{'loss': '0.6349', 'grad_norm': '1.474', 'learning_rate': '8.268e-05', 'epoch': '1.989'}
+{'loss': '0.6168', 'grad_norm': '1.278', 'learning_rate': '8.266e-05', 'epoch': '1.99'}
+{'loss': '0.5403', 'grad_norm': '1.088', 'learning_rate': '8.265e-05', 'epoch': '1.991'}
+{'loss': '0.6327', 'grad_norm': '1.13', 'learning_rate': '8.263e-05', 'epoch': '1.992'}
+{'loss': '0.6707', 'grad_norm': '1.201', 'learning_rate': '8.261e-05', 'epoch': '1.992'}
+{'loss': '0.6825', 'grad_norm': '1.225', 'learning_rate': '8.259e-05', 'epoch': '1.993'}
+{'loss': '0.5959', 'grad_norm': '1.12', 'learning_rate': '8.257e-05', 'epoch': '1.994'}
+{'loss': '0.5317', 'grad_norm': '1.174', 'learning_rate': '8.255e-05', 'epoch': '1.995'}
+{'loss': '0.6555', 'grad_norm': '1.161', 'learning_rate': '8.253e-05', 'epoch': '1.996'}
+{'loss': '0.6569', 'grad_norm': '1.152', 'learning_rate': '8.252e-05', 'epoch': '1.997'}
+{'loss': '0.6312', 'grad_norm': '1.074', 'learning_rate': '8.25e-05', 'epoch': '1.998'}
+{'loss': '0.6245', 'grad_norm': '1.107', 'learning_rate': '8.248e-05', 'epoch': '1.998'}
+{'loss': '0.5207', 'grad_norm': '0.87', 'learning_rate': '8.246e-05', 'epoch': '1.999'}
+{'loss': '0.5883', 'grad_norm': '1.627', 'learning_rate': '8.244e-05', 'epoch': '2'}
+{'loss': '0.6075', 'grad_norm': '1.344', 'learning_rate': '8.242e-05', 'epoch': '2.001'}
+{'loss': '0.603', 'grad_norm': '1.126', 'learning_rate': '8.24e-05', 'epoch': '2.002'}
+{'loss': '0.6273', 'grad_norm': '1.049', 'learning_rate': '8.238e-05', 'epoch': '2.003'}
+{'loss': '0.5028', 'grad_norm': '1.095', 'learning_rate': '8.237e-05', 'epoch': '2.003'}
+{'loss': '0.5689', 'grad_norm': '1.155', 'learning_rate': '8.235e-05', 'epoch': '2.004'}
+{'loss': '0.612', 'grad_norm': '1.175', 'learning_rate': '8.233e-05', 'epoch': '2.005'}
+{'loss': '0.6186', 'grad_norm': '1.125', 'learning_rate': '8.231e-05', 'epoch': '2.006'}
+{'loss': '0.5442', 'grad_norm': '1.118', 'learning_rate': '8.229e-05', 'epoch': '2.007'}
+{'loss': '0.5793', 'grad_norm': '1.207', 'learning_rate': '8.227e-05', 'epoch': '2.008'}
+{'loss': '0.6538', 'grad_norm': '1.174', 'learning_rate': '8.225e-05', 'epoch': '2.009'}
+{'loss': '0.615', 'grad_norm': '1.052', 'learning_rate': '8.223e-05', 'epoch': '2.009'}
+{'loss': '0.5928', 'grad_norm': '1.063', 'learning_rate': '8.222e-05', 'epoch': '2.01'}
+{'loss': '0.5732', 'grad_norm': '1.122', 'learning_rate': '8.22e-05', 'epoch': '2.011'}
+{'loss': '0.5882', 'grad_norm': '0.9991', 'learning_rate': '8.218e-05', 'epoch': '2.012'}
+{'loss': '0.5597', 'grad_norm': '1.2', 'learning_rate': '8.216e-05', 'epoch': '2.013'}
+{'loss': '0.5848', 'grad_norm': '1.19', 'learning_rate': '8.214e-05', 'epoch': '2.014'}
+{'loss': '0.6188', 'grad_norm': '1.139', 'learning_rate': '8.212e-05', 'epoch': '2.015'}
+{'loss': '0.6273', 'grad_norm': '1.023', 'learning_rate': '8.21e-05', 'epoch': '2.016'}
+{'loss': '0.6496', 'grad_norm': '1.242', 'learning_rate': '8.208e-05', 'epoch': '2.016'}
+{'loss': '0.4905', 'grad_norm': '0.862', 'learning_rate': '8.207e-05', 'epoch': '2.017'}
+{'loss': '0.6047', 'grad_norm': '1.251', 'learning_rate': '8.205e-05', 'epoch': '2.018'}
+{'loss': '0.6835', 'grad_norm': '1.105', 'learning_rate': '8.203e-05', 'epoch': '2.019'}
+{'loss': '0.588', 'grad_norm': '1.11', 'learning_rate': '8.201e-05', 'epoch': '2.02'}
+{'loss': '0.5747', 'grad_norm': '1.238', 'learning_rate': '8.199e-05', 'epoch': '2.021'}
+{'loss': '0.5815', 'grad_norm': '0.9958', 'learning_rate': '8.197e-05', 'epoch': '2.022'}
+{'loss': '0.5182', 'grad_norm': '1.049', 'learning_rate': '8.195e-05', 'epoch': '2.022'}
+{'loss': '0.5116', 'grad_norm': '1.133', 'learning_rate': '8.193e-05', 'epoch': '2.023'}
+{'loss': '0.6048', 'grad_norm': '1.223', 'learning_rate': '8.191e-05', 'epoch': '2.024'}
+{'loss': '0.5399', 'grad_norm': '1.126', 'learning_rate': '8.19e-05', 'epoch': '2.025'}
+{'loss': '0.5849', 'grad_norm': '1.176', 'learning_rate': '8.188e-05', 'epoch': '2.026'}
+{'loss': '0.6259', 'grad_norm': '1.245', 'learning_rate': '8.186e-05', 'epoch': '2.027'}
+{'loss': '0.5777', 'grad_norm': '1.243', 'learning_rate': '8.184e-05', 'epoch': '2.028'}
+{'eval_loss': '0.6861', 'eval_runtime': '498.1', 'eval_samples_per_second': '4.136', 'eval_steps_per_second': '4.136', 'epoch': '2.028'}
+{'loss': '0.5753', 'grad_norm': '1.228', 'learning_rate': '8.182e-05', 'epoch': '2.028'}
+{'loss': '0.5633', 'grad_norm': '1.313', 'learning_rate': '8.18e-05', 'epoch': '2.029'}
+{'loss': '0.5915', 'grad_norm': '1.066', 'learning_rate': '8.178e-05', 'epoch': '2.03'}
+{'loss': '0.5696', 'grad_norm': '1.314', 'learning_rate': '8.176e-05', 'epoch': '2.031'}
+{'loss': '0.6609', 'grad_norm': '1.162', 'learning_rate': '8.174e-05', 'epoch': '2.032'}
+{'loss': '0.6298', 'grad_norm': '1.521', 'learning_rate': '8.173e-05', 'epoch': '2.033'}
+{'loss': '0.5098', 'grad_norm': '1.114', 'learning_rate': '8.171e-05', 'epoch': '2.034'}
+{'loss': '0.6025', 'grad_norm': '1.325', 'learning_rate': '8.169e-05', 'epoch': '2.035'}
+{'loss': '0.6645', 'grad_norm': '1.248', 'learning_rate': '8.167e-05', 'epoch': '2.035'}
+{'loss': '0.6095', 'grad_norm': '1.149', 'learning_rate': '8.165e-05', 'epoch': '2.036'}
+{'loss': '0.6366', 'grad_norm': '1.196', 'learning_rate': '8.163e-05', 'epoch': '2.037'}
+{'loss': '0.6125', 'grad_norm': '1.381', 'learning_rate': '8.161e-05', 'epoch': '2.038'}
+{'loss': '0.5961', 'grad_norm': '1.109', 'learning_rate': '8.159e-05', 'epoch': '2.039'}
+{'loss': '0.5667', 'grad_norm': '1.193', 'learning_rate': '8.157e-05', 'epoch': '2.04'}
+{'loss': '0.6202', 'grad_norm': '1.369', 'learning_rate': '8.155e-05', 'epoch': '2.041'}
+{'loss': '0.6066', 'grad_norm': '1.38', 'learning_rate': '8.154e-05', 'epoch': '2.041'}
+{'loss': '0.5777', 'grad_norm': '1.231', 'learning_rate': '8.152e-05', 'epoch': '2.042'}
+{'loss': '0.5413', 'grad_norm': '1.244', 'learning_rate': '8.15e-05', 'epoch': '2.043'}
+{'loss': '0.5795', 'grad_norm': '1.36', 'learning_rate': '8.148e-05', 'epoch': '2.044'}
+{'loss': '0.617', 'grad_norm': '1.199', 'learning_rate': '8.146e-05', 'epoch': '2.045'}
+{'loss': '0.59', 'grad_norm': '1.187', 'learning_rate': '8.144e-05', 'epoch': '2.046'}
+{'loss': '0.6028', 'grad_norm': '1.102', 'learning_rate': '8.142e-05', 'epoch': '2.047'}
+{'loss': '0.6313', 'grad_norm': '1.081', 'learning_rate': '8.14e-05', 'epoch': '2.047'}
+{'loss': '0.6083', 'grad_norm': '4.683', 'learning_rate': '8.138e-05', 'epoch': '2.048'}
+{'loss': '0.6332', 'grad_norm': '1.442', 'learning_rate': '8.136e-05', 'epoch': '2.049'}
+{'loss': '0.5841', 'grad_norm': '1.156', 'learning_rate': '8.134e-05', 'epoch': '2.05'}
+{'loss': '0.5268', 'grad_norm': '1.011', 'learning_rate': '8.133e-05', 'epoch': '2.051'}
+{'loss': '0.6303', 'grad_norm': '1.179', 'learning_rate': '8.131e-05', 'epoch': '2.052'}
+{'loss': '0.5571', 'grad_norm': '1.114', 'learning_rate': '8.129e-05', 'epoch': '2.053'}
+{'loss': '0.5323', 'grad_norm': '1.044', 'learning_rate': '8.127e-05', 'epoch': '2.054'}
+{'loss': '0.6114', 'grad_norm': '1.26', 'learning_rate': '8.125e-05', 'epoch': '2.054'}
+{'loss': '0.5971', 'grad_norm': '1.077', 'learning_rate': '8.123e-05', 'epoch': '2.055'}
+{'loss': '0.575', 'grad_norm': '1.112', 'learning_rate': '8.121e-05', 'epoch': '2.056'}
+{'loss': '0.6124', 'grad_norm': '1.144', 'learning_rate': '8.119e-05', 'epoch': '2.057'}
+{'loss': '0.5593', 'grad_norm': '1.258', 'learning_rate': '8.117e-05', 'epoch': '2.058'}
+{'loss': '0.5903', 'grad_norm': '1.181', 'learning_rate': '8.115e-05', 'epoch': '2.059'}
+{'loss': '0.5889', 'grad_norm': '1.535', 'learning_rate': '8.113e-05', 'epoch': '2.06'}
+{'loss': '0.5815', 'grad_norm': '1.087', 'learning_rate': '8.111e-05', 'epoch': '2.06'}
+{'loss': '0.5762', 'grad_norm': '1.211', 'learning_rate': '8.109e-05', 'epoch': '2.061'}
+{'loss': '0.6051', 'grad_norm': '1.324', 'learning_rate': '8.108e-05', 'epoch': '2.062'}
+{'loss': '0.6266', 'grad_norm': '1.138', 'learning_rate': '8.106e-05', 'epoch': '2.063'}
+{'loss': '0.6093', 'grad_norm': '1.372', 'learning_rate': '8.104e-05', 'epoch': '2.064'}
+{'loss': '0.573', 'grad_norm': '1.097', 'learning_rate': '8.102e-05', 'epoch': '2.065'}
+{'loss': '0.5691', 'grad_norm': '1.243', 'learning_rate': '8.1e-05', 'epoch': '2.066'}
+{'loss': '0.5794', 'grad_norm': '1.329', 'learning_rate': '8.098e-05', 'epoch': '2.066'}
+{'loss': '0.5537', 'grad_norm': '1.124', 'learning_rate': '8.096e-05', 'epoch': '2.067'}
+{'loss': '0.6083', 'grad_norm': '1.245', 'learning_rate': '8.094e-05', 'epoch': '2.068'}
+{'loss': '0.5528', 'grad_norm': '1.302', 'learning_rate': '8.092e-05', 'epoch': '2.069'}
+{'loss': '0.5502', 'grad_norm': '1.179', 'learning_rate': '8.09e-05', 'epoch': '2.07'}
+{'loss': '0.5832', 'grad_norm': '1.153', 'learning_rate': '8.088e-05', 'epoch': '2.071'}
+{'eval_loss': '0.6874', 'eval_runtime': '497.8', 'eval_samples_per_second': '4.138', 'eval_steps_per_second': '4.138', 'epoch': '2.071'}
+{'train_runtime': '8.017e+04', 'train_samples_per_second': '1.388', 'train_steps_per_second': '0.173', 'train_loss': '0.7331', 'epoch': '2.071'}
+Saved best adapter -> runs/instruct_run_24b/best_adapter
+Final eval_loss=0.6851214170455933, ppl=1.9840127138756343
+--- Merge: runs/instruct_run_24b/best_adapter + CPT/runs/cpt_run_v1/merged_24b_cpt_lora -> runs/instruct_run_24b/merged_24b_instruct_lora ---
+`torch_dtype` is deprecated! Use `dtype` instead!
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 921, in <module>
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 912, in main
+    merge_adapter(cfg, base_dir, best_adapter_dir, final_dir)
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 692, in merge_adapter
+    base = AutoModelForCausalLM.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 376, in from_pretrained
+    raise ValueError(
+ValueError: Unrecognized configuration class <class 'transformers.models.mistral3.configuration_mistral3.Mistral3Config'> for this kind of AutoModel: AutoModelForCausalLM.
+Model type should be one of AfmoeConfig, ApertusConfig, ArceeConfig, AriaTextConfig, BambaConfig, BartConfig, BertConfig, BertGenerationConfig, BigBirdConfig, BigBirdPegasusConfig, BioGptConfig, BitNetConfig, BlenderbotConfig, BlenderbotSmallConfig, BloomConfig, BltConfig, CamembertConfig, LlamaConfig, CodeGenConfig, CohereConfig, Cohere2Config, CpmAntConfig, CTRLConfig, CwmConfig, Data2VecTextConfig, DbrxConfig, DeepseekV2Config, DeepseekV3Config, DiffLlamaConfig, DogeConfig, Dots1Config, ElectraConfig, Emu3Config, ErnieConfig, Ernie4_5Config, Ernie4_5_MoeConfig, Exaone4Config, FalconConfig, FalconH1Config, FalconMambaConfig, FlexOlmoConfig, FuyuConfig, GemmaConfig, Gemma2Config, Gemma3Config, Gemma3TextConfig, Gemma3nConfig, Gemma3nTextConfig, GitConfig, GlmConfig, Glm4Config, Glm4MoeConfig, GotOcr2Config, GPT2Config, GPT2Config, GPTBigCodeConfig, GPTNeoConfig, GPTNeoXConfig, GPTNeoXJapaneseConfig, GptOssConfig, GPTJConfig, GraniteConfig, GraniteMoeConfig, GraniteMoeHybridConfig, GraniteMoeSharedConfig, HeliumConfig, HunYuanDenseV1Config, HunYuanMoEV1Config, Jais2Config, JambaConfig, JetMoeConfig, Lfm2Config, Lfm2MoeConfig, LlamaConfig, Llama4Config, Llama4TextConfig, LongcatFlashConfig, MambaConfig, Mamba2Config, MarianConfig, MBartConfig, MegatronBertConfig, MiniMaxConfig, MinistralConfig, Ministral3Config, MistralConfig, MixtralConfig, MllamaConfig, ModernBertDecoderConfig, MoshiConfig, MptConfig, MusicgenConfig, MusicgenMelodyConfig, MvpConfig, NanoChatConfig, NemotronConfig, OlmoConfig, Olmo2Config, Olmo3Config, OlmoeConfig, OpenAIGPTConfig, OPTConfig, PegasusConfig, PersimmonConfig, PhiConfig, Phi3Config, Phi4MultimodalConfig, PhimoeConfig, PLBartConfig, ProphetNetConfig, Qwen2Config, Qwen2MoeConfig, Qwen3Config, Qwen3MoeConfig, Qwen3NextConfig, RecurrentGemmaConfig, ReformerConfig, RemBertConfig, RobertaConfig, RobertaPreLayerNormConfig, RoCBertConfig, RoFormerConfig, RwkvConfig, SeedOssConfig, SmolLM3Config, StableLmConfig, Starcoder2Config, TrOCRConfig, VaultGemmaConfig, WhisperConfig, XGLMConfig, XLMConfig, XLMRobertaConfig, XLMRobertaXLConfig, XLNetConfig, xLSTMConfig, XmodConfig, ZambaConfig, Zamba2Config.
+Traceback (most recent call last):
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 921, in <module>
+    main()
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 912, in main
+    merge_adapter(cfg, base_dir, best_adapter_dir, final_dir)
+  File "/workspace/Trainer-kit/SFT/run_instruct.py", line 692, in merge_adapter
+    base = AutoModelForCausalLM.from_pretrained(
+  File "/workspace/CPT-env/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py", line 376, in from_pretrained
+    raise ValueError(
+ValueError: Unrecognized configuration class <class 'transformers.models.mistral3.configuration_mistral3.Mistral3Config'> for this kind of AutoModel: AutoModelForCausalLM.
+Model type should be one of AfmoeConfig, ApertusConfig, ArceeConfig, AriaTextConfig, BambaConfig, BartConfig, BertConfig, BertGenerationConfig, BigBirdConfig, BigBirdPegasusConfig, BioGptConfig, BitNetConfig, BlenderbotConfig, BlenderbotSmallConfig, BloomConfig, BltConfig, CamembertConfig, LlamaConfig, CodeGenConfig, CohereConfig, Cohere2Config, CpmAntConfig, CTRLConfig, CwmConfig, Data2VecTextConfig, DbrxConfig, DeepseekV2Config, DeepseekV3Config, DiffLlamaConfig, DogeConfig, Dots1Config, ElectraConfig, Emu3Config, ErnieConfig, Ernie4_5Config, Ernie4_5_MoeConfig, Exaone4Config, FalconConfig, FalconH1Config, FalconMambaConfig, FlexOlmoConfig, FuyuConfig, GemmaConfig, Gemma2Config, Gemma3Config, Gemma3TextConfig, Gemma3nConfig, Gemma3nTextConfig, GitConfig, GlmConfig, Glm4Config, Glm4MoeConfig, GotOcr2Config, GPT2Config, GPT2Config, GPTBigCodeConfig, GPTNeoConfig, GPTNeoXConfig, GPTNeoXJapaneseConfig, GptOssConfig, GPTJConfig, GraniteConfig, GraniteMoeConfig, GraniteMoeHybridConfig, GraniteMoeSharedConfig, HeliumConfig, HunYuanDenseV1Config, HunYuanMoEV1Config, Jais2Config, JambaConfig, JetMoeConfig, Lfm2Config, Lfm2MoeConfig, LlamaConfig, Llama4Config, Llama4TextConfig, LongcatFlashConfig, MambaConfig, Mamba2Config, MarianConfig, MBartConfig, MegatronBertConfig, MiniMaxConfig, MinistralConfig, Ministral3Config, MistralConfig, MixtralConfig, MllamaConfig, ModernBertDecoderConfig, MoshiConfig, MptConfig, MusicgenConfig, MusicgenMelodyConfig, MvpConfig, NanoChatConfig, NemotronConfig, OlmoConfig, Olmo2Config, Olmo3Config, OlmoeConfig, OpenAIGPTConfig, OPTConfig, PegasusConfig, PersimmonConfig, PhiConfig, Phi3Config, Phi4MultimodalConfig, PhimoeConfig, PLBartConfig, ProphetNetConfig, Qwen2Config, Qwen2MoeConfig, Qwen3Config, Qwen3MoeConfig, Qwen3NextConfig, RecurrentGemmaConfig, ReformerConfig, RemBertConfig, RobertaConfig, RobertaPreLayerNormConfig, RoCBertConfig, RoFormerConfig, RwkvConfig, SeedOssConfig, SmolLM3Config, StableLmConfig, Starcoder2Config, TrOCRConfig, VaultGemmaConfig, WhisperConfig, XGLMConfig, XLMConfig, XLMRobertaConfig, XLMRobertaXLConfig, XLNetConfig, xLSTMConfig, XmodConfig, ZambaConfig, Zamba2Config.
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/requirements.txt b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e7931173f4304b55fcec981001d950eebb55e97c
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/requirements.txt
@@ -0,0 +1,90 @@
+exceptiongroup==1.3.1
+python-dateutil==2.9.0.post0
+nvidia-ml-py==13.580.82
+huggingface_hub==1.2.3
+idna==3.11
+click==8.3.1
+numpy==2.2.6
+httpx==0.28.1
+nvidia-nvshmem-cu12==3.3.20
+tokenizers==0.22.1
+nvidia-cufile-cu12==1.13.1.3
+nvidia-cublas-cu12==12.8.4.1
+MarkupSafe==3.0.3
+yarl==1.22.0
+async-timeout==5.0.1
+sympy==1.14.0
+datasets==4.4.2
+platformdirs==4.5.1
+nvidia-cusolver-cu12==11.7.3.90
+smmap==5.0.2
+accelerate==1.12.0
+requests==2.32.5
+nvidia-nccl-cu12==2.27.5
+nvidia-cuda-nvrtc-cu12==12.8.93
+aiohttp==3.13.2
+bitsandbytes==0.49.0
+mpmath==1.3.0
+typing-inspection==0.4.2
+nvidia-cudnn-cu12==9.10.2.21
+GitPython==3.1.45
+xxhash==3.6.0
+pydantic_core==2.41.5
+setuptools==59.6.0
+six==1.17.0
+typing_extensions==4.15.0
+filelock==3.20.1
+charset-normalizer==3.4.4
+nvitop==1.6.1
+wandb==0.23.1
+regex==2025.11.3
+nvidia-cuda-runtime-cu12==12.8.90
+absl-py==2.3.1
+pytz==2025.2
+rouge-score==0.1.2
+torch==2.9.1
+Jinja2==3.1.6
+nvidia-cusparse-cu12==12.5.8.93
+psutil==7.1.3
+nltk==3.9.2
+packaging==25.0
+safetensors==0.7.0
+sentry-sdk==2.48.0
+gitdb==4.0.12
+httpcore==1.0.9
+anyio==4.12.0
+transformers==5.0.0.dev0
+pydantic==2.12.5
+fsspec==2025.10.0
+PyYAML==6.0.3
+hf-xet==1.2.0
+typer-slim==0.20.1
+triton==3.5.1
+nvidia-nvtx-cu12==12.8.90
+tqdm==4.67.1
+attrs==25.4.0
+peft==0.18.0
+aiohappyeyeballs==2.6.1
+networkx==3.4.2
+nvidia-cufft-cu12==11.3.3.83
+certifi==2025.11.12
+pyarrow==22.0.0
+dill==0.4.0
+protobuf==6.33.2
+aiosignal==1.4.0
+frozenlist==1.8.0
+urllib3==2.6.2
+propcache==0.4.1
+tzdata==2025.3
+pandas==2.3.3
+annotated-types==0.7.0
+shellingham==1.5.4
+multidict==6.7.0
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+nvidia-nvjitlink-cu12==12.8.93
+h11==0.16.0
+multiprocess==0.70.18
+nvidia-curand-cu12==10.3.9.90
+pip==22.0.2
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/wandb-metadata.json b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/wandb-metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..27536d6107604e586157512bf504652453e7e8fa
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/wandb-metadata.json
@@ -0,0 +1,51 @@
+{
+  "os":  "Linux-6.12.46+-x86_64-with-glibc2.35",
+  "python":  "CPython 3.10.12",
+  "startedAt":  "2025-12-23T14:22:35.730910Z",
+  "args":  [
+    "--config",
+    "./SFT/config_instruct.yaml"
+  ],
+  "program":  "/workspace/Trainer-kit/SFT/run_instruct.py",
+  "codePath":  "SFT/run_instruct.py",
+  "codePathLocal":  "SFT/run_instruct.py",
+  "git":  {
+    "remote":  "https://github.com/Ashx098/Trainer-kit.git",
+    "commit":  "92382202abe01938bfeff39a3bf7b82e6e786325"
+  },
+  "email":  "shaiksirajuddin9949@gmail.com",
+  "root":  "runs/instruct_run_24b",
+  "host":  "a100-2gpu-shell-session-9c959f5f-mx8rm",
+  "executable":  "/workspace/CPT-env/bin/python",
+  "cpu_count":  12,
+  "cpu_count_logical":  24,
+  "gpu":  "NVIDIA A100-SXM4-80GB",
+  "gpu_count":  2,
+  "disk":  {
+    "/":  {
+      "total":  "791251738624",
+      "used":  "329782792192"
+    }
+  },
+  "memory":  {
+    "total":  "359047892992"
+  },
+  "gpu_nvidia":  [
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-989794b0-ec3b-13bf-db9f-3fbe341497ba"
+    },
+    {
+      "name":  "NVIDIA A100-SXM4-80GB",
+      "memoryTotal":  "85899345920",
+      "cudaCores":  6912,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-3790aa64-60ef-9eac-b0b1-b278ee8c0d40"
+    }
+  ],
+  "cudaVersion":  "13.0",
+  "writerId":  "dkyx2p7rbyqdypnvz73q5ob1r5e53xmv"
+}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/wandb-summary.json b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/wandb-summary.json
new file mode 100644
index 0000000000000000000000000000000000000000..22c1184cf6f667a524c61a95ba24acc153a12fa6
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/files/wandb-summary.json
@@ -0,0 +1 @@
+{"train_samples_per_second":1.388,"_timestamp":1.766580468251779e+09,"total_flos":5.005098268090491e+18,"eval/loss":0.6851214170455933,"train_loss":0.733072043210268,"_step":2449,"train_steps_per_second":0.173,"_wandb":{"runtime":80711},"train_runtime":80168.2654,"train/loss":0.5831656455993652,"eval/runtime":497.0682,"eval/samples_per_second":4.144,"train/grad_norm":1.1529462337493896,"eval/steps_per_second":4.144,"_runtime":80711,"train/global_step":4800,"train/learning_rate":8.08829405466816e-05,"train/epoch":2.0707659115426105}
\ No newline at end of file
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug-core.log b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..ebe69d7d712e3bd4e70f4e1b3b69361aadaa621f
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug-core.log
@@ -0,0 +1,14 @@
+{"time":"2025-12-23T14:22:35.996237193Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpv19qn5za/port-924716.txt","pid":924716,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-12-23T14:22:35.996809485Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":924716}
+{"time":"2025-12-23T14:22:35.996808097Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-924716-924825-960609214/socket","Net":"unix"}}
+{"time":"2025-12-23T14:22:36.183022345Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-12-23T14:22:36.190620051Z","level":"INFO","msg":"handleInformInit: received","streamId":"cktxoubm","id":"1(@)"}
+{"time":"2025-12-23T14:22:36.346686902Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"cktxoubm","id":"1(@)"}
+{"time":"2025-12-24T12:47:48.478245431Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-12-24T12:47:48.478346661Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-12-24T12:47:48.47834953Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-12-24T12:47:48.478427188Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-12-24T12:47:48.47850559Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-924716-924825-960609214/socket","Net":"unix"}}
+{"time":"2025-12-24T12:47:48.810776344Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-12-24T12:47:48.81080799Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-12-24T12:47:48.810823552Z","level":"INFO","msg":"server is closed"}
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug-internal.log b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..4a3e538e7581e1086374e9b27a45569ef0521436
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug-internal.log
@@ -0,0 +1,11 @@
+{"time":"2025-12-23T14:22:36.190806696Z","level":"INFO","msg":"stream: starting","core version":"0.23.1"}
+{"time":"2025-12-23T14:22:36.346431407Z","level":"INFO","msg":"stream: created new stream","id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346514258Z","level":"INFO","msg":"handler: started","stream_id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346675637Z","level":"INFO","msg":"stream: started","id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346701299Z","level":"INFO","msg":"writer: started","stream_id":"cktxoubm"}
+{"time":"2025-12-23T14:22:36.346725215Z","level":"INFO","msg":"sender: started","stream_id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.47834979Z","level":"INFO","msg":"stream: closing","id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.717701027Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2025-12-24T12:47:48.809988924Z","level":"INFO","msg":"handler: closed","stream_id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.810086512Z","level":"INFO","msg":"sender: closed","stream_id":"cktxoubm"}
+{"time":"2025-12-24T12:47:48.810099591Z","level":"INFO","msg":"stream: closed","id":"cktxoubm"}
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug.log b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..38dc783a2520c4ef56c29b4b4862fe4ba48f4669
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/logs/debug.log
@@ -0,0 +1,26 @@
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Current SDK version is 0.23.1
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Configure stats pid to 924716
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Loading settings from /root/.config/wandb/settings
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Loading settings from /workspace/Trainer-kit/wandb/settings
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_setup.py:_flush():80] Loading settings from environment variables
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:setup_run_log_directory():714] Logging user logs to runs/instruct_run_24b/wandb/run-20251223_142235-cktxoubm/logs/debug.log
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:setup_run_log_directory():715] Logging internal logs to runs/instruct_run_24b/wandb/run-20251223_142235-cktxoubm/logs/debug-internal.log
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:init():841] calling init triggers
+2025-12-23 14:22:35,732 INFO    MainThread:924716 [wandb_init.py:init():846] wandb.init called with sweep_config: {}
+config: {'model': {'repo_id': './CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'base_local_dir': 'base_model', 'trust_remote_code': True, 'tokenizer_use_fast': True, 'device_map': 'auto', 'torch_dtype': 'bfloat16', 'use_4bit': False, 'bnb_4bit_quant_type': 'nf4', 'bnb_4bit_use_double_quant': False, 'bnb_4bit_compute_dtype': 'bfloat16', 'attn_implementation': None}, 'data': {'train_jsonl': '../sft_dataset.jsonl', 'eval_jsonl': None, 'eval_split_ratio': 0.1, 'instruction_field': 'instruction', 'input_field': 'input', 'output_field': 'output', 'format_type': 'custom', 'system_prompt': 'You are a Hyperswitch Rust code analyzer. Identify functions/structs that need modification for a given task.\n\n## Output Format\n\n##OUTPUT\nExplain the data flow and why each component must change:\n- Flow: [Input → Processing → Output with arrows]\n- For each component: "The [ComponentName] ([path]) must [action] because [reason]—without this, [consequence]"\n- Explain coupling between components\n\n##SELECT\nmodify::crates/path/to/file.rs::impl::ComponentName\nadd::crates/another/file.rs::function::AnotherComponent\n<EOS>\n\n## Rules\n\n1. Use full paths: `remove::crates/folder/file.rs::Type::Name`\n2. Use `::` for nested items: `status::StructName::Type::Name`\n3. Always explain "must change because" and "without this"\n3. Types of components: function, struct, enum, impl, trait\n4. If there is extra information (e.g., enum variants), include that too.\n5. Start with ##OUTPUT, end with ##SELECT, terminate with <EOS>\n\n## Example\n\n##TASK\nAdd webhook subscription support\n\n##OUTPUT\nThe webhook system routes events via EventClass enum. Flow: webhook → EventClass → handler → processing. The EventClass enum (crates/common_enums/src/enums.rs::EventClass) must add Subscriptions variant because it defines event routing—without this, subscription events cannot be processed. The SubscriptionStatus impl (crates/common_enums/src/transformers.rs::SubscriptionStatus) must map to EventType because it converts status to events—without this, status changes don\'t trigger webhooks. These are coupled: EventClass routes to handlers that use SubscriptionStatus mappings.\n\n##SELECT\ncrates/common_enums/src/enums.rs::EventClass\ncrates/common_enums/src/transformers.rs::SubscriptionStatus\n<EOS>\n', 'custom_template': '##INSTRUCTION\n{instruction}<|im_end|>\n##TASK\n{input}<|im_end|>\n##OUTPUT\n{output}<|im_end|>', 'max_length': 2048, 'shuffle': True, 'num_proc': 4}, 'peft': {'enabled': True, 'r': 8, 'lora_alpha': 16, 'lora_dropout': 0.05, 'bias': 'none', 'target_modules': 'auto'}, 'train': {'num_train_epochs': 6, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'learning_rate': '1e-4', 'weight_decay': 0.0, 'warmup_ratio': 0.08, 'lr_scheduler_type': 'cosine', 'optim': 'adamw_torch', 'max_grad_norm': 0.8, 'gradient_checkpointing': True, 'logging_steps': 2, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'evaluation_strategy': 'steps', 'eval_steps': 100, 'load_best_model_at_end': True, 'early_stopping': {'enabled': True, 'patience': 3, 'min_delta': 0.001, 'metric': 'eval_loss', 'mode': 'min'}, 'resume_from_checkpoint': 'auto'}, 'run_dir': 'runs/instruct_run_24b', '_wandb': {}}
+2025-12-23 14:22:35,733 INFO    MainThread:924716 [wandb_init.py:init():889] starting backend
+2025-12-23 14:22:36,183 INFO    MainThread:924716 [wandb_init.py:init():892] sending inform_init request
+2025-12-23 14:22:36,187 INFO    MainThread:924716 [wandb_init.py:init():900] backend started and connected
+2025-12-23 14:22:36,189 INFO    MainThread:924716 [wandb_init.py:init():970] updated telemetry
+2025-12-23 14:22:36,193 INFO    MainThread:924716 [wandb_init.py:init():994] communicating run to backend with 90.0 second timeout
+2025-12-23 14:22:36,531 INFO    MainThread:924716 [wandb_init.py:init():1041] starting run threads in backend
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_console_start():2521] atexit reg
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_redirect():2369] redirect: wrap_raw
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_redirect():2438] Wrapping output streams.
+2025-12-23 14:22:36,634 INFO    MainThread:924716 [wandb_run.py:_redirect():2461] Redirects installed.
+2025-12-23 14:22:36,639 INFO    MainThread:924716 [wandb_init.py:init():1081] run started, returning control to user process
+2025-12-23 14:23:22,767 INFO    MainThread:924716 [wandb_run.py:_config_callback():1396] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.0', 'base_model_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'revision': None, 'inference_mode': False, 'r': 8, 'target_modules': ['v_proj', 'k_proj', 'q_proj', 'o_proj'], 'exclude_modules': None, 'lora_alpha': 16, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': None, 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False}}, 'image_token_index': 10, 'projector_hidden_act': 'gelu', 'vision_feature_layer': -1, 'vision_config': {'hidden_size': 1024, 'intermediate_size': 4096, 'num_hidden_layers': 24, 'num_attention_heads': 16, 'num_channels': 3, 'patch_size': 14, 'image_size': 1540, 'attention_dropout': 0.0, 'hidden_act': 'silu', 'head_dim': 64, 'initializer_range': 0.02, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'pixtral', 'output_attentions': False}, 'text_config': {'vocab_size': 131072, 'max_position_embeddings': 393216, 'hidden_size': 5120, 'intermediate_size': 32768, 'num_hidden_layers': 40, 'num_attention_heads': 32, 'sliding_window': None, 'head_dim': 128, 'num_key_value_heads': 8, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-05, 'use_cache': True, 'attention_dropout': 0.0, 'rope_parameters': {'beta_fast': 32.0, 'beta_slow': 1.0, 'factor': 48.0, 'llama_4_scaling_beta': 0.1, 'mscale': 1.0, 'mscale_all_dim': 1.0, 'original_max_position_embeddings': 8192, 'rope_theta': 100000000.0, 'rope_type': 'yarn', 'type': 'yarn'}, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 11, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': '', 'model_type': 'ministral3', 'output_attentions': False}, 'multimodal_projector_bias': False, 'spatial_merge_size': 2, 'return_dict': True, 'output_hidden_states': False, 'dtype': 'bfloat16', 'tie_word_embeddings': False, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'architectures': ['Mistral3ForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, '_name_or_path': 'CPT/runs/cpt_run_v1/merged_24b_cpt_lora', 'transformers_version': '5.0.0.dev0', 'model_type': 'mistral3', 'use_cache': False, 'output_attentions': False, 'output_dir': 'runs/instruct_run_24b/checkpoints', 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 1, 'per_device_eval_batch_size': 1, 'gradient_accumulation_steps': 8, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 0.0001, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 0.8, 'num_train_epochs': 6.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.08, 'warmup_steps': 0.08, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': None, 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 2, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 20, 'enable_jit_checkpoint': False, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'use_cpu': False, 'seed': 42, 'data_seed': None, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': -1, 'ddp_backend': None, 'debug': [], 'dataloader_drop_last': False, 'eval_steps': 100, 'dataloader_num_workers': 0, 'dataloader_prefetch_factor': None, 'run_name': None, 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': True, 'metric_for_best_model': 'eval_loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'auto_find_batch_size': False, 'full_determinism': False, 'ddp_timeout': 1800, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True}
+2025-12-23 14:23:22,778 INFO    MainThread:924716 [wandb_config.py:__setitem__():154] [no run ID] config set model/num_parameters = 24022764544 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x7f588665f580>>
+2025-12-23 14:23:22,779 INFO    MainThread:924716 [wandb_run.py:_config_callback():1396] config_cb model/num_parameters 24022764544 None
+2025-12-24 12:47:48,477 INFO    wandb-AsyncioManager-main:924716 [service_client.py:_forward_responses():80] Reached EOF.
+2025-12-24 12:47:48,478 INFO    wandb-AsyncioManager-main:924716 [mailbox.py:close():137] Closing mailbox, abandoning 1 handles.
diff --git a/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/run-cktxoubm.wandb b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/run-cktxoubm.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..a51307db1410b733be963b78b0bf70ed5a403b97
--- /dev/null
+++ b/sft_devstral_24B/wandb/run-20251223_142235-cktxoubm/run-cktxoubm.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f86cd347785241337118dce3dd05ea3c638be6d488cca04105eea087e46c7b1d
+size 45224266