Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,61 @@
+---
+library_name: transformers
+license: other
+base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: Proactive-Interactive-R1-SFT-7B
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# Proactive-Interactive-R1-SFT-7B
+
+This model is a fine-tuned version of [deepseek-ai/DeepSeek-R1-Distill-Qwen-7B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B) on the Proactive-Interactive-R1-SFT-7B dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- max_samples: 4000
+- learning_rate: 1e-05
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 32
+- total_eval_batch_size: 32
+- optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.03
+- num_epochs: 3
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.55.0
+- Pytorch 2.8.0+cu128
+- Datasets 3.6.0
+- Tokenizers 0.21.1

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 62777166004224.0,
+    "train_loss": 1.4248257904052735,
+    "train_runtime": 9087.8291,
+    "train_samples_per_second": 1.32,
+    "train_steps_per_second": 0.041
+}

chat_template.jinja

@@ -0,0 +1 @@
+{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set ns = namespace(is_first=false, is_tool=false, is_output_first=true, system_prompt='') %}{%- for message in messages %}{%- if message['role'] == 'system' %}{% set ns.system_prompt = message['content'] %}{%- endif %}{%- endfor %}{{bos_token}}{{ns.system_prompt}}{%- for message in messages %}{%- if message['role'] == 'user' %}{%- set ns.is_tool = false -%}{{'<｜User｜>' + message['content']}}{%- endif %}{%- if message['role'] == 'assistant' and message['content'] is none %}{%- set ns.is_tool = false -%}{%- for tool in message['tool_calls']%}{%- if not ns.is_first %}{{'<｜Assistant｜><｜tool▁calls▁begin｜><｜tool▁call▁begin｜>' + tool['type'] + '<｜tool▁sep｜>' + tool['function']['name'] + '\n' + '```json' + '\n' + tool['function']['arguments'] + '\n' + '```' + '<｜tool▁call▁end｜>'}}{%- set ns.is_first = true -%}{%- else %}{{'\n' + '<｜tool▁call▁begin｜>' + tool['type'] + '<｜tool▁sep｜>' + tool['function']['name'] + '\n' + '```json' + '\n' + tool['function']['arguments'] + '\n' + '```' + '<｜tool▁call▁end｜>'}}{{'<｜tool▁calls▁end｜><｜end▁of▁sentence｜>'}}{%- endif %}{%- endfor %}{%- endif %}{%- if message['role'] == 'assistant' and message['content'] is not none %}{%- if ns.is_tool %}{{'<｜tool▁outputs▁end｜>' + message['content'] + '<｜end▁of▁sentence｜>'}}{%- set ns.is_tool = false -%}{%- else %}{% set content = message['content'] %}{% if '</think>' in content %}{% set content = content.split('</think>')[-1] %}{% endif %}{{'<｜Assistant｜>' + content + '<｜end▁of▁sentence｜>'}}{%- endif %}{%- endif %}{%- if message['role'] == 'tool' %}{%- set ns.is_tool = true -%}{%- if ns.is_output_first %}{{'<｜tool▁outputs▁begin｜><｜tool▁output▁begin｜>' + message['content'] + '<｜tool▁output▁end｜>'}}{%- set ns.is_output_first = false %}{%- else %}{{'\n<｜tool▁output▁begin｜>' + message['content'] + '<｜tool▁output▁end｜>'}}{%- endif %}{%- endif %}{%- endfor -%}{% if ns.is_tool %}{{'<｜tool▁outputs▁end｜>'}}{% endif %}{% if add_generation_prompt and not ns.is_tool %}{{'<｜Assistant｜><think>\n'}}{% endif %}

config.json

@@ -0,0 +1,59 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 131072,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 10000,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.55.0",
+  "use_cache": false,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 151646,
+  "do_sample": true,
+  "eos_token_id": 151643,
+  "temperature": 0.6,
+  "top_p": 0.95,
+  "transformers_version": "4.55.0"
+}

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91d76e01cc39e9a5be9e51f64ff46395ca4a1d14e91d2afc16c34eef87f992ea
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4bafc232f862b482bff2d07a11dc62f7780f90e44a1afb13de3f8fde18905e5
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee8e907be773f54717181a73f0428b1c620c1d55ea171c78c55f7125f4d35f8c
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e6aef40c495a0e8ff87ebe0119bd2df314361e75057946be3a4da3f1c85b4b8
+size 1089994880

model.safetensors.index.json

@@ -0,0 +1,347 @@
+{
+  "metadata": {
+    "total_parameters": 333312,
+    "total_size": 15231233024
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00004-of-00004.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.norm.weight": "model-00003-of-00004.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<｜begin▁of▁sentence｜>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<｜end▁of▁sentence｜>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<｜end▁of▁sentence｜>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c432089e073c2d6ebc34e2a62d8cb5b4760009134644dd89dc8d413bcf5c3c11
+size 11423522

tokenizer_config.json

@@ -0,0 +1,228 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<｜end▁of▁sentence｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<｜User｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151645": {
+      "content": "<｜Assistant｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151646": {
+      "content": "<｜begin▁of▁sentence｜>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|EOT|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151648": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151649": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<asking>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</asking>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<response>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</response>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "bos_token": "<｜begin▁of▁sentence｜>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<｜end▁of▁sentence｜>",
+  "extra_special_tokens": {},
+  "legacy": true,
+  "model_max_length": 16384,
+  "pad_token": "<｜end▁of▁sentence｜>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "split_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizerFast",
+  "unk_token": null,
+  "use_default_system_prompt": false
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 62777166004224.0,
+    "train_loss": 1.4248257904052735,
+    "train_runtime": 9087.8291,
+    "train_samples_per_second": 1.32,
+    "train_steps_per_second": 0.041
+}

trainer_log.jsonl

@@ -0,0 +1,376 @@
+{"current_steps": 1, "total_steps": 375, "loss": 2.3147, "lr": 0.0, "epoch": 0.008, "percentage": 0.27, "elapsed_time": "0:00:36", "remaining_time": "3:44:43"}
+{"current_steps": 2, "total_steps": 375, "loss": 2.2195, "lr": 8.333333333333333e-07, "epoch": 0.016, "percentage": 0.53, "elapsed_time": "0:01:01", "remaining_time": "3:10:06"}
+{"current_steps": 3, "total_steps": 375, "loss": 2.3793, "lr": 1.6666666666666667e-06, "epoch": 0.024, "percentage": 0.8, "elapsed_time": "0:01:25", "remaining_time": "2:56:31"}
+{"current_steps": 4, "total_steps": 375, "loss": 2.2879, "lr": 2.5e-06, "epoch": 0.032, "percentage": 1.07, "elapsed_time": "0:01:50", "remaining_time": "2:50:15"}
+{"current_steps": 5, "total_steps": 375, "loss": 2.2428, "lr": 3.3333333333333333e-06, "epoch": 0.04, "percentage": 1.33, "elapsed_time": "0:02:14", "remaining_time": "2:45:33"}
+{"current_steps": 6, "total_steps": 375, "loss": 2.2197, "lr": 4.166666666666667e-06, "epoch": 0.048, "percentage": 1.6, "elapsed_time": "0:02:38", "remaining_time": "2:42:22"}
+{"current_steps": 7, "total_steps": 375, "loss": 2.0332, "lr": 5e-06, "epoch": 0.056, "percentage": 1.87, "elapsed_time": "0:03:03", "remaining_time": "2:40:56"}
+{"current_steps": 8, "total_steps": 375, "loss": 2.1247, "lr": 5.833333333333334e-06, "epoch": 0.064, "percentage": 2.13, "elapsed_time": "0:03:30", "remaining_time": "2:40:49"}
+{"current_steps": 9, "total_steps": 375, "loss": 2.1694, "lr": 6.666666666666667e-06, "epoch": 0.072, "percentage": 2.4, "elapsed_time": "0:03:54", "remaining_time": "2:38:43"}
+{"current_steps": 10, "total_steps": 375, "loss": 2.1108, "lr": 7.500000000000001e-06, "epoch": 0.08, "percentage": 2.67, "elapsed_time": "0:04:17", "remaining_time": "2:36:41"}
+{"current_steps": 11, "total_steps": 375, "loss": 2.0756, "lr": 8.333333333333334e-06, "epoch": 0.088, "percentage": 2.93, "elapsed_time": "0:04:41", "remaining_time": "2:35:08"}
+{"current_steps": 12, "total_steps": 375, "loss": 2.0886, "lr": 9.166666666666666e-06, "epoch": 0.096, "percentage": 3.2, "elapsed_time": "0:05:05", "remaining_time": "2:33:54"}
+{"current_steps": 13, "total_steps": 375, "loss": 1.9143, "lr": 1e-05, "epoch": 0.104, "percentage": 3.47, "elapsed_time": "0:05:30", "remaining_time": "2:33:24"}
+{"current_steps": 14, "total_steps": 375, "loss": 1.7421, "lr": 9.999812749151968e-06, "epoch": 0.112, "percentage": 3.73, "elapsed_time": "0:05:54", "remaining_time": "2:32:15"}
+{"current_steps": 15, "total_steps": 375, "loss": 1.9555, "lr": 9.99925101063302e-06, "epoch": 0.12, "percentage": 4.0, "elapsed_time": "0:06:17", "remaining_time": "2:30:56"}
+{"current_steps": 16, "total_steps": 375, "loss": 1.7679, "lr": 9.998314826517564e-06, "epoch": 0.128, "percentage": 4.27, "elapsed_time": "0:06:41", "remaining_time": "2:30:18"}
+{"current_steps": 17, "total_steps": 375, "loss": 1.8436, "lr": 9.997004266926105e-06, "epoch": 0.136, "percentage": 4.53, "elapsed_time": "0:07:05", "remaining_time": "2:29:25"}
+{"current_steps": 18, "total_steps": 375, "loss": 1.6908, "lr": 9.995319430020004e-06, "epoch": 0.144, "percentage": 4.8, "elapsed_time": "0:07:31", "remaining_time": "2:29:18"}
+{"current_steps": 19, "total_steps": 375, "loss": 1.8093, "lr": 9.993260441994116e-06, "epoch": 0.152, "percentage": 5.07, "elapsed_time": "0:07:55", "remaining_time": "2:28:27"}
+{"current_steps": 20, "total_steps": 375, "loss": 1.8674, "lr": 9.990827457067342e-06, "epoch": 0.16, "percentage": 5.33, "elapsed_time": "0:08:18", "remaining_time": "2:27:25"}
+{"current_steps": 21, "total_steps": 375, "loss": 1.8742, "lr": 9.988020657471078e-06, "epoch": 0.168, "percentage": 5.6, "elapsed_time": "0:08:40", "remaining_time": "2:26:11"}
+{"current_steps": 22, "total_steps": 375, "loss": 1.7627, "lr": 9.984840253435569e-06, "epoch": 0.176, "percentage": 5.87, "elapsed_time": "0:09:03", "remaining_time": "2:25:21"}
+{"current_steps": 23, "total_steps": 375, "loss": 1.7632, "lr": 9.98128648317415e-06, "epoch": 0.184, "percentage": 6.13, "elapsed_time": "0:09:30", "remaining_time": "2:25:24"}
+{"current_steps": 24, "total_steps": 375, "loss": 1.7992, "lr": 9.977359612865424e-06, "epoch": 0.192, "percentage": 6.4, "elapsed_time": "0:09:55", "remaining_time": "2:25:06"}
+{"current_steps": 25, "total_steps": 375, "loss": 1.7215, "lr": 9.973059936633308e-06, "epoch": 0.2, "percentage": 6.67, "elapsed_time": "0:10:21", "remaining_time": "2:25:02"}
+{"current_steps": 26, "total_steps": 375, "loss": 1.7456, "lr": 9.968387776525009e-06, "epoch": 0.208, "percentage": 6.93, "elapsed_time": "0:10:43", "remaining_time": "2:24:01"}
+{"current_steps": 27, "total_steps": 375, "loss": 1.7512, "lr": 9.963343482486907e-06, "epoch": 0.216, "percentage": 7.2, "elapsed_time": "0:11:05", "remaining_time": "2:23:03"}
+{"current_steps": 28, "total_steps": 375, "loss": 1.801, "lr": 9.957927432338332e-06, "epoch": 0.224, "percentage": 7.47, "elapsed_time": "0:11:28", "remaining_time": "2:22:16"}
+{"current_steps": 29, "total_steps": 375, "loss": 1.7266, "lr": 9.952140031743282e-06, "epoch": 0.232, "percentage": 7.73, "elapsed_time": "0:11:52", "remaining_time": "2:21:46"}
+{"current_steps": 30, "total_steps": 375, "loss": 1.8852, "lr": 9.945981714180021e-06, "epoch": 0.24, "percentage": 8.0, "elapsed_time": "0:12:15", "remaining_time": "2:20:53"}
+{"current_steps": 31, "total_steps": 375, "loss": 1.6649, "lr": 9.939452940908627e-06, "epoch": 0.248, "percentage": 8.27, "elapsed_time": "0:12:36", "remaining_time": "2:19:51"}
+{"current_steps": 32, "total_steps": 375, "loss": 1.7861, "lr": 9.932554200936428e-06, "epoch": 0.256, "percentage": 8.53, "elapsed_time": "0:12:59", "remaining_time": "2:19:18"}
+{"current_steps": 33, "total_steps": 375, "loss": 1.7849, "lr": 9.925286010981394e-06, "epoch": 0.264, "percentage": 8.8, "elapsed_time": "0:13:23", "remaining_time": "2:18:42"}
+{"current_steps": 34, "total_steps": 375, "loss": 1.8268, "lr": 9.917648915433413e-06, "epoch": 0.272, "percentage": 9.07, "elapsed_time": "0:13:48", "remaining_time": "2:18:27"}
+{"current_steps": 35, "total_steps": 375, "loss": 1.7176, "lr": 9.909643486313533e-06, "epoch": 0.28, "percentage": 9.33, "elapsed_time": "0:14:13", "remaining_time": "2:18:13"}
+{"current_steps": 36, "total_steps": 375, "loss": 1.7909, "lr": 9.901270323231114e-06, "epoch": 0.288, "percentage": 9.6, "elapsed_time": "0:14:38", "remaining_time": "2:17:52"}
+{"current_steps": 37, "total_steps": 375, "loss": 1.7618, "lr": 9.892530053338909e-06, "epoch": 0.296, "percentage": 9.87, "elapsed_time": "0:15:03", "remaining_time": "2:17:33"}
+{"current_steps": 38, "total_steps": 375, "loss": 1.7316, "lr": 9.883423331286096e-06, "epoch": 0.304, "percentage": 10.13, "elapsed_time": "0:15:27", "remaining_time": "2:17:02"}
+{"current_steps": 39, "total_steps": 375, "loss": 1.6342, "lr": 9.873950839169248e-06, "epoch": 0.312, "percentage": 10.4, "elapsed_time": "0:15:52", "remaining_time": "2:16:50"}
+{"current_steps": 40, "total_steps": 375, "loss": 1.7302, "lr": 9.864113286481237e-06, "epoch": 0.32, "percentage": 10.67, "elapsed_time": "0:16:18", "remaining_time": "2:16:32"}
+{"current_steps": 41, "total_steps": 375, "loss": 1.5871, "lr": 9.853911410058097e-06, "epoch": 0.328, "percentage": 10.93, "elapsed_time": "0:16:43", "remaining_time": "2:16:14"}
+{"current_steps": 42, "total_steps": 375, "loss": 1.7055, "lr": 9.843345974023833e-06, "epoch": 0.336, "percentage": 11.2, "elapsed_time": "0:17:10", "remaining_time": "2:16:11"}
+{"current_steps": 43, "total_steps": 375, "loss": 1.725, "lr": 9.832417769733185e-06, "epoch": 0.344, "percentage": 11.47, "elapsed_time": "0:17:33", "remaining_time": "2:15:31"}
+{"current_steps": 44, "total_steps": 375, "loss": 1.7716, "lr": 9.821127615712365e-06, "epoch": 0.352, "percentage": 11.73, "elapsed_time": "0:17:56", "remaining_time": "2:14:58"}
+{"current_steps": 45, "total_steps": 375, "loss": 1.6597, "lr": 9.809476357597738e-06, "epoch": 0.36, "percentage": 12.0, "elapsed_time": "0:18:23", "remaining_time": "2:14:53"}
+{"current_steps": 46, "total_steps": 375, "loss": 1.7454, "lr": 9.797464868072489e-06, "epoch": 0.368, "percentage": 12.27, "elapsed_time": "0:18:48", "remaining_time": "2:14:32"}
+{"current_steps": 47, "total_steps": 375, "loss": 1.813, "lr": 9.785094046801256e-06, "epoch": 0.376, "percentage": 12.53, "elapsed_time": "0:19:11", "remaining_time": "2:13:55"}
+{"current_steps": 48, "total_steps": 375, "loss": 1.6276, "lr": 9.77236482036275e-06, "epoch": 0.384, "percentage": 12.8, "elapsed_time": "0:19:35", "remaining_time": "2:13:26"}
+{"current_steps": 49, "total_steps": 375, "loss": 1.6772, "lr": 9.759278142180348e-06, "epoch": 0.392, "percentage": 13.07, "elapsed_time": "0:19:57", "remaining_time": "2:12:48"}
+{"current_steps": 50, "total_steps": 375, "loss": 1.7564, "lr": 9.745834992450688e-06, "epoch": 0.4, "percentage": 13.33, "elapsed_time": "0:20:19", "remaining_time": "2:12:08"}
+{"current_steps": 51, "total_steps": 375, "loss": 1.7262, "lr": 9.732036378070243e-06, "epoch": 0.408, "percentage": 13.6, "elapsed_time": "0:20:44", "remaining_time": "2:11:44"}
+{"current_steps": 52, "total_steps": 375, "loss": 1.5396, "lr": 9.717883332559911e-06, "epoch": 0.416, "percentage": 13.87, "elapsed_time": "0:21:08", "remaining_time": "2:11:20"}
+{"current_steps": 53, "total_steps": 375, "loss": 1.7466, "lr": 9.703376915987601e-06, "epoch": 0.424, "percentage": 14.13, "elapsed_time": "0:21:32", "remaining_time": "2:10:53"}
+{"current_steps": 54, "total_steps": 375, "loss": 1.6408, "lr": 9.688518214888836e-06, "epoch": 0.432, "percentage": 14.4, "elapsed_time": "0:21:56", "remaining_time": "2:10:26"}
+{"current_steps": 55, "total_steps": 375, "loss": 1.5324, "lr": 9.673308342185366e-06, "epoch": 0.44, "percentage": 14.67, "elapsed_time": "0:22:22", "remaining_time": "2:10:09"}
+{"current_steps": 56, "total_steps": 375, "loss": 1.5823, "lr": 9.657748437101819e-06, "epoch": 0.448, "percentage": 14.93, "elapsed_time": "0:22:44", "remaining_time": "2:09:31"}
+{"current_steps": 57, "total_steps": 375, "loss": 1.7311, "lr": 9.641839665080363e-06, "epoch": 0.456, "percentage": 15.2, "elapsed_time": "0:23:06", "remaining_time": "2:08:53"}
+{"current_steps": 58, "total_steps": 375, "loss": 1.8164, "lr": 9.625583217693419e-06, "epoch": 0.464, "percentage": 15.47, "elapsed_time": "0:23:30", "remaining_time": "2:08:27"}
+{"current_steps": 59, "total_steps": 375, "loss": 1.6059, "lr": 9.60898031255441e-06, "epoch": 0.472, "percentage": 15.73, "elapsed_time": "0:23:54", "remaining_time": "2:08:05"}
+{"current_steps": 60, "total_steps": 375, "loss": 1.7617, "lr": 9.592032193226564e-06, "epoch": 0.48, "percentage": 16.0, "elapsed_time": "0:24:19", "remaining_time": "2:07:40"}
+{"current_steps": 61, "total_steps": 375, "loss": 1.7207, "lr": 9.574740129129767e-06, "epoch": 0.488, "percentage": 16.27, "elapsed_time": "0:24:43", "remaining_time": "2:07:18"}
+{"current_steps": 62, "total_steps": 375, "loss": 1.7482, "lr": 9.557105415445485e-06, "epoch": 0.496, "percentage": 16.53, "elapsed_time": "0:25:07", "remaining_time": "2:06:49"}
+{"current_steps": 63, "total_steps": 375, "loss": 1.7692, "lr": 9.539129373019755e-06, "epoch": 0.504, "percentage": 16.8, "elapsed_time": "0:25:32", "remaining_time": "2:06:29"}
+{"current_steps": 64, "total_steps": 375, "loss": 1.737, "lr": 9.520813348264252e-06, "epoch": 0.512, "percentage": 17.07, "elapsed_time": "0:25:54", "remaining_time": "2:05:55"}
+{"current_steps": 65, "total_steps": 375, "loss": 1.6523, "lr": 9.502158713055444e-06, "epoch": 0.52, "percentage": 17.33, "elapsed_time": "0:26:18", "remaining_time": "2:05:28"}
+{"current_steps": 66, "total_steps": 375, "loss": 1.7734, "lr": 9.483166864631837e-06, "epoch": 0.528, "percentage": 17.6, "elapsed_time": "0:26:40", "remaining_time": "2:04:52"}
+{"current_steps": 67, "total_steps": 375, "loss": 1.6183, "lr": 9.46383922548932e-06, "epoch": 0.536, "percentage": 17.87, "elapsed_time": "0:27:05", "remaining_time": "2:04:31"}
+{"current_steps": 68, "total_steps": 375, "loss": 1.6967, "lr": 9.444177243274619e-06, "epoch": 0.544, "percentage": 18.13, "elapsed_time": "0:27:28", "remaining_time": "2:04:03"}
+{"current_steps": 69, "total_steps": 375, "loss": 1.5285, "lr": 9.424182390676872e-06, "epoch": 0.552, "percentage": 18.4, "elapsed_time": "0:27:53", "remaining_time": "2:03:41"}
+{"current_steps": 70, "total_steps": 375, "loss": 1.7399, "lr": 9.403856165317322e-06, "epoch": 0.56, "percentage": 18.67, "elapsed_time": "0:28:18", "remaining_time": "2:03:19"}
+{"current_steps": 71, "total_steps": 375, "loss": 1.6845, "lr": 9.383200089637143e-06, "epoch": 0.568, "percentage": 18.93, "elapsed_time": "0:28:43", "remaining_time": "2:02:57"}
+{"current_steps": 72, "total_steps": 375, "loss": 1.6052, "lr": 9.362215710783411e-06, "epoch": 0.576, "percentage": 19.2, "elapsed_time": "0:29:06", "remaining_time": "2:02:31"}
+{"current_steps": 73, "total_steps": 375, "loss": 1.6624, "lr": 9.34090460049322e-06, "epoch": 0.584, "percentage": 19.47, "elapsed_time": "0:29:30", "remaining_time": "2:02:05"}
+{"current_steps": 74, "total_steps": 375, "loss": 1.581, "lr": 9.319268354975958e-06, "epoch": 0.592, "percentage": 19.73, "elapsed_time": "0:29:54", "remaining_time": "2:01:41"}
+{"current_steps": 75, "total_steps": 375, "loss": 1.6495, "lr": 9.297308594793757e-06, "epoch": 0.6, "percentage": 20.0, "elapsed_time": "0:30:17", "remaining_time": "2:01:10"}
+{"current_steps": 76, "total_steps": 375, "loss": 1.6056, "lr": 9.275026964740101e-06, "epoch": 0.608, "percentage": 20.27, "elapsed_time": "0:30:42", "remaining_time": "2:00:50"}
+{"current_steps": 77, "total_steps": 375, "loss": 1.679, "lr": 9.252425133716639e-06, "epoch": 0.616, "percentage": 20.53, "elapsed_time": "0:31:05", "remaining_time": "2:00:21"}
+{"current_steps": 78, "total_steps": 375, "loss": 1.6168, "lr": 9.229504794608182e-06, "epoch": 0.624, "percentage": 20.8, "elapsed_time": "0:31:32", "remaining_time": "2:00:07"}
+{"current_steps": 79, "total_steps": 375, "loss": 1.6812, "lr": 9.206267664155906e-06, "epoch": 0.632, "percentage": 21.07, "elapsed_time": "0:32:00", "remaining_time": "1:59:54"}
+{"current_steps": 80, "total_steps": 375, "loss": 1.6214, "lr": 9.182715482828764e-06, "epoch": 0.64, "percentage": 21.33, "elapsed_time": "0:32:24", "remaining_time": "1:59:29"}
+{"current_steps": 81, "total_steps": 375, "loss": 1.6537, "lr": 9.158850014693123e-06, "epoch": 0.648, "percentage": 21.6, "elapsed_time": "0:32:45", "remaining_time": "1:58:54"}
+{"current_steps": 82, "total_steps": 375, "loss": 1.602, "lr": 9.134673047280644e-06, "epoch": 0.656, "percentage": 21.87, "elapsed_time": "0:33:09", "remaining_time": "1:58:28"}
+{"current_steps": 83, "total_steps": 375, "loss": 1.7139, "lr": 9.110186391454389e-06, "epoch": 0.664, "percentage": 22.13, "elapsed_time": "0:33:32", "remaining_time": "1:58:00"}
+{"current_steps": 84, "total_steps": 375, "loss": 1.6621, "lr": 9.085391881273182e-06, "epoch": 0.672, "percentage": 22.4, "elapsed_time": "0:33:56", "remaining_time": "1:57:34"}
+{"current_steps": 85, "total_steps": 375, "loss": 1.6323, "lr": 9.060291373854252e-06, "epoch": 0.68, "percentage": 22.67, "elapsed_time": "0:34:20", "remaining_time": "1:57:10"}
+{"current_steps": 86, "total_steps": 375, "loss": 1.6435, "lr": 9.034886749234112e-06, "epoch": 0.688, "percentage": 22.93, "elapsed_time": "0:34:44", "remaining_time": "1:56:45"}
+{"current_steps": 87, "total_steps": 375, "loss": 1.6756, "lr": 9.009179910227767e-06, "epoch": 0.696, "percentage": 23.2, "elapsed_time": "0:35:08", "remaining_time": "1:56:19"}
+{"current_steps": 88, "total_steps": 375, "loss": 1.6813, "lr": 8.98317278228618e-06, "epoch": 0.704, "percentage": 23.47, "elapsed_time": "0:35:32", "remaining_time": "1:55:54"}
+{"current_steps": 89, "total_steps": 375, "loss": 1.6627, "lr": 8.956867313352055e-06, "epoch": 0.712, "percentage": 23.73, "elapsed_time": "0:35:56", "remaining_time": "1:55:29"}
+{"current_steps": 90, "total_steps": 375, "loss": 1.8571, "lr": 8.930265473713939e-06, "epoch": 0.72, "percentage": 24.0, "elapsed_time": "0:36:19", "remaining_time": "1:55:03"}
+{"current_steps": 91, "total_steps": 375, "loss": 1.4989, "lr": 8.90336925585864e-06, "epoch": 0.728, "percentage": 24.27, "elapsed_time": "0:36:43", "remaining_time": "1:54:36"}
+{"current_steps": 92, "total_steps": 375, "loss": 1.6913, "lr": 8.876180674322006e-06, "epoch": 0.736, "percentage": 24.53, "elapsed_time": "0:37:06", "remaining_time": "1:54:09"}
+{"current_steps": 93, "total_steps": 375, "loss": 1.6397, "lr": 8.84870176553801e-06, "epoch": 0.744, "percentage": 24.8, "elapsed_time": "0:37:28", "remaining_time": "1:53:38"}
+{"current_steps": 94, "total_steps": 375, "loss": 1.5152, "lr": 8.820934587686247e-06, "epoch": 0.752, "percentage": 25.07, "elapsed_time": "0:37:54", "remaining_time": "1:53:19"}
+{"current_steps": 95, "total_steps": 375, "loss": 1.5435, "lr": 8.792881220537752e-06, "epoch": 0.76, "percentage": 25.33, "elapsed_time": "0:38:19", "remaining_time": "1:52:57"}
+{"current_steps": 96, "total_steps": 375, "loss": 1.5283, "lr": 8.764543765299245e-06, "epoch": 0.768, "percentage": 25.6, "elapsed_time": "0:38:42", "remaining_time": "1:52:28"}
+{"current_steps": 97, "total_steps": 375, "loss": 1.6756, "lr": 8.735924344455732e-06, "epoch": 0.776, "percentage": 25.87, "elapsed_time": "0:39:04", "remaining_time": "1:51:59"}
+{"current_steps": 98, "total_steps": 375, "loss": 1.613, "lr": 8.707025101611546e-06, "epoch": 0.784, "percentage": 26.13, "elapsed_time": "0:39:29", "remaining_time": "1:51:37"}
+{"current_steps": 99, "total_steps": 375, "loss": 1.5942, "lr": 8.677848201329775e-06, "epoch": 0.792, "percentage": 26.4, "elapsed_time": "0:39:55", "remaining_time": "1:51:19"}
+{"current_steps": 100, "total_steps": 375, "loss": 1.7216, "lr": 8.64839582897015e-06, "epoch": 0.8, "percentage": 26.67, "elapsed_time": "0:40:19", "remaining_time": "1:50:52"}
+{"current_steps": 101, "total_steps": 375, "loss": 1.4699, "lr": 8.61867019052535e-06, "epoch": 0.808, "percentage": 26.93, "elapsed_time": "0:40:44", "remaining_time": "1:50:30"}
+{"current_steps": 102, "total_steps": 375, "loss": 1.7439, "lr": 8.588673512455781e-06, "epoch": 0.816, "percentage": 27.2, "elapsed_time": "0:41:08", "remaining_time": "1:50:06"}
+{"current_steps": 103, "total_steps": 375, "loss": 1.696, "lr": 8.558408041522801e-06, "epoch": 0.824, "percentage": 27.47, "elapsed_time": "0:41:32", "remaining_time": "1:49:41"}
+{"current_steps": 104, "total_steps": 375, "loss": 1.713, "lr": 8.527876044620453e-06, "epoch": 0.832, "percentage": 27.73, "elapsed_time": "0:41:56", "remaining_time": "1:49:16"}
+{"current_steps": 105, "total_steps": 375, "loss": 1.7213, "lr": 8.497079808605659e-06, "epoch": 0.84, "percentage": 28.0, "elapsed_time": "0:42:20", "remaining_time": "1:48:51"}
+{"current_steps": 106, "total_steps": 375, "loss": 1.6091, "lr": 8.466021640126946e-06, "epoch": 0.848, "percentage": 28.27, "elapsed_time": "0:42:45", "remaining_time": "1:48:30"}
+{"current_steps": 107, "total_steps": 375, "loss": 1.6345, "lr": 8.434703865451666e-06, "epoch": 0.856, "percentage": 28.53, "elapsed_time": "0:43:10", "remaining_time": "1:48:09"}
+{"current_steps": 108, "total_steps": 375, "loss": 1.6443, "lr": 8.403128830291767e-06, "epoch": 0.864, "percentage": 28.8, "elapsed_time": "0:43:36", "remaining_time": "1:47:48"}
+{"current_steps": 109, "total_steps": 375, "loss": 1.6869, "lr": 8.371298899628091e-06, "epoch": 0.872, "percentage": 29.07, "elapsed_time": "0:43:59", "remaining_time": "1:47:21"}
+{"current_steps": 110, "total_steps": 375, "loss": 1.5866, "lr": 8.339216457533244e-06, "epoch": 0.88, "percentage": 29.33, "elapsed_time": "0:44:22", "remaining_time": "1:46:54"}
+{"current_steps": 111, "total_steps": 375, "loss": 1.6483, "lr": 8.306883906993022e-06, "epoch": 0.888, "percentage": 29.6, "elapsed_time": "0:44:45", "remaining_time": "1:46:28"}
+{"current_steps": 112, "total_steps": 375, "loss": 1.594, "lr": 8.274303669726427e-06, "epoch": 0.896, "percentage": 29.87, "elapsed_time": "0:45:08", "remaining_time": "1:46:01"}
+{"current_steps": 113, "total_steps": 375, "loss": 1.4755, "lr": 8.24147818600428e-06, "epoch": 0.904, "percentage": 30.13, "elapsed_time": "0:45:32", "remaining_time": "1:45:36"}
+{"current_steps": 114, "total_steps": 375, "loss": 1.6598, "lr": 8.20840991446645e-06, "epoch": 0.912, "percentage": 30.4, "elapsed_time": "0:45:57", "remaining_time": "1:45:12"}
+{"current_steps": 115, "total_steps": 375, "loss": 1.5304, "lr": 8.175101331937692e-06, "epoch": 0.92, "percentage": 30.67, "elapsed_time": "0:46:21", "remaining_time": "1:44:47"}
+{"current_steps": 116, "total_steps": 375, "loss": 1.6354, "lr": 8.141554933242135e-06, "epoch": 0.928, "percentage": 30.93, "elapsed_time": "0:46:46", "remaining_time": "1:44:26"}
+{"current_steps": 117, "total_steps": 375, "loss": 1.5989, "lr": 8.10777323101642e-06, "epoch": 0.936, "percentage": 31.2, "elapsed_time": "0:47:10", "remaining_time": "1:44:01"}
+{"current_steps": 118, "total_steps": 375, "loss": 1.6564, "lr": 8.073758755521506e-06, "epoch": 0.944, "percentage": 31.47, "elapsed_time": "0:47:32", "remaining_time": "1:43:33"}
+{"current_steps": 119, "total_steps": 375, "loss": 1.5884, "lr": 8.03951405445314e-06, "epoch": 0.952, "percentage": 31.73, "elapsed_time": "0:47:59", "remaining_time": "1:43:14"}
+{"current_steps": 120, "total_steps": 375, "loss": 1.5128, "lr": 8.005041692751055e-06, "epoch": 0.96, "percentage": 32.0, "elapsed_time": "0:48:24", "remaining_time": "1:42:51"}
+{"current_steps": 121, "total_steps": 375, "loss": 1.7014, "lr": 7.970344252406832e-06, "epoch": 0.968, "percentage": 32.27, "elapsed_time": "0:48:46", "remaining_time": "1:42:24"}
+{"current_steps": 122, "total_steps": 375, "loss": 1.6341, "lr": 7.935424332270523e-06, "epoch": 0.976, "percentage": 32.53, "elapsed_time": "0:49:11", "remaining_time": "1:41:59"}
+{"current_steps": 123, "total_steps": 375, "loss": 1.6521, "lr": 7.900284547855992e-06, "epoch": 0.984, "percentage": 32.8, "elapsed_time": "0:49:35", "remaining_time": "1:41:35"}
+{"current_steps": 124, "total_steps": 375, "loss": 1.7013, "lr": 7.864927531145012e-06, "epoch": 0.992, "percentage": 33.07, "elapsed_time": "0:49:58", "remaining_time": "1:41:09"}
+{"current_steps": 125, "total_steps": 375, "loss": 1.5216, "lr": 7.829355930390126e-06, "epoch": 1.0, "percentage": 33.33, "elapsed_time": "0:50:23", "remaining_time": "1:40:46"}
+{"current_steps": 126, "total_steps": 375, "loss": 1.4109, "lr": 7.7935724099163e-06, "epoch": 1.008, "percentage": 33.6, "elapsed_time": "0:50:56", "remaining_time": "1:40:40"}
+{"current_steps": 127, "total_steps": 375, "loss": 1.4167, "lr": 7.757579649921354e-06, "epoch": 1.016, "percentage": 33.87, "elapsed_time": "0:51:20", "remaining_time": "1:40:16"}
+{"current_steps": 128, "total_steps": 375, "loss": 1.4185, "lr": 7.721380346275221e-06, "epoch": 1.024, "percentage": 34.13, "elapsed_time": "0:51:44", "remaining_time": "1:39:50"}
+{"current_steps": 129, "total_steps": 375, "loss": 1.5076, "lr": 7.684977210318024e-06, "epoch": 1.032, "percentage": 34.4, "elapsed_time": "0:52:07", "remaining_time": "1:39:24"}
+{"current_steps": 130, "total_steps": 375, "loss": 1.4809, "lr": 7.648372968656995e-06, "epoch": 1.04, "percentage": 34.67, "elapsed_time": "0:52:31", "remaining_time": "1:38:59"}
+{"current_steps": 131, "total_steps": 375, "loss": 1.4613, "lr": 7.611570362962247e-06, "epoch": 1.048, "percentage": 34.93, "elapsed_time": "0:52:55", "remaining_time": "1:38:34"}
+{"current_steps": 132, "total_steps": 375, "loss": 1.3367, "lr": 7.574572149761437e-06, "epoch": 1.056, "percentage": 35.2, "elapsed_time": "0:53:19", "remaining_time": "1:38:09"}
+{"current_steps": 133, "total_steps": 375, "loss": 1.3333, "lr": 7.5373811002332785e-06, "epoch": 1.064, "percentage": 35.47, "elapsed_time": "0:53:40", "remaining_time": "1:37:40"}
+{"current_steps": 134, "total_steps": 375, "loss": 1.2811, "lr": 7.500000000000001e-06, "epoch": 1.072, "percentage": 35.73, "elapsed_time": "0:54:06", "remaining_time": "1:37:18"}
+{"current_steps": 135, "total_steps": 375, "loss": 1.2576, "lr": 7.462431648918689e-06, "epoch": 1.08, "percentage": 36.0, "elapsed_time": "0:54:31", "remaining_time": "1:36:55"}
+{"current_steps": 136, "total_steps": 375, "loss": 1.3759, "lr": 7.424678860871584e-06, "epoch": 1.088, "percentage": 36.27, "elapsed_time": "0:54:57", "remaining_time": "1:36:35"}
+{"current_steps": 137, "total_steps": 375, "loss": 1.2774, "lr": 7.3867444635553165e-06, "epoch": 1.096, "percentage": 36.53, "elapsed_time": "0:55:22", "remaining_time": "1:36:11"}
+{"current_steps": 138, "total_steps": 375, "loss": 1.346, "lr": 7.3486312982691134e-06, "epoch": 1.104, "percentage": 36.8, "elapsed_time": "0:55:46", "remaining_time": "1:35:46"}
+{"current_steps": 139, "total_steps": 375, "loss": 1.3977, "lr": 7.310342219701981e-06, "epoch": 1.112, "percentage": 37.07, "elapsed_time": "0:56:11", "remaining_time": "1:35:23"}
+{"current_steps": 140, "total_steps": 375, "loss": 1.3203, "lr": 7.271880095718895e-06, "epoch": 1.12, "percentage": 37.33, "elapsed_time": "0:56:38", "remaining_time": "1:35:04"}
+{"current_steps": 141, "total_steps": 375, "loss": 1.3563, "lr": 7.233247807145989e-06, "epoch": 1.1280000000000001, "percentage": 37.6, "elapsed_time": "0:57:02", "remaining_time": "1:34:40"}
+{"current_steps": 142, "total_steps": 375, "loss": 1.4031, "lr": 7.19444824755478e-06, "epoch": 1.1360000000000001, "percentage": 37.87, "elapsed_time": "0:57:30", "remaining_time": "1:34:21"}
+{"current_steps": 143, "total_steps": 375, "loss": 1.3123, "lr": 7.155484323045442e-06, "epoch": 1.144, "percentage": 38.13, "elapsed_time": "0:57:54", "remaining_time": "1:33:56"}
+{"current_steps": 144, "total_steps": 375, "loss": 1.3359, "lr": 7.11635895202914e-06, "epoch": 1.152, "percentage": 38.4, "elapsed_time": "0:58:17", "remaining_time": "1:33:31"}
+{"current_steps": 145, "total_steps": 375, "loss": 1.4203, "lr": 7.0770750650094335e-06, "epoch": 1.16, "percentage": 38.67, "elapsed_time": "0:58:42", "remaining_time": "1:33:07"}
+{"current_steps": 146, "total_steps": 375, "loss": 1.3475, "lr": 7.037635604362786e-06, "epoch": 1.168, "percentage": 38.93, "elapsed_time": "0:59:05", "remaining_time": "1:32:40"}
+{"current_steps": 147, "total_steps": 375, "loss": 1.3728, "lr": 6.9980435241181785e-06, "epoch": 1.176, "percentage": 39.2, "elapsed_time": "0:59:27", "remaining_time": "1:32:13"}
+{"current_steps": 148, "total_steps": 375, "loss": 1.4341, "lr": 6.958301789735853e-06, "epoch": 1.184, "percentage": 39.47, "elapsed_time": "0:59:49", "remaining_time": "1:31:45"}
+{"current_steps": 149, "total_steps": 375, "loss": 1.4365, "lr": 6.918413377885193e-06, "epoch": 1.192, "percentage": 39.73, "elapsed_time": "1:00:13", "remaining_time": "1:31:21"}
+{"current_steps": 150, "total_steps": 375, "loss": 1.3273, "lr": 6.878381276221777e-06, "epoch": 1.2, "percentage": 40.0, "elapsed_time": "1:00:38", "remaining_time": "1:30:58"}
+{"current_steps": 151, "total_steps": 375, "loss": 1.3297, "lr": 6.838208483163601e-06, "epoch": 1.208, "percentage": 40.27, "elapsed_time": "1:01:03", "remaining_time": "1:30:34"}
+{"current_steps": 152, "total_steps": 375, "loss": 1.3977, "lr": 6.797898007666493e-06, "epoch": 1.216, "percentage": 40.53, "elapsed_time": "1:01:25", "remaining_time": "1:30:06"}
+{"current_steps": 153, "total_steps": 375, "loss": 1.2802, "lr": 6.757452868998737e-06, "epoch": 1.224, "percentage": 40.8, "elapsed_time": "1:01:49", "remaining_time": "1:29:41"}
+{"current_steps": 154, "total_steps": 375, "loss": 1.3741, "lr": 6.716876096514944e-06, "epoch": 1.232, "percentage": 41.07, "elapsed_time": "1:02:13", "remaining_time": "1:29:17"}
+{"current_steps": 155, "total_steps": 375, "loss": 1.2866, "lr": 6.676170729429132e-06, "epoch": 1.24, "percentage": 41.33, "elapsed_time": "1:02:35", "remaining_time": "1:28:50"}
+{"current_steps": 156, "total_steps": 375, "loss": 1.3524, "lr": 6.635339816587109e-06, "epoch": 1.248, "percentage": 41.6, "elapsed_time": "1:02:59", "remaining_time": "1:28:25"}
+{"current_steps": 157, "total_steps": 375, "loss": 1.4385, "lr": 6.594386416238095e-06, "epoch": 1.256, "percentage": 41.87, "elapsed_time": "1:03:21", "remaining_time": "1:27:58"}
+{"current_steps": 158, "total_steps": 375, "loss": 1.3711, "lr": 6.553313595805666e-06, "epoch": 1.264, "percentage": 42.13, "elapsed_time": "1:03:44", "remaining_time": "1:27:32"}
+{"current_steps": 159, "total_steps": 375, "loss": 1.392, "lr": 6.512124431658006e-06, "epoch": 1.272, "percentage": 42.4, "elapsed_time": "1:04:07", "remaining_time": "1:27:06"}
+{"current_steps": 160, "total_steps": 375, "loss": 1.3345, "lr": 6.470822008877482e-06, "epoch": 1.28, "percentage": 42.67, "elapsed_time": "1:04:31", "remaining_time": "1:26:42"}
+{"current_steps": 161, "total_steps": 375, "loss": 1.4048, "lr": 6.4294094210295725e-06, "epoch": 1.288, "percentage": 42.93, "elapsed_time": "1:04:55", "remaining_time": "1:26:18"}
+{"current_steps": 162, "total_steps": 375, "loss": 1.4097, "lr": 6.3878897699311525e-06, "epoch": 1.296, "percentage": 43.2, "elapsed_time": "1:05:21", "remaining_time": "1:25:55"}
+{"current_steps": 163, "total_steps": 375, "loss": 1.3149, "lr": 6.346266165418173e-06, "epoch": 1.304, "percentage": 43.47, "elapsed_time": "1:05:44", "remaining_time": "1:25:30"}
+{"current_steps": 164, "total_steps": 375, "loss": 1.4207, "lr": 6.304541725112734e-06, "epoch": 1.312, "percentage": 43.73, "elapsed_time": "1:06:07", "remaining_time": "1:25:04"}
+{"current_steps": 165, "total_steps": 375, "loss": 1.4244, "lr": 6.262719574189564e-06, "epoch": 1.32, "percentage": 44.0, "elapsed_time": "1:06:31", "remaining_time": "1:24:40"}
+{"current_steps": 166, "total_steps": 375, "loss": 1.2237, "lr": 6.2208028451419575e-06, "epoch": 1.328, "percentage": 44.27, "elapsed_time": "1:06:54", "remaining_time": "1:24:13"}
+{"current_steps": 167, "total_steps": 375, "loss": 1.3877, "lr": 6.178794677547138e-06, "epoch": 1.336, "percentage": 44.53, "elapsed_time": "1:07:18", "remaining_time": "1:23:50"}
+{"current_steps": 168, "total_steps": 375, "loss": 1.4125, "lr": 6.136698217831106e-06, "epoch": 1.3439999999999999, "percentage": 44.8, "elapsed_time": "1:07:45", "remaining_time": "1:23:29"}
+{"current_steps": 169, "total_steps": 375, "loss": 1.3165, "lr": 6.094516619032975e-06, "epoch": 1.3519999999999999, "percentage": 45.07, "elapsed_time": "1:08:12", "remaining_time": "1:23:08"}
+{"current_steps": 170, "total_steps": 375, "loss": 1.2087, "lr": 6.052253040568804e-06, "epoch": 1.3599999999999999, "percentage": 45.33, "elapsed_time": "1:08:36", "remaining_time": "1:22:43"}
+{"current_steps": 171, "total_steps": 375, "loss": 1.298, "lr": 6.009910647994956e-06, "epoch": 1.3679999999999999, "percentage": 45.6, "elapsed_time": "1:08:59", "remaining_time": "1:22:18"}
+{"current_steps": 172, "total_steps": 375, "loss": 1.478, "lr": 5.967492612770999e-06, "epoch": 1.376, "percentage": 45.87, "elapsed_time": "1:09:23", "remaining_time": "1:21:54"}
+{"current_steps": 173, "total_steps": 375, "loss": 1.3841, "lr": 5.925002112022158e-06, "epoch": 1.384, "percentage": 46.13, "elapsed_time": "1:09:46", "remaining_time": "1:21:28"}
+{"current_steps": 174, "total_steps": 375, "loss": 1.3204, "lr": 5.882442328301356e-06, "epoch": 1.392, "percentage": 46.4, "elapsed_time": "1:10:08", "remaining_time": "1:21:01"}
+{"current_steps": 175, "total_steps": 375, "loss": 1.2993, "lr": 5.839816449350824e-06, "epoch": 1.4, "percentage": 46.67, "elapsed_time": "1:10:34", "remaining_time": "1:20:39"}
+{"current_steps": 176, "total_steps": 375, "loss": 1.3931, "lr": 5.7971276678633625e-06, "epoch": 1.408, "percentage": 46.93, "elapsed_time": "1:10:58", "remaining_time": "1:20:15"}
+{"current_steps": 177, "total_steps": 375, "loss": 1.3798, "lr": 5.754379181243179e-06, "epoch": 1.416, "percentage": 47.2, "elapsed_time": "1:11:21", "remaining_time": "1:19:49"}
+{"current_steps": 178, "total_steps": 375, "loss": 1.3581, "lr": 5.711574191366427e-06, "epoch": 1.424, "percentage": 47.47, "elapsed_time": "1:11:43", "remaining_time": "1:19:23"}
+{"current_steps": 179, "total_steps": 375, "loss": 1.333, "lr": 5.668715904341365e-06, "epoch": 1.432, "percentage": 47.73, "elapsed_time": "1:12:08", "remaining_time": "1:18:59"}
+{"current_steps": 180, "total_steps": 375, "loss": 1.369, "lr": 5.62580753026823e-06, "epoch": 1.44, "percentage": 48.0, "elapsed_time": "1:12:32", "remaining_time": "1:18:34"}
+{"current_steps": 181, "total_steps": 375, "loss": 1.207, "lr": 5.5828522829987965e-06, "epoch": 1.448, "percentage": 48.27, "elapsed_time": "1:12:56", "remaining_time": "1:18:10"}
+{"current_steps": 182, "total_steps": 375, "loss": 1.3628, "lr": 5.539853379895656e-06, "epoch": 1.456, "percentage": 48.53, "elapsed_time": "1:13:20", "remaining_time": "1:17:46"}
+{"current_steps": 183, "total_steps": 375, "loss": 1.3174, "lr": 5.496814041591234e-06, "epoch": 1.464, "percentage": 48.8, "elapsed_time": "1:13:45", "remaining_time": "1:17:22"}
+{"current_steps": 184, "total_steps": 375, "loss": 1.2956, "lr": 5.453737491746572e-06, "epoch": 1.472, "percentage": 49.07, "elapsed_time": "1:14:10", "remaining_time": "1:16:59"}
+{"current_steps": 185, "total_steps": 375, "loss": 1.4769, "lr": 5.410626956809864e-06, "epoch": 1.48, "percentage": 49.33, "elapsed_time": "1:14:33", "remaining_time": "1:16:34"}
+{"current_steps": 186, "total_steps": 375, "loss": 1.2785, "lr": 5.367485665774802e-06, "epoch": 1.488, "percentage": 49.6, "elapsed_time": "1:14:56", "remaining_time": "1:16:09"}
+{"current_steps": 187, "total_steps": 375, "loss": 1.3007, "lr": 5.324316849938715e-06, "epoch": 1.496, "percentage": 49.87, "elapsed_time": "1:15:20", "remaining_time": "1:15:44"}
+{"current_steps": 188, "total_steps": 375, "loss": 1.3652, "lr": 5.281123742660558e-06, "epoch": 1.504, "percentage": 50.13, "elapsed_time": "1:15:43", "remaining_time": "1:15:19"}
+{"current_steps": 189, "total_steps": 375, "loss": 1.4483, "lr": 5.237909579118713e-06, "epoch": 1.512, "percentage": 50.4, "elapsed_time": "1:16:05", "remaining_time": "1:14:53"}
+{"current_steps": 190, "total_steps": 375, "loss": 1.3331, "lr": 5.194677596068689e-06, "epoch": 1.52, "percentage": 50.67, "elapsed_time": "1:16:28", "remaining_time": "1:14:27"}
+{"current_steps": 191, "total_steps": 375, "loss": 1.3594, "lr": 5.1514310316006835e-06, "epoch": 1.528, "percentage": 50.93, "elapsed_time": "1:16:52", "remaining_time": "1:14:03"}
+{"current_steps": 192, "total_steps": 375, "loss": 1.3753, "lr": 5.1081731248970435e-06, "epoch": 1.536, "percentage": 51.2, "elapsed_time": "1:17:16", "remaining_time": "1:13:39"}
+{"current_steps": 193, "total_steps": 375, "loss": 1.3207, "lr": 5.064907115989655e-06, "epoch": 1.544, "percentage": 51.47, "elapsed_time": "1:17:42", "remaining_time": "1:13:16"}
+{"current_steps": 194, "total_steps": 375, "loss": 1.2969, "lr": 5.021636245517261e-06, "epoch": 1.552, "percentage": 51.73, "elapsed_time": "1:18:09", "remaining_time": "1:12:55"}
+{"current_steps": 195, "total_steps": 375, "loss": 1.3814, "lr": 4.978363754482741e-06, "epoch": 1.56, "percentage": 52.0, "elapsed_time": "1:18:33", "remaining_time": "1:12:31"}
+{"current_steps": 196, "total_steps": 375, "loss": 1.3583, "lr": 4.935092884010347e-06, "epoch": 1.568, "percentage": 52.27, "elapsed_time": "1:18:57", "remaining_time": "1:12:06"}
+{"current_steps": 197, "total_steps": 375, "loss": 1.4385, "lr": 4.891826875102958e-06, "epoch": 1.576, "percentage": 52.53, "elapsed_time": "1:19:22", "remaining_time": "1:11:42"}
+{"current_steps": 198, "total_steps": 375, "loss": 1.4347, "lr": 4.848568968399317e-06, "epoch": 1.584, "percentage": 52.8, "elapsed_time": "1:19:44", "remaining_time": "1:11:17"}
+{"current_steps": 199, "total_steps": 375, "loss": 1.3173, "lr": 4.805322403931312e-06, "epoch": 1.592, "percentage": 53.07, "elapsed_time": "1:20:07", "remaining_time": "1:10:52"}
+{"current_steps": 200, "total_steps": 375, "loss": 1.4581, "lr": 4.762090420881289e-06, "epoch": 1.6, "percentage": 53.33, "elapsed_time": "1:20:29", "remaining_time": "1:10:26"}
+{"current_steps": 201, "total_steps": 375, "loss": 1.3553, "lr": 4.718876257339444e-06, "epoch": 1.608, "percentage": 53.6, "elapsed_time": "1:20:54", "remaining_time": "1:10:02"}
+{"current_steps": 202, "total_steps": 375, "loss": 1.2876, "lr": 4.6756831500612846e-06, "epoch": 1.616, "percentage": 53.87, "elapsed_time": "1:21:19", "remaining_time": "1:09:38"}
+{"current_steps": 203, "total_steps": 375, "loss": 1.4021, "lr": 4.632514334225201e-06, "epoch": 1.624, "percentage": 54.13, "elapsed_time": "1:21:42", "remaining_time": "1:09:13"}
+{"current_steps": 204, "total_steps": 375, "loss": 1.492, "lr": 4.589373043190137e-06, "epoch": 1.6320000000000001, "percentage": 54.4, "elapsed_time": "1:22:06", "remaining_time": "1:08:49"}
+{"current_steps": 205, "total_steps": 375, "loss": 1.3503, "lr": 4.546262508253429e-06, "epoch": 1.6400000000000001, "percentage": 54.67, "elapsed_time": "1:22:30", "remaining_time": "1:08:24"}
+{"current_steps": 206, "total_steps": 375, "loss": 1.3706, "lr": 4.503185958408767e-06, "epoch": 1.6480000000000001, "percentage": 54.93, "elapsed_time": "1:22:51", "remaining_time": "1:07:58"}
+{"current_steps": 207, "total_steps": 375, "loss": 1.2973, "lr": 4.460146620104347e-06, "epoch": 1.6560000000000001, "percentage": 55.2, "elapsed_time": "1:23:13", "remaining_time": "1:07:32"}
+{"current_steps": 208, "total_steps": 375, "loss": 1.3999, "lr": 4.417147717001205e-06, "epoch": 1.6640000000000001, "percentage": 55.47, "elapsed_time": "1:23:40", "remaining_time": "1:07:10"}
+{"current_steps": 209, "total_steps": 375, "loss": 1.3405, "lr": 4.374192469731771e-06, "epoch": 1.6720000000000002, "percentage": 55.73, "elapsed_time": "1:24:05", "remaining_time": "1:06:47"}
+{"current_steps": 210, "total_steps": 375, "loss": 1.2838, "lr": 4.331284095658637e-06, "epoch": 1.6800000000000002, "percentage": 56.0, "elapsed_time": "1:24:28", "remaining_time": "1:06:22"}
+{"current_steps": 211, "total_steps": 375, "loss": 1.3207, "lr": 4.2884258086335755e-06, "epoch": 1.688, "percentage": 56.27, "elapsed_time": "1:24:51", "remaining_time": "1:05:56"}
+{"current_steps": 212, "total_steps": 375, "loss": 1.3182, "lr": 4.245620818756822e-06, "epoch": 1.696, "percentage": 56.53, "elapsed_time": "1:25:14", "remaining_time": "1:05:32"}
+{"current_steps": 213, "total_steps": 375, "loss": 1.299, "lr": 4.202872332136639e-06, "epoch": 1.704, "percentage": 56.8, "elapsed_time": "1:25:40", "remaining_time": "1:05:09"}
+{"current_steps": 214, "total_steps": 375, "loss": 1.3005, "lr": 4.160183550649176e-06, "epoch": 1.712, "percentage": 57.07, "elapsed_time": "1:26:03", "remaining_time": "1:04:45"}
+{"current_steps": 215, "total_steps": 375, "loss": 1.2494, "lr": 4.117557671698648e-06, "epoch": 1.72, "percentage": 57.33, "elapsed_time": "1:26:26", "remaining_time": "1:04:20"}
+{"current_steps": 216, "total_steps": 375, "loss": 1.4299, "lr": 4.074997887977843e-06, "epoch": 1.728, "percentage": 57.6, "elapsed_time": "1:26:50", "remaining_time": "1:03:55"}
+{"current_steps": 217, "total_steps": 375, "loss": 1.379, "lr": 4.032507387229002e-06, "epoch": 1.736, "percentage": 57.87, "elapsed_time": "1:27:12", "remaining_time": "1:03:30"}
+{"current_steps": 218, "total_steps": 375, "loss": 1.3061, "lr": 3.9900893520050446e-06, "epoch": 1.744, "percentage": 58.13, "elapsed_time": "1:27:36", "remaining_time": "1:03:05"}
+{"current_steps": 219, "total_steps": 375, "loss": 1.3921, "lr": 3.9477469594311975e-06, "epoch": 1.752, "percentage": 58.4, "elapsed_time": "1:27:58", "remaining_time": "1:02:39"}
+{"current_steps": 220, "total_steps": 375, "loss": 1.3521, "lr": 3.905483380967027e-06, "epoch": 1.76, "percentage": 58.67, "elapsed_time": "1:28:22", "remaining_time": "1:02:15"}
+{"current_steps": 221, "total_steps": 375, "loss": 1.3919, "lr": 3.863301782168896e-06, "epoch": 1.768, "percentage": 58.93, "elapsed_time": "1:28:46", "remaining_time": "1:01:51"}
+{"current_steps": 222, "total_steps": 375, "loss": 1.4446, "lr": 3.821205322452863e-06, "epoch": 1.776, "percentage": 59.2, "elapsed_time": "1:29:11", "remaining_time": "1:01:28"}
+{"current_steps": 223, "total_steps": 375, "loss": 1.4121, "lr": 3.779197154858044e-06, "epoch": 1.784, "percentage": 59.47, "elapsed_time": "1:29:34", "remaining_time": "1:01:03"}
+{"current_steps": 224, "total_steps": 375, "loss": 1.3486, "lr": 3.7372804258104367e-06, "epoch": 1.792, "percentage": 59.73, "elapsed_time": "1:29:58", "remaining_time": "1:00:39"}
+{"current_steps": 225, "total_steps": 375, "loss": 1.3536, "lr": 3.695458274887268e-06, "epoch": 1.8, "percentage": 60.0, "elapsed_time": "1:30:22", "remaining_time": "1:00:14"}
+{"current_steps": 226, "total_steps": 375, "loss": 1.3906, "lr": 3.6537338345818273e-06, "epoch": 1.808, "percentage": 60.27, "elapsed_time": "1:30:45", "remaining_time": "0:59:50"}
+{"current_steps": 227, "total_steps": 375, "loss": 1.4416, "lr": 3.6121102300688504e-06, "epoch": 1.8159999999999998, "percentage": 60.53, "elapsed_time": "1:31:09", "remaining_time": "0:59:26"}
+{"current_steps": 228, "total_steps": 375, "loss": 1.4409, "lr": 3.5705905789704296e-06, "epoch": 1.8239999999999998, "percentage": 60.8, "elapsed_time": "1:31:31", "remaining_time": "0:59:00"}
+{"current_steps": 229, "total_steps": 375, "loss": 1.4683, "lr": 3.529177991122519e-06, "epoch": 1.8319999999999999, "percentage": 61.07, "elapsed_time": "1:31:55", "remaining_time": "0:58:36"}
+{"current_steps": 230, "total_steps": 375, "loss": 1.3358, "lr": 3.487875568341995e-06, "epoch": 1.8399999999999999, "percentage": 61.33, "elapsed_time": "1:32:19", "remaining_time": "0:58:12"}
+{"current_steps": 231, "total_steps": 375, "loss": 1.441, "lr": 3.446686404194337e-06, "epoch": 1.8479999999999999, "percentage": 61.6, "elapsed_time": "1:32:46", "remaining_time": "0:57:50"}
+{"current_steps": 232, "total_steps": 375, "loss": 1.3806, "lr": 3.4056135837619077e-06, "epoch": 1.8559999999999999, "percentage": 61.87, "elapsed_time": "1:33:12", "remaining_time": "0:57:27"}
+{"current_steps": 233, "total_steps": 375, "loss": 1.2891, "lr": 3.3646601834128924e-06, "epoch": 1.8639999999999999, "percentage": 62.13, "elapsed_time": "1:33:36", "remaining_time": "0:57:03"}
+{"current_steps": 234, "total_steps": 375, "loss": 1.3966, "lr": 3.3238292705708675e-06, "epoch": 1.8719999999999999, "percentage": 62.4, "elapsed_time": "1:34:01", "remaining_time": "0:56:39"}
+{"current_steps": 235, "total_steps": 375, "loss": 1.3219, "lr": 3.2831239034850593e-06, "epoch": 1.88, "percentage": 62.67, "elapsed_time": "1:34:26", "remaining_time": "0:56:15"}
+{"current_steps": 236, "total_steps": 375, "loss": 1.5404, "lr": 3.2425471310012645e-06, "epoch": 1.888, "percentage": 62.93, "elapsed_time": "1:34:50", "remaining_time": "0:55:51"}
+{"current_steps": 237, "total_steps": 375, "loss": 1.2135, "lr": 3.2021019923335093e-06, "epoch": 1.896, "percentage": 63.2, "elapsed_time": "1:35:14", "remaining_time": "0:55:27"}
+{"current_steps": 238, "total_steps": 375, "loss": 1.2094, "lr": 3.1617915168363994e-06, "epoch": 1.904, "percentage": 63.47, "elapsed_time": "1:35:39", "remaining_time": "0:55:03"}
+{"current_steps": 239, "total_steps": 375, "loss": 1.3233, "lr": 3.121618723778225e-06, "epoch": 1.912, "percentage": 63.73, "elapsed_time": "1:36:01", "remaining_time": "0:54:38"}
+{"current_steps": 240, "total_steps": 375, "loss": 1.3909, "lr": 3.081586622114809e-06, "epoch": 1.92, "percentage": 64.0, "elapsed_time": "1:36:26", "remaining_time": "0:54:14"}
+{"current_steps": 241, "total_steps": 375, "loss": 1.3041, "lr": 3.041698210264149e-06, "epoch": 1.928, "percentage": 64.27, "elapsed_time": "1:36:47", "remaining_time": "0:53:49"}
+{"current_steps": 242, "total_steps": 375, "loss": 1.256, "lr": 3.001956475881822e-06, "epoch": 1.936, "percentage": 64.53, "elapsed_time": "1:37:12", "remaining_time": "0:53:25"}
+{"current_steps": 243, "total_steps": 375, "loss": 1.4119, "lr": 2.962364395637216e-06, "epoch": 1.944, "percentage": 64.8, "elapsed_time": "1:37:38", "remaining_time": "0:53:02"}
+{"current_steps": 244, "total_steps": 375, "loss": 1.238, "lr": 2.9229249349905686e-06, "epoch": 1.952, "percentage": 65.07, "elapsed_time": "1:38:01", "remaining_time": "0:52:37"}
+{"current_steps": 245, "total_steps": 375, "loss": 1.3517, "lr": 2.8836410479708625e-06, "epoch": 1.96, "percentage": 65.33, "elapsed_time": "1:38:23", "remaining_time": "0:52:12"}
+{"current_steps": 246, "total_steps": 375, "loss": 1.2968, "lr": 2.84451567695456e-06, "epoch": 1.968, "percentage": 65.6, "elapsed_time": "1:38:49", "remaining_time": "0:51:49"}
+{"current_steps": 247, "total_steps": 375, "loss": 1.3458, "lr": 2.805551752445222e-06, "epoch": 1.976, "percentage": 65.87, "elapsed_time": "1:39:13", "remaining_time": "0:51:25"}
+{"current_steps": 248, "total_steps": 375, "loss": 1.424, "lr": 2.766752192854012e-06, "epoch": 1.984, "percentage": 66.13, "elapsed_time": "1:39:37", "remaining_time": "0:51:01"}
+{"current_steps": 249, "total_steps": 375, "loss": 1.2664, "lr": 2.728119904281105e-06, "epoch": 1.992, "percentage": 66.4, "elapsed_time": "1:40:02", "remaining_time": "0:50:37"}
+{"current_steps": 250, "total_steps": 375, "loss": 1.3434, "lr": 2.689657780298019e-06, "epoch": 2.0, "percentage": 66.67, "elapsed_time": "1:40:30", "remaining_time": "0:50:15"}
+{"current_steps": 251, "total_steps": 375, "loss": 1.2695, "lr": 2.651368701730889e-06, "epoch": 2.008, "percentage": 66.93, "elapsed_time": "1:41:05", "remaining_time": "0:49:56"}
+{"current_steps": 252, "total_steps": 375, "loss": 1.1521, "lr": 2.6132555364446856e-06, "epoch": 2.016, "percentage": 67.2, "elapsed_time": "1:41:30", "remaining_time": "0:49:32"}
+{"current_steps": 253, "total_steps": 375, "loss": 1.2569, "lr": 2.5753211391284172e-06, "epoch": 2.024, "percentage": 67.47, "elapsed_time": "1:41:53", "remaining_time": "0:49:07"}
+{"current_steps": 254, "total_steps": 375, "loss": 1.1739, "lr": 2.537568351081311e-06, "epoch": 2.032, "percentage": 67.73, "elapsed_time": "1:42:17", "remaining_time": "0:48:43"}
+{"current_steps": 255, "total_steps": 375, "loss": 1.1113, "lr": 2.5000000000000015e-06, "epoch": 2.04, "percentage": 68.0, "elapsed_time": "1:42:39", "remaining_time": "0:48:18"}
+{"current_steps": 256, "total_steps": 375, "loss": 1.2154, "lr": 2.4626188997667224e-06, "epoch": 2.048, "percentage": 68.27, "elapsed_time": "1:43:04", "remaining_time": "0:47:54"}
+{"current_steps": 257, "total_steps": 375, "loss": 1.1986, "lr": 2.425427850238565e-06, "epoch": 2.056, "percentage": 68.53, "elapsed_time": "1:43:27", "remaining_time": "0:47:30"}
+{"current_steps": 258, "total_steps": 375, "loss": 1.1513, "lr": 2.388429637037753e-06, "epoch": 2.064, "percentage": 68.8, "elapsed_time": "1:43:53", "remaining_time": "0:47:06"}
+{"current_steps": 259, "total_steps": 375, "loss": 1.1515, "lr": 2.3516270313430085e-06, "epoch": 2.072, "percentage": 69.07, "elapsed_time": "1:44:20", "remaining_time": "0:46:43"}
+{"current_steps": 260, "total_steps": 375, "loss": 1.1398, "lr": 2.3150227896819782e-06, "epoch": 2.08, "percentage": 69.33, "elapsed_time": "1:44:43", "remaining_time": "0:46:19"}
+{"current_steps": 261, "total_steps": 375, "loss": 1.2116, "lr": 2.278619653724781e-06, "epoch": 2.088, "percentage": 69.6, "elapsed_time": "1:45:08", "remaining_time": "0:45:55"}
+{"current_steps": 262, "total_steps": 375, "loss": 1.1741, "lr": 2.2424203500786473e-06, "epoch": 2.096, "percentage": 69.87, "elapsed_time": "1:45:32", "remaining_time": "0:45:31"}
+{"current_steps": 263, "total_steps": 375, "loss": 1.2057, "lr": 2.206427590083703e-06, "epoch": 2.104, "percentage": 70.13, "elapsed_time": "1:45:55", "remaining_time": "0:45:06"}
+{"current_steps": 264, "total_steps": 375, "loss": 1.1549, "lr": 2.170644069609876e-06, "epoch": 2.112, "percentage": 70.4, "elapsed_time": "1:46:18", "remaining_time": "0:44:42"}
+{"current_steps": 265, "total_steps": 375, "loss": 1.2437, "lr": 2.1350724688549906e-06, "epoch": 2.12, "percentage": 70.67, "elapsed_time": "1:46:43", "remaining_time": "0:44:18"}
+{"current_steps": 266, "total_steps": 375, "loss": 1.1539, "lr": 2.09971545214401e-06, "epoch": 2.128, "percentage": 70.93, "elapsed_time": "1:47:09", "remaining_time": "0:43:54"}
+{"current_steps": 267, "total_steps": 375, "loss": 1.186, "lr": 2.0645756677294788e-06, "epoch": 2.136, "percentage": 71.2, "elapsed_time": "1:47:32", "remaining_time": "0:43:30"}
+{"current_steps": 268, "total_steps": 375, "loss": 1.2699, "lr": 2.029655747593169e-06, "epoch": 2.144, "percentage": 71.47, "elapsed_time": "1:47:55", "remaining_time": "0:43:05"}
+{"current_steps": 269, "total_steps": 375, "loss": 1.1699, "lr": 1.9949583072489455e-06, "epoch": 2.152, "percentage": 71.73, "elapsed_time": "1:48:20", "remaining_time": "0:42:41"}
+{"current_steps": 270, "total_steps": 375, "loss": 1.2022, "lr": 1.9604859455468587e-06, "epoch": 2.16, "percentage": 72.0, "elapsed_time": "1:48:43", "remaining_time": "0:42:17"}
+{"current_steps": 271, "total_steps": 375, "loss": 1.1245, "lr": 1.926241244478496e-06, "epoch": 2.168, "percentage": 72.27, "elapsed_time": "1:49:08", "remaining_time": "0:41:52"}
+{"current_steps": 272, "total_steps": 375, "loss": 1.1375, "lr": 1.8922267689835806e-06, "epoch": 2.176, "percentage": 72.53, "elapsed_time": "1:49:32", "remaining_time": "0:41:28"}
+{"current_steps": 273, "total_steps": 375, "loss": 1.178, "lr": 1.8584450667578656e-06, "epoch": 2.184, "percentage": 72.8, "elapsed_time": "1:49:59", "remaining_time": "0:41:05"}
+{"current_steps": 274, "total_steps": 375, "loss": 1.1776, "lr": 1.8248986680623077e-06, "epoch": 2.192, "percentage": 73.07, "elapsed_time": "1:50:24", "remaining_time": "0:40:41"}
+{"current_steps": 275, "total_steps": 375, "loss": 1.0931, "lr": 1.7915900855335506e-06, "epoch": 2.2, "percentage": 73.33, "elapsed_time": "1:50:48", "remaining_time": "0:40:17"}
+{"current_steps": 276, "total_steps": 375, "loss": 1.1597, "lr": 1.7585218139957205e-06, "epoch": 2.208, "percentage": 73.6, "elapsed_time": "1:51:11", "remaining_time": "0:39:53"}
+{"current_steps": 277, "total_steps": 375, "loss": 1.2446, "lr": 1.7256963302735752e-06, "epoch": 2.216, "percentage": 73.87, "elapsed_time": "1:51:36", "remaining_time": "0:39:29"}
+{"current_steps": 278, "total_steps": 375, "loss": 1.202, "lr": 1.6931160930069789e-06, "epoch": 2.224, "percentage": 74.13, "elapsed_time": "1:51:59", "remaining_time": "0:39:04"}
+{"current_steps": 279, "total_steps": 375, "loss": 1.1506, "lr": 1.6607835424667578e-06, "epoch": 2.232, "percentage": 74.4, "elapsed_time": "1:52:25", "remaining_time": "0:38:40"}
+{"current_steps": 280, "total_steps": 375, "loss": 1.093, "lr": 1.6287011003719105e-06, "epoch": 2.24, "percentage": 74.67, "elapsed_time": "1:52:48", "remaining_time": "0:38:16"}
+{"current_steps": 281, "total_steps": 375, "loss": 1.2059, "lr": 1.596871169708235e-06, "epoch": 2.248, "percentage": 74.93, "elapsed_time": "1:53:11", "remaining_time": "0:37:51"}
+{"current_steps": 282, "total_steps": 375, "loss": 1.1799, "lr": 1.5652961345483353e-06, "epoch": 2.2560000000000002, "percentage": 75.2, "elapsed_time": "1:53:35", "remaining_time": "0:37:27"}
+{"current_steps": 283, "total_steps": 375, "loss": 1.1929, "lr": 1.5339783598730568e-06, "epoch": 2.2640000000000002, "percentage": 75.47, "elapsed_time": "1:53:58", "remaining_time": "0:37:03"}
+{"current_steps": 284, "total_steps": 375, "loss": 1.2047, "lr": 1.5029201913943425e-06, "epoch": 2.2720000000000002, "percentage": 75.73, "elapsed_time": "1:54:22", "remaining_time": "0:36:38"}
+{"current_steps": 285, "total_steps": 375, "loss": 1.2801, "lr": 1.4721239553795485e-06, "epoch": 2.2800000000000002, "percentage": 76.0, "elapsed_time": "1:54:44", "remaining_time": "0:36:14"}
+{"current_steps": 286, "total_steps": 375, "loss": 1.1508, "lr": 1.4415919584771999e-06, "epoch": 2.288, "percentage": 76.27, "elapsed_time": "1:55:07", "remaining_time": "0:35:49"}
+{"current_steps": 287, "total_steps": 375, "loss": 1.171, "lr": 1.4113264875442201e-06, "epoch": 2.296, "percentage": 76.53, "elapsed_time": "1:55:29", "remaining_time": "0:35:24"}
+{"current_steps": 288, "total_steps": 375, "loss": 1.0647, "lr": 1.3813298094746491e-06, "epoch": 2.304, "percentage": 76.8, "elapsed_time": "1:55:53", "remaining_time": "0:35:00"}
+{"current_steps": 289, "total_steps": 375, "loss": 1.21, "lr": 1.35160417102985e-06, "epoch": 2.312, "percentage": 77.07, "elapsed_time": "1:56:16", "remaining_time": "0:34:35"}
+{"current_steps": 290, "total_steps": 375, "loss": 1.3148, "lr": 1.3221517986702249e-06, "epoch": 2.32, "percentage": 77.33, "elapsed_time": "1:56:38", "remaining_time": "0:34:11"}
+{"current_steps": 291, "total_steps": 375, "loss": 1.133, "lr": 1.292974898388456e-06, "epoch": 2.328, "percentage": 77.6, "elapsed_time": "1:57:01", "remaining_time": "0:33:46"}
+{"current_steps": 292, "total_steps": 375, "loss": 1.1668, "lr": 1.2640756555442684e-06, "epoch": 2.336, "percentage": 77.87, "elapsed_time": "1:57:24", "remaining_time": "0:33:22"}
+{"current_steps": 293, "total_steps": 375, "loss": 1.1914, "lr": 1.235456234700756e-06, "epoch": 2.344, "percentage": 78.13, "elapsed_time": "1:57:50", "remaining_time": "0:32:58"}
+{"current_steps": 294, "total_steps": 375, "loss": 1.1777, "lr": 1.207118779462248e-06, "epoch": 2.352, "percentage": 78.4, "elapsed_time": "1:58:13", "remaining_time": "0:32:34"}
+{"current_steps": 295, "total_steps": 375, "loss": 1.1946, "lr": 1.1790654123137552e-06, "epoch": 2.36, "percentage": 78.67, "elapsed_time": "1:58:36", "remaining_time": "0:32:09"}
+{"current_steps": 296, "total_steps": 375, "loss": 1.2927, "lr": 1.1512982344619904e-06, "epoch": 2.368, "percentage": 78.93, "elapsed_time": "1:59:01", "remaining_time": "0:31:46"}
+{"current_steps": 297, "total_steps": 375, "loss": 1.3264, "lr": 1.1238193256779955e-06, "epoch": 2.376, "percentage": 79.2, "elapsed_time": "1:59:24", "remaining_time": "0:31:21"}
+{"current_steps": 298, "total_steps": 375, "loss": 1.1768, "lr": 1.0966307441413598e-06, "epoch": 2.384, "percentage": 79.47, "elapsed_time": "1:59:47", "remaining_time": "0:30:57"}
+{"current_steps": 299, "total_steps": 375, "loss": 1.1419, "lr": 1.0697345262860638e-06, "epoch": 2.392, "percentage": 79.73, "elapsed_time": "2:00:10", "remaining_time": "0:30:32"}
+{"current_steps": 300, "total_steps": 375, "loss": 1.1746, "lr": 1.0431326866479457e-06, "epoch": 2.4, "percentage": 80.0, "elapsed_time": "2:00:34", "remaining_time": "0:30:08"}
+{"current_steps": 301, "total_steps": 375, "loss": 1.2397, "lr": 1.01682721771382e-06, "epoch": 2.408, "percentage": 80.27, "elapsed_time": "2:00:56", "remaining_time": "0:29:44"}
+{"current_steps": 302, "total_steps": 375, "loss": 1.1136, "lr": 9.908200897722332e-07, "epoch": 2.416, "percentage": 80.53, "elapsed_time": "2:01:20", "remaining_time": "0:29:19"}
+{"current_steps": 303, "total_steps": 375, "loss": 1.2358, "lr": 9.6511325076589e-07, "epoch": 2.424, "percentage": 80.8, "elapsed_time": "2:01:43", "remaining_time": "0:28:55"}
+{"current_steps": 304, "total_steps": 375, "loss": 1.1577, "lr": 9.397086261457511e-07, "epoch": 2.432, "percentage": 81.07, "elapsed_time": "2:02:07", "remaining_time": "0:28:31"}
+{"current_steps": 305, "total_steps": 375, "loss": 1.1313, "lr": 9.146081187268185e-07, "epoch": 2.44, "percentage": 81.33, "elapsed_time": "2:02:31", "remaining_time": "0:28:07"}
+{"current_steps": 306, "total_steps": 375, "loss": 1.2469, "lr": 8.898136085456127e-07, "epoch": 2.448, "percentage": 81.6, "elapsed_time": "2:02:55", "remaining_time": "0:27:43"}
+{"current_steps": 307, "total_steps": 375, "loss": 1.1897, "lr": 8.65326952719357e-07, "epoch": 2.456, "percentage": 81.87, "elapsed_time": "2:03:20", "remaining_time": "0:27:19"}
+{"current_steps": 308, "total_steps": 375, "loss": 1.2205, "lr": 8.411499853068783e-07, "epoch": 2.464, "percentage": 82.13, "elapsed_time": "2:03:45", "remaining_time": "0:26:55"}
+{"current_steps": 309, "total_steps": 375, "loss": 1.1335, "lr": 8.172845171712379e-07, "epoch": 2.472, "percentage": 82.4, "elapsed_time": "2:04:12", "remaining_time": "0:26:31"}
+{"current_steps": 310, "total_steps": 375, "loss": 1.1261, "lr": 7.937323358440935e-07, "epoch": 2.48, "percentage": 82.67, "elapsed_time": "2:04:36", "remaining_time": "0:26:07"}
+{"current_steps": 311, "total_steps": 375, "loss": 1.1149, "lr": 7.70495205391818e-07, "epoch": 2.488, "percentage": 82.93, "elapsed_time": "2:05:00", "remaining_time": "0:25:43"}
+{"current_steps": 312, "total_steps": 375, "loss": 1.2633, "lr": 7.475748662833615e-07, "epoch": 2.496, "percentage": 83.2, "elapsed_time": "2:05:24", "remaining_time": "0:25:19"}
+{"current_steps": 313, "total_steps": 375, "loss": 1.2781, "lr": 7.249730352599e-07, "epoch": 2.504, "percentage": 83.47, "elapsed_time": "2:05:48", "remaining_time": "0:24:55"}
+{"current_steps": 314, "total_steps": 375, "loss": 1.1741, "lr": 7.026914052062433e-07, "epoch": 2.512, "percentage": 83.73, "elapsed_time": "2:06:14", "remaining_time": "0:24:31"}
+{"current_steps": 315, "total_steps": 375, "loss": 1.183, "lr": 6.807316450240425e-07, "epoch": 2.52, "percentage": 84.0, "elapsed_time": "2:06:38", "remaining_time": "0:24:07"}
+{"current_steps": 316, "total_steps": 375, "loss": 1.2157, "lr": 6.590953995067812e-07, "epoch": 2.528, "percentage": 84.27, "elapsed_time": "2:07:02", "remaining_time": "0:23:43"}
+{"current_steps": 317, "total_steps": 375, "loss": 1.1956, "lr": 6.377842892165892e-07, "epoch": 2.536, "percentage": 84.53, "elapsed_time": "2:07:24", "remaining_time": "0:23:18"}
+{"current_steps": 318, "total_steps": 375, "loss": 1.1124, "lr": 6.167999103628569e-07, "epoch": 2.544, "percentage": 84.8, "elapsed_time": "2:07:51", "remaining_time": "0:22:55"}
+{"current_steps": 319, "total_steps": 375, "loss": 1.2367, "lr": 5.961438346826792e-07, "epoch": 2.552, "percentage": 85.07, "elapsed_time": "2:08:18", "remaining_time": "0:22:31"}
+{"current_steps": 320, "total_steps": 375, "loss": 1.1769, "lr": 5.758176093231294e-07, "epoch": 2.56, "percentage": 85.33, "elapsed_time": "2:08:41", "remaining_time": "0:22:07"}
+{"current_steps": 321, "total_steps": 375, "loss": 1.1362, "lr": 5.558227567253832e-07, "epoch": 2.568, "percentage": 85.6, "elapsed_time": "2:09:06", "remaining_time": "0:21:43"}
+{"current_steps": 322, "total_steps": 375, "loss": 1.0455, "lr": 5.361607745106817e-07, "epoch": 2.576, "percentage": 85.87, "elapsed_time": "2:09:31", "remaining_time": "0:21:19"}
+{"current_steps": 323, "total_steps": 375, "loss": 1.2309, "lr": 5.168331353681643e-07, "epoch": 2.584, "percentage": 86.13, "elapsed_time": "2:09:55", "remaining_time": "0:20:54"}
+{"current_steps": 324, "total_steps": 375, "loss": 1.1187, "lr": 4.97841286944557e-07, "epoch": 2.592, "percentage": 86.4, "elapsed_time": "2:10:19", "remaining_time": "0:20:30"}
+{"current_steps": 325, "total_steps": 375, "loss": 1.2398, "lr": 4.791866517357491e-07, "epoch": 2.6, "percentage": 86.67, "elapsed_time": "2:10:42", "remaining_time": "0:20:06"}
+{"current_steps": 326, "total_steps": 375, "loss": 1.1547, "lr": 4.608706269802471e-07, "epoch": 2.608, "percentage": 86.93, "elapsed_time": "2:11:06", "remaining_time": "0:19:42"}
+{"current_steps": 327, "total_steps": 375, "loss": 1.2571, "lr": 4.428945845545168e-07, "epoch": 2.616, "percentage": 87.2, "elapsed_time": "2:11:29", "remaining_time": "0:19:18"}
+{"current_steps": 328, "total_steps": 375, "loss": 1.2648, "lr": 4.2525987087023433e-07, "epoch": 2.624, "percentage": 87.47, "elapsed_time": "2:11:53", "remaining_time": "0:18:53"}
+{"current_steps": 329, "total_steps": 375, "loss": 1.211, "lr": 4.0796780677343606e-07, "epoch": 2.632, "percentage": 87.73, "elapsed_time": "2:12:20", "remaining_time": "0:18:30"}
+{"current_steps": 330, "total_steps": 375, "loss": 1.1852, "lr": 3.910196874455896e-07, "epoch": 2.64, "percentage": 88.0, "elapsed_time": "2:12:43", "remaining_time": "0:18:05"}
+{"current_steps": 331, "total_steps": 375, "loss": 1.2286, "lr": 3.744167823065814e-07, "epoch": 2.648, "percentage": 88.27, "elapsed_time": "2:13:06", "remaining_time": "0:17:41"}
+{"current_steps": 332, "total_steps": 375, "loss": 1.1621, "lr": 3.581603349196372e-07, "epoch": 2.656, "percentage": 88.53, "elapsed_time": "2:13:29", "remaining_time": "0:17:17"}
+{"current_steps": 333, "total_steps": 375, "loss": 1.2425, "lr": 3.4225156289818096e-07, "epoch": 2.664, "percentage": 88.8, "elapsed_time": "2:13:53", "remaining_time": "0:16:53"}
+{"current_steps": 334, "total_steps": 375, "loss": 1.1193, "lr": 3.26691657814634e-07, "epoch": 2.672, "percentage": 89.07, "elapsed_time": "2:14:19", "remaining_time": "0:16:29"}
+{"current_steps": 335, "total_steps": 375, "loss": 1.1053, "lr": 3.1148178511116624e-07, "epoch": 2.68, "percentage": 89.33, "elapsed_time": "2:14:44", "remaining_time": "0:16:05"}
+{"current_steps": 336, "total_steps": 375, "loss": 1.1874, "lr": 2.966230840124007e-07, "epoch": 2.6879999999999997, "percentage": 89.6, "elapsed_time": "2:15:06", "remaining_time": "0:15:40"}
+{"current_steps": 337, "total_steps": 375, "loss": 1.1105, "lr": 2.821166674400905e-07, "epoch": 2.6959999999999997, "percentage": 89.87, "elapsed_time": "2:15:31", "remaining_time": "0:15:16"}
+{"current_steps": 338, "total_steps": 375, "loss": 1.1718, "lr": 2.6796362192975766e-07, "epoch": 2.7039999999999997, "percentage": 90.13, "elapsed_time": "2:15:55", "remaining_time": "0:14:52"}
+{"current_steps": 339, "total_steps": 375, "loss": 1.0511, "lr": 2.5416500754931294e-07, "epoch": 2.7119999999999997, "percentage": 90.4, "elapsed_time": "2:16:19", "remaining_time": "0:14:28"}
+{"current_steps": 340, "total_steps": 375, "loss": 1.2698, "lr": 2.407218578196524e-07, "epoch": 2.7199999999999998, "percentage": 90.67, "elapsed_time": "2:16:43", "remaining_time": "0:14:04"}
+{"current_steps": 341, "total_steps": 375, "loss": 1.2182, "lr": 2.2763517963725169e-07, "epoch": 2.7279999999999998, "percentage": 90.93, "elapsed_time": "2:17:06", "remaining_time": "0:13:40"}
+{"current_steps": 342, "total_steps": 375, "loss": 1.2383, "lr": 2.1490595319874574e-07, "epoch": 2.7359999999999998, "percentage": 91.2, "elapsed_time": "2:17:34", "remaining_time": "0:13:16"}
+{"current_steps": 343, "total_steps": 375, "loss": 1.1606, "lr": 2.0253513192751374e-07, "epoch": 2.7439999999999998, "percentage": 91.47, "elapsed_time": "2:17:59", "remaining_time": "0:12:52"}
+{"current_steps": 344, "total_steps": 375, "loss": 1.1782, "lr": 1.905236424022633e-07, "epoch": 2.752, "percentage": 91.73, "elapsed_time": "2:18:23", "remaining_time": "0:12:28"}
+{"current_steps": 345, "total_steps": 375, "loss": 1.1847, "lr": 1.7887238428763553e-07, "epoch": 2.76, "percentage": 92.0, "elapsed_time": "2:18:48", "remaining_time": "0:12:04"}
+{"current_steps": 346, "total_steps": 375, "loss": 1.2057, "lr": 1.6758223026681507e-07, "epoch": 2.768, "percentage": 92.27, "elapsed_time": "2:19:13", "remaining_time": "0:11:40"}
+{"current_steps": 347, "total_steps": 375, "loss": 1.0974, "lr": 1.5665402597616842e-07, "epoch": 2.776, "percentage": 92.53, "elapsed_time": "2:19:38", "remaining_time": "0:11:16"}
+{"current_steps": 348, "total_steps": 375, "loss": 1.169, "lr": 1.4608858994190344e-07, "epoch": 2.784, "percentage": 92.8, "elapsed_time": "2:20:02", "remaining_time": "0:10:51"}
+{"current_steps": 349, "total_steps": 375, "loss": 1.1029, "lr": 1.358867135187636e-07, "epoch": 2.792, "percentage": 93.07, "elapsed_time": "2:20:28", "remaining_time": "0:10:27"}
+{"current_steps": 350, "total_steps": 375, "loss": 1.1654, "lr": 1.2604916083075236e-07, "epoch": 2.8, "percentage": 93.33, "elapsed_time": "2:20:50", "remaining_time": "0:10:03"}
+{"current_steps": 351, "total_steps": 375, "loss": 1.0817, "lr": 1.1657666871390471e-07, "epoch": 2.808, "percentage": 93.6, "elapsed_time": "2:21:15", "remaining_time": "0:09:39"}
+{"current_steps": 352, "total_steps": 375, "loss": 1.3023, "lr": 1.0746994666109234e-07, "epoch": 2.816, "percentage": 93.87, "elapsed_time": "2:21:39", "remaining_time": "0:09:15"}
+{"current_steps": 353, "total_steps": 375, "loss": 1.1863, "lr": 9.872967676888611e-08, "epoch": 2.824, "percentage": 94.13, "elapsed_time": "2:22:02", "remaining_time": "0:08:51"}
+{"current_steps": 354, "total_steps": 375, "loss": 1.1599, "lr": 9.035651368646647e-08, "epoch": 2.832, "percentage": 94.4, "elapsed_time": "2:22:24", "remaining_time": "0:08:26"}
+{"current_steps": 355, "total_steps": 375, "loss": 1.2885, "lr": 8.235108456658814e-08, "epoch": 2.84, "percentage": 94.67, "elapsed_time": "2:22:48", "remaining_time": "0:08:02"}
+{"current_steps": 356, "total_steps": 375, "loss": 1.2252, "lr": 7.471398901860772e-08, "epoch": 2.848, "percentage": 94.93, "elapsed_time": "2:23:13", "remaining_time": "0:07:38"}
+{"current_steps": 357, "total_steps": 375, "loss": 1.2234, "lr": 6.744579906357185e-08, "epoch": 2.856, "percentage": 95.2, "elapsed_time": "2:23:37", "remaining_time": "0:07:14"}
+{"current_steps": 358, "total_steps": 375, "loss": 1.1826, "lr": 6.054705909137426e-08, "epoch": 2.864, "percentage": 95.47, "elapsed_time": "2:24:02", "remaining_time": "0:06:50"}
+{"current_steps": 359, "total_steps": 375, "loss": 1.1835, "lr": 5.401828581997948e-08, "epoch": 2.872, "percentage": 95.73, "elapsed_time": "2:24:24", "remaining_time": "0:06:26"}
+{"current_steps": 360, "total_steps": 375, "loss": 1.0867, "lr": 4.7859968256719344e-08, "epoch": 2.88, "percentage": 96.0, "elapsed_time": "2:24:50", "remaining_time": "0:06:02"}
+{"current_steps": 361, "total_steps": 375, "loss": 1.1871, "lr": 4.207256766166845e-08, "epoch": 2.888, "percentage": 96.27, "elapsed_time": "2:25:16", "remaining_time": "0:05:38"}
+{"current_steps": 362, "total_steps": 375, "loss": 1.2517, "lr": 3.665651751309451e-08, "epoch": 2.896, "percentage": 96.53, "elapsed_time": "2:25:40", "remaining_time": "0:05:13"}
+{"current_steps": 363, "total_steps": 375, "loss": 1.2392, "lr": 3.16122234749916e-08, "epoch": 2.904, "percentage": 96.8, "elapsed_time": "2:26:03", "remaining_time": "0:04:49"}
+{"current_steps": 364, "total_steps": 375, "loss": 1.0573, "lr": 2.6940063366693303e-08, "epoch": 2.912, "percentage": 97.07, "elapsed_time": "2:26:27", "remaining_time": "0:04:25"}
+{"current_steps": 365, "total_steps": 375, "loss": 1.1827, "lr": 2.264038713457706e-08, "epoch": 2.92, "percentage": 97.33, "elapsed_time": "2:26:52", "remaining_time": "0:04:01"}
+{"current_steps": 366, "total_steps": 375, "loss": 1.1677, "lr": 1.8713516825851207e-08, "epoch": 2.928, "percentage": 97.6, "elapsed_time": "2:27:16", "remaining_time": "0:03:37"}
+{"current_steps": 367, "total_steps": 375, "loss": 1.15, "lr": 1.51597465644332e-08, "epoch": 2.936, "percentage": 97.87, "elapsed_time": "2:27:38", "remaining_time": "0:03:13"}
+{"current_steps": 368, "total_steps": 375, "loss": 1.067, "lr": 1.1979342528922189e-08, "epoch": 2.944, "percentage": 98.13, "elapsed_time": "2:28:02", "remaining_time": "0:02:48"}
+{"current_steps": 369, "total_steps": 375, "loss": 1.1662, "lr": 9.1725429326589e-09, "epoch": 2.952, "percentage": 98.4, "elapsed_time": "2:28:28", "remaining_time": "0:02:24"}
+{"current_steps": 370, "total_steps": 375, "loss": 1.2158, "lr": 6.739558005884883e-09, "epoch": 2.96, "percentage": 98.67, "elapsed_time": "2:28:51", "remaining_time": "0:02:00"}
+{"current_steps": 371, "total_steps": 375, "loss": 1.2243, "lr": 4.6805699799967744e-09, "epoch": 2.968, "percentage": 98.93, "elapsed_time": "2:29:14", "remaining_time": "0:01:36"}
+{"current_steps": 372, "total_steps": 375, "loss": 1.2542, "lr": 2.995733073895557e-09, "epoch": 2.976, "percentage": 99.2, "elapsed_time": "2:29:37", "remaining_time": "0:01:12"}
+{"current_steps": 373, "total_steps": 375, "loss": 1.1568, "lr": 1.6851734824380184e-09, "epoch": 2.984, "percentage": 99.47, "elapsed_time": "2:30:02", "remaining_time": "0:00:48"}
+{"current_steps": 374, "total_steps": 375, "loss": 1.1133, "lr": 7.48989366980979e-10, "epoch": 2.992, "percentage": 99.73, "elapsed_time": "2:30:26", "remaining_time": "0:00:24"}
+{"current_steps": 375, "total_steps": 375, "loss": 1.1486, "lr": 1.872508480332824e-10, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "2:30:52", "remaining_time": "0:00:00"}
+{"current_steps": 375, "total_steps": 375, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "2:31:27", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,2668 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 375,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008,
+      "grad_norm": 6.165421485900879,
+      "learning_rate": 0.0,
+      "loss": 2.3147,
+      "step": 1
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 5.735071659088135,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 2.2195,
+      "step": 2
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 6.245882987976074,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 2.3793,
+      "step": 3
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 6.186586856842041,
+      "learning_rate": 2.5e-06,
+      "loss": 2.2879,
+      "step": 4
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.804585933685303,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.2428,
+      "step": 5
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 5.387149333953857,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 2.2197,
+      "step": 6
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 4.933386325836182,
+      "learning_rate": 5e-06,
+      "loss": 2.0332,
+      "step": 7
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 3.527738571166992,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 2.1247,
+      "step": 8
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 3.6139931678771973,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 2.1694,
+      "step": 9
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.918950080871582,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 2.1108,
+      "step": 10
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 2.8728139400482178,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 2.0756,
+      "step": 11
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 2.7820944786071777,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 2.0886,
+      "step": 12
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 2.8887734413146973,
+      "learning_rate": 1e-05,
+      "loss": 1.9143,
+      "step": 13
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 2.7269833087921143,
+      "learning_rate": 9.999812749151968e-06,
+      "loss": 1.7421,
+      "step": 14
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4814541339874268,
+      "learning_rate": 9.99925101063302e-06,
+      "loss": 1.9555,
+      "step": 15
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 2.188281297683716,
+      "learning_rate": 9.998314826517564e-06,
+      "loss": 1.7679,
+      "step": 16
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 2.072674036026001,
+      "learning_rate": 9.997004266926105e-06,
+      "loss": 1.8436,
+      "step": 17
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 1.929784893989563,
+      "learning_rate": 9.995319430020004e-06,
+      "loss": 1.6908,
+      "step": 18
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 1.532141089439392,
+      "learning_rate": 9.993260441994116e-06,
+      "loss": 1.8093,
+      "step": 19
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5804815292358398,
+      "learning_rate": 9.990827457067342e-06,
+      "loss": 1.8674,
+      "step": 20
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 1.6191236972808838,
+      "learning_rate": 9.988020657471078e-06,
+      "loss": 1.8742,
+      "step": 21
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 1.451058030128479,
+      "learning_rate": 9.984840253435569e-06,
+      "loss": 1.7627,
+      "step": 22
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 1.297680139541626,
+      "learning_rate": 9.98128648317415e-06,
+      "loss": 1.7632,
+      "step": 23
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.2567354440689087,
+      "learning_rate": 9.977359612865424e-06,
+      "loss": 1.7992,
+      "step": 24
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2900863885879517,
+      "learning_rate": 9.973059936633308e-06,
+      "loss": 1.7215,
+      "step": 25
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 1.2776153087615967,
+      "learning_rate": 9.968387776525009e-06,
+      "loss": 1.7456,
+      "step": 26
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 1.263024926185608,
+      "learning_rate": 9.963343482486907e-06,
+      "loss": 1.7512,
+      "step": 27
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 1.2118524312973022,
+      "learning_rate": 9.957927432338332e-06,
+      "loss": 1.801,
+      "step": 28
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 1.1955608129501343,
+      "learning_rate": 9.952140031743282e-06,
+      "loss": 1.7266,
+      "step": 29
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2853139638900757,
+      "learning_rate": 9.945981714180021e-06,
+      "loss": 1.8852,
+      "step": 30
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 1.1693278551101685,
+      "learning_rate": 9.939452940908627e-06,
+      "loss": 1.6649,
+      "step": 31
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.2286041975021362,
+      "learning_rate": 9.932554200936428e-06,
+      "loss": 1.7861,
+      "step": 32
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 1.34682297706604,
+      "learning_rate": 9.925286010981394e-06,
+      "loss": 1.7849,
+      "step": 33
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 1.2271584272384644,
+      "learning_rate": 9.917648915433413e-06,
+      "loss": 1.8268,
+      "step": 34
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2259082794189453,
+      "learning_rate": 9.909643486313533e-06,
+      "loss": 1.7176,
+      "step": 35
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.1778770685195923,
+      "learning_rate": 9.901270323231114e-06,
+      "loss": 1.7909,
+      "step": 36
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 1.1178228855133057,
+      "learning_rate": 9.892530053338909e-06,
+      "loss": 1.7618,
+      "step": 37
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 1.1982990503311157,
+      "learning_rate": 9.883423331286096e-06,
+      "loss": 1.7316,
+      "step": 38
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 1.0350874662399292,
+      "learning_rate": 9.873950839169248e-06,
+      "loss": 1.6342,
+      "step": 39
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.188277244567871,
+      "learning_rate": 9.864113286481237e-06,
+      "loss": 1.7302,
+      "step": 40
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 1.1097241640090942,
+      "learning_rate": 9.853911410058097e-06,
+      "loss": 1.5871,
+      "step": 41
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.1486974954605103,
+      "learning_rate": 9.843345974023833e-06,
+      "loss": 1.7055,
+      "step": 42
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 1.1390248537063599,
+      "learning_rate": 9.832417769733185e-06,
+      "loss": 1.725,
+      "step": 43
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 1.1239397525787354,
+      "learning_rate": 9.821127615712365e-06,
+      "loss": 1.7716,
+      "step": 44
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.095049262046814,
+      "learning_rate": 9.809476357597738e-06,
+      "loss": 1.6597,
+      "step": 45
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 1.2124348878860474,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 1.7454,
+      "step": 46
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 1.1528229713439941,
+      "learning_rate": 9.785094046801256e-06,
+      "loss": 1.813,
+      "step": 47
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.1076029539108276,
+      "learning_rate": 9.77236482036275e-06,
+      "loss": 1.6276,
+      "step": 48
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 1.1525791883468628,
+      "learning_rate": 9.759278142180348e-06,
+      "loss": 1.6772,
+      "step": 49
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1833432912826538,
+      "learning_rate": 9.745834992450688e-06,
+      "loss": 1.7564,
+      "step": 50
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 1.220745325088501,
+      "learning_rate": 9.732036378070243e-06,
+      "loss": 1.7262,
+      "step": 51
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 1.1543320417404175,
+      "learning_rate": 9.717883332559911e-06,
+      "loss": 1.5396,
+      "step": 52
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 1.1147841215133667,
+      "learning_rate": 9.703376915987601e-06,
+      "loss": 1.7466,
+      "step": 53
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 1.0171838998794556,
+      "learning_rate": 9.688518214888836e-06,
+      "loss": 1.6408,
+      "step": 54
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0049159526824951,
+      "learning_rate": 9.673308342185366e-06,
+      "loss": 1.5324,
+      "step": 55
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.069504737854004,
+      "learning_rate": 9.657748437101819e-06,
+      "loss": 1.5823,
+      "step": 56
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 1.0831701755523682,
+      "learning_rate": 9.641839665080363e-06,
+      "loss": 1.7311,
+      "step": 57
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 1.0674666166305542,
+      "learning_rate": 9.625583217693419e-06,
+      "loss": 1.8164,
+      "step": 58
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 1.0751819610595703,
+      "learning_rate": 9.60898031255441e-06,
+      "loss": 1.6059,
+      "step": 59
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0138529539108276,
+      "learning_rate": 9.592032193226564e-06,
+      "loss": 1.7617,
+      "step": 60
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 1.0597162246704102,
+      "learning_rate": 9.574740129129767e-06,
+      "loss": 1.7207,
+      "step": 61
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 1.0368247032165527,
+      "learning_rate": 9.557105415445485e-06,
+      "loss": 1.7482,
+      "step": 62
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 1.0020555257797241,
+      "learning_rate": 9.539129373019755e-06,
+      "loss": 1.7692,
+      "step": 63
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 1.0732332468032837,
+      "learning_rate": 9.520813348264252e-06,
+      "loss": 1.737,
+      "step": 64
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9923004508018494,
+      "learning_rate": 9.502158713055444e-06,
+      "loss": 1.6523,
+      "step": 65
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 1.0406478643417358,
+      "learning_rate": 9.483166864631837e-06,
+      "loss": 1.7734,
+      "step": 66
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 1.030851125717163,
+      "learning_rate": 9.46383922548932e-06,
+      "loss": 1.6183,
+      "step": 67
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 1.0091995000839233,
+      "learning_rate": 9.444177243274619e-06,
+      "loss": 1.6967,
+      "step": 68
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 1.0016194581985474,
+      "learning_rate": 9.424182390676872e-06,
+      "loss": 1.5285,
+      "step": 69
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0569149255752563,
+      "learning_rate": 9.403856165317322e-06,
+      "loss": 1.7399,
+      "step": 70
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 1.0487375259399414,
+      "learning_rate": 9.383200089637143e-06,
+      "loss": 1.6845,
+      "step": 71
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.0257350206375122,
+      "learning_rate": 9.362215710783411e-06,
+      "loss": 1.6052,
+      "step": 72
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 0.9415053129196167,
+      "learning_rate": 9.34090460049322e-06,
+      "loss": 1.6624,
+      "step": 73
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.982798159122467,
+      "learning_rate": 9.319268354975958e-06,
+      "loss": 1.581,
+      "step": 74
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0241587162017822,
+      "learning_rate": 9.297308594793757e-06,
+      "loss": 1.6495,
+      "step": 75
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 1.0798351764678955,
+      "learning_rate": 9.275026964740101e-06,
+      "loss": 1.6056,
+      "step": 76
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 1.0368587970733643,
+      "learning_rate": 9.252425133716639e-06,
+      "loss": 1.679,
+      "step": 77
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 1.019658088684082,
+      "learning_rate": 9.229504794608182e-06,
+      "loss": 1.6168,
+      "step": 78
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 0.9923626184463501,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 1.6812,
+      "step": 79
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9651780724525452,
+      "learning_rate": 9.182715482828764e-06,
+      "loss": 1.6214,
+      "step": 80
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 1.065513014793396,
+      "learning_rate": 9.158850014693123e-06,
+      "loss": 1.6537,
+      "step": 81
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.9438872933387756,
+      "learning_rate": 9.134673047280644e-06,
+      "loss": 1.602,
+      "step": 82
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 1.0228830575942993,
+      "learning_rate": 9.110186391454389e-06,
+      "loss": 1.7139,
+      "step": 83
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 1.0108044147491455,
+      "learning_rate": 9.085391881273182e-06,
+      "loss": 1.6621,
+      "step": 84
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9967405200004578,
+      "learning_rate": 9.060291373854252e-06,
+      "loss": 1.6323,
+      "step": 85
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 1.0818958282470703,
+      "learning_rate": 9.034886749234112e-06,
+      "loss": 1.6435,
+      "step": 86
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 1.012364149093628,
+      "learning_rate": 9.009179910227767e-06,
+      "loss": 1.6756,
+      "step": 87
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.084316372871399,
+      "learning_rate": 8.98317278228618e-06,
+      "loss": 1.6813,
+      "step": 88
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 1.0937138795852661,
+      "learning_rate": 8.956867313352055e-06,
+      "loss": 1.6627,
+      "step": 89
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.081781268119812,
+      "learning_rate": 8.930265473713939e-06,
+      "loss": 1.8571,
+      "step": 90
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 0.9865168929100037,
+      "learning_rate": 8.90336925585864e-06,
+      "loss": 1.4989,
+      "step": 91
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 1.0438846349716187,
+      "learning_rate": 8.876180674322006e-06,
+      "loss": 1.6913,
+      "step": 92
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 1.0505565404891968,
+      "learning_rate": 8.84870176553801e-06,
+      "loss": 1.6397,
+      "step": 93
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 0.908542275428772,
+      "learning_rate": 8.820934587686247e-06,
+      "loss": 1.5152,
+      "step": 94
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0538814067840576,
+      "learning_rate": 8.792881220537752e-06,
+      "loss": 1.5435,
+      "step": 95
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 1.0938372611999512,
+      "learning_rate": 8.764543765299245e-06,
+      "loss": 1.5283,
+      "step": 96
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 1.1297836303710938,
+      "learning_rate": 8.735924344455732e-06,
+      "loss": 1.6756,
+      "step": 97
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 0.9867210984230042,
+      "learning_rate": 8.707025101611546e-06,
+      "loss": 1.613,
+      "step": 98
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 1.0464468002319336,
+      "learning_rate": 8.677848201329775e-06,
+      "loss": 1.5942,
+      "step": 99
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0653105974197388,
+      "learning_rate": 8.64839582897015e-06,
+      "loss": 1.7216,
+      "step": 100
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 0.9832385778427124,
+      "learning_rate": 8.61867019052535e-06,
+      "loss": 1.4699,
+      "step": 101
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.9331985116004944,
+      "learning_rate": 8.588673512455781e-06,
+      "loss": 1.7439,
+      "step": 102
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 0.9516758322715759,
+      "learning_rate": 8.558408041522801e-06,
+      "loss": 1.696,
+      "step": 103
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 1.0405412912368774,
+      "learning_rate": 8.527876044620453e-06,
+      "loss": 1.713,
+      "step": 104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9289470314979553,
+      "learning_rate": 8.497079808605659e-06,
+      "loss": 1.7213,
+      "step": 105
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 0.9201377034187317,
+      "learning_rate": 8.466021640126946e-06,
+      "loss": 1.6091,
+      "step": 106
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 0.9733519554138184,
+      "learning_rate": 8.434703865451666e-06,
+      "loss": 1.6345,
+      "step": 107
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 1.0086345672607422,
+      "learning_rate": 8.403128830291767e-06,
+      "loss": 1.6443,
+      "step": 108
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 1.031723141670227,
+      "learning_rate": 8.371298899628091e-06,
+      "loss": 1.6869,
+      "step": 109
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.98936527967453,
+      "learning_rate": 8.339216457533244e-06,
+      "loss": 1.5866,
+      "step": 110
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 0.962455689907074,
+      "learning_rate": 8.306883906993022e-06,
+      "loss": 1.6483,
+      "step": 111
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.9804932475090027,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 1.594,
+      "step": 112
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 0.9484031200408936,
+      "learning_rate": 8.24147818600428e-06,
+      "loss": 1.4755,
+      "step": 113
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 1.0301589965820312,
+      "learning_rate": 8.20840991446645e-06,
+      "loss": 1.6598,
+      "step": 114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9916971325874329,
+      "learning_rate": 8.175101331937692e-06,
+      "loss": 1.5304,
+      "step": 115
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.9457253813743591,
+      "learning_rate": 8.141554933242135e-06,
+      "loss": 1.6354,
+      "step": 116
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 1.002668023109436,
+      "learning_rate": 8.10777323101642e-06,
+      "loss": 1.5989,
+      "step": 117
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 1.0599579811096191,
+      "learning_rate": 8.073758755521506e-06,
+      "loss": 1.6564,
+      "step": 118
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 0.9575379490852356,
+      "learning_rate": 8.03951405445314e-06,
+      "loss": 1.5884,
+      "step": 119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8948677182197571,
+      "learning_rate": 8.005041692751055e-06,
+      "loss": 1.5128,
+      "step": 120
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 0.9918161034584045,
+      "learning_rate": 7.970344252406832e-06,
+      "loss": 1.7014,
+      "step": 121
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 1.0054118633270264,
+      "learning_rate": 7.935424332270523e-06,
+      "loss": 1.6341,
+      "step": 122
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 0.9967445731163025,
+      "learning_rate": 7.900284547855992e-06,
+      "loss": 1.6521,
+      "step": 123
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 1.0111421346664429,
+      "learning_rate": 7.864927531145012e-06,
+      "loss": 1.7013,
+      "step": 124
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9393289685249329,
+      "learning_rate": 7.829355930390126e-06,
+      "loss": 1.5216,
+      "step": 125
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 0.9452322125434875,
+      "learning_rate": 7.7935724099163e-06,
+      "loss": 1.4109,
+      "step": 126
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 1.041080355644226,
+      "learning_rate": 7.757579649921354e-06,
+      "loss": 1.4167,
+      "step": 127
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 1.0035260915756226,
+      "learning_rate": 7.721380346275221e-06,
+      "loss": 1.4185,
+      "step": 128
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.9582903385162354,
+      "learning_rate": 7.684977210318024e-06,
+      "loss": 1.5076,
+      "step": 129
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9640058279037476,
+      "learning_rate": 7.648372968656995e-06,
+      "loss": 1.4809,
+      "step": 130
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 0.9298117160797119,
+      "learning_rate": 7.611570362962247e-06,
+      "loss": 1.4613,
+      "step": 131
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 0.922681450843811,
+      "learning_rate": 7.574572149761437e-06,
+      "loss": 1.3367,
+      "step": 132
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 0.9627285599708557,
+      "learning_rate": 7.5373811002332785e-06,
+      "loss": 1.3333,
+      "step": 133
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 0.9147132635116577,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.2811,
+      "step": 134
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.942388117313385,
+      "learning_rate": 7.462431648918689e-06,
+      "loss": 1.2576,
+      "step": 135
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.9468157291412354,
+      "learning_rate": 7.424678860871584e-06,
+      "loss": 1.3759,
+      "step": 136
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 0.8806234002113342,
+      "learning_rate": 7.3867444635553165e-06,
+      "loss": 1.2774,
+      "step": 137
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 0.962507426738739,
+      "learning_rate": 7.3486312982691134e-06,
+      "loss": 1.346,
+      "step": 138
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 0.9227285385131836,
+      "learning_rate": 7.310342219701981e-06,
+      "loss": 1.3977,
+      "step": 139
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9752129316329956,
+      "learning_rate": 7.271880095718895e-06,
+      "loss": 1.3203,
+      "step": 140
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.9896626472473145,
+      "learning_rate": 7.233247807145989e-06,
+      "loss": 1.3563,
+      "step": 141
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 0.9603149890899658,
+      "learning_rate": 7.19444824755478e-06,
+      "loss": 1.4031,
+      "step": 142
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 0.9318594336509705,
+      "learning_rate": 7.155484323045442e-06,
+      "loss": 1.3123,
+      "step": 143
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.9754460453987122,
+      "learning_rate": 7.11635895202914e-06,
+      "loss": 1.3359,
+      "step": 144
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9400285482406616,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 1.4203,
+      "step": 145
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 0.9439907073974609,
+      "learning_rate": 7.037635604362786e-06,
+      "loss": 1.3475,
+      "step": 146
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 0.9856473803520203,
+      "learning_rate": 6.9980435241181785e-06,
+      "loss": 1.3728,
+      "step": 147
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 1.0352829694747925,
+      "learning_rate": 6.958301789735853e-06,
+      "loss": 1.4341,
+      "step": 148
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 1.0069676637649536,
+      "learning_rate": 6.918413377885193e-06,
+      "loss": 1.4365,
+      "step": 149
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9149236679077148,
+      "learning_rate": 6.878381276221777e-06,
+      "loss": 1.3273,
+      "step": 150
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 0.9450361132621765,
+      "learning_rate": 6.838208483163601e-06,
+      "loss": 1.3297,
+      "step": 151
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 1.0741515159606934,
+      "learning_rate": 6.797898007666493e-06,
+      "loss": 1.3977,
+      "step": 152
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.9067620038986206,
+      "learning_rate": 6.757452868998737e-06,
+      "loss": 1.2802,
+      "step": 153
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 0.9164926409721375,
+      "learning_rate": 6.716876096514944e-06,
+      "loss": 1.3741,
+      "step": 154
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9673436880111694,
+      "learning_rate": 6.676170729429132e-06,
+      "loss": 1.2866,
+      "step": 155
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 0.8857837915420532,
+      "learning_rate": 6.635339816587109e-06,
+      "loss": 1.3524,
+      "step": 156
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 1.0112680196762085,
+      "learning_rate": 6.594386416238095e-06,
+      "loss": 1.4385,
+      "step": 157
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 0.9922051429748535,
+      "learning_rate": 6.553313595805666e-06,
+      "loss": 1.3711,
+      "step": 158
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 0.9454922676086426,
+      "learning_rate": 6.512124431658006e-06,
+      "loss": 1.392,
+      "step": 159
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9144309759140015,
+      "learning_rate": 6.470822008877482e-06,
+      "loss": 1.3345,
+      "step": 160
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 1.0063748359680176,
+      "learning_rate": 6.4294094210295725e-06,
+      "loss": 1.4048,
+      "step": 161
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 1.0005673170089722,
+      "learning_rate": 6.3878897699311525e-06,
+      "loss": 1.4097,
+      "step": 162
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 0.9271945357322693,
+      "learning_rate": 6.346266165418173e-06,
+      "loss": 1.3149,
+      "step": 163
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 0.997925341129303,
+      "learning_rate": 6.304541725112734e-06,
+      "loss": 1.4207,
+      "step": 164
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0391231775283813,
+      "learning_rate": 6.262719574189564e-06,
+      "loss": 1.4244,
+      "step": 165
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 1.0015145540237427,
+      "learning_rate": 6.2208028451419575e-06,
+      "loss": 1.2237,
+      "step": 166
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 0.9355135560035706,
+      "learning_rate": 6.178794677547138e-06,
+      "loss": 1.3877,
+      "step": 167
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.9757426381111145,
+      "learning_rate": 6.136698217831106e-06,
+      "loss": 1.4125,
+      "step": 168
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 0.9164122939109802,
+      "learning_rate": 6.094516619032975e-06,
+      "loss": 1.3165,
+      "step": 169
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.902611255645752,
+      "learning_rate": 6.052253040568804e-06,
+      "loss": 1.2087,
+      "step": 170
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 0.9562743902206421,
+      "learning_rate": 6.009910647994956e-06,
+      "loss": 1.298,
+      "step": 171
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 0.9646859169006348,
+      "learning_rate": 5.967492612770999e-06,
+      "loss": 1.478,
+      "step": 172
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 0.9848045110702515,
+      "learning_rate": 5.925002112022158e-06,
+      "loss": 1.3841,
+      "step": 173
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 0.9551551342010498,
+      "learning_rate": 5.882442328301356e-06,
+      "loss": 1.3204,
+      "step": 174
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9004006385803223,
+      "learning_rate": 5.839816449350824e-06,
+      "loss": 1.2993,
+      "step": 175
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.9321911334991455,
+      "learning_rate": 5.7971276678633625e-06,
+      "loss": 1.3931,
+      "step": 176
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.9299284219741821,
+      "learning_rate": 5.754379181243179e-06,
+      "loss": 1.3798,
+      "step": 177
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 0.9809712171554565,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 1.3581,
+      "step": 178
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 0.9836567640304565,
+      "learning_rate": 5.668715904341365e-06,
+      "loss": 1.333,
+      "step": 179
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9797928333282471,
+      "learning_rate": 5.62580753026823e-06,
+      "loss": 1.369,
+      "step": 180
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 0.8885632753372192,
+      "learning_rate": 5.5828522829987965e-06,
+      "loss": 1.207,
+      "step": 181
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 0.9868252277374268,
+      "learning_rate": 5.539853379895656e-06,
+      "loss": 1.3628,
+      "step": 182
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 0.9249109029769897,
+      "learning_rate": 5.496814041591234e-06,
+      "loss": 1.3174,
+      "step": 183
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.9424538612365723,
+      "learning_rate": 5.453737491746572e-06,
+      "loss": 1.2956,
+      "step": 184
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0274248123168945,
+      "learning_rate": 5.410626956809864e-06,
+      "loss": 1.4769,
+      "step": 185
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 0.9153403043746948,
+      "learning_rate": 5.367485665774802e-06,
+      "loss": 1.2785,
+      "step": 186
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 0.9761044383049011,
+      "learning_rate": 5.324316849938715e-06,
+      "loss": 1.3007,
+      "step": 187
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 0.9353325366973877,
+      "learning_rate": 5.281123742660558e-06,
+      "loss": 1.3652,
+      "step": 188
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 1.0336346626281738,
+      "learning_rate": 5.237909579118713e-06,
+      "loss": 1.4483,
+      "step": 189
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.968720555305481,
+      "learning_rate": 5.194677596068689e-06,
+      "loss": 1.3331,
+      "step": 190
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 0.9536193609237671,
+      "learning_rate": 5.1514310316006835e-06,
+      "loss": 1.3594,
+      "step": 191
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.9640070199966431,
+      "learning_rate": 5.1081731248970435e-06,
+      "loss": 1.3753,
+      "step": 192
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 0.8998177647590637,
+      "learning_rate": 5.064907115989655e-06,
+      "loss": 1.3207,
+      "step": 193
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 1.0492256879806519,
+      "learning_rate": 5.021636245517261e-06,
+      "loss": 1.2969,
+      "step": 194
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9381954073905945,
+      "learning_rate": 4.978363754482741e-06,
+      "loss": 1.3814,
+      "step": 195
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 0.9293028712272644,
+      "learning_rate": 4.935092884010347e-06,
+      "loss": 1.3583,
+      "step": 196
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 1.0378097295761108,
+      "learning_rate": 4.891826875102958e-06,
+      "loss": 1.4385,
+      "step": 197
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 0.9952184557914734,
+      "learning_rate": 4.848568968399317e-06,
+      "loss": 1.4347,
+      "step": 198
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 0.9569420218467712,
+      "learning_rate": 4.805322403931312e-06,
+      "loss": 1.3173,
+      "step": 199
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.0088790655136108,
+      "learning_rate": 4.762090420881289e-06,
+      "loss": 1.4581,
+      "step": 200
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.9280454516410828,
+      "learning_rate": 4.718876257339444e-06,
+      "loss": 1.3553,
+      "step": 201
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 0.9970977306365967,
+      "learning_rate": 4.6756831500612846e-06,
+      "loss": 1.2876,
+      "step": 202
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 0.9597547650337219,
+      "learning_rate": 4.632514334225201e-06,
+      "loss": 1.4021,
+      "step": 203
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 1.0041824579238892,
+      "learning_rate": 4.589373043190137e-06,
+      "loss": 1.492,
+      "step": 204
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.9870665669441223,
+      "learning_rate": 4.546262508253429e-06,
+      "loss": 1.3503,
+      "step": 205
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 0.9604246616363525,
+      "learning_rate": 4.503185958408767e-06,
+      "loss": 1.3706,
+      "step": 206
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 1.0384790897369385,
+      "learning_rate": 4.460146620104347e-06,
+      "loss": 1.2973,
+      "step": 207
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.9975892901420593,
+      "learning_rate": 4.417147717001205e-06,
+      "loss": 1.3999,
+      "step": 208
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 0.9257301092147827,
+      "learning_rate": 4.374192469731771e-06,
+      "loss": 1.3405,
+      "step": 209
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.914090633392334,
+      "learning_rate": 4.331284095658637e-06,
+      "loss": 1.2838,
+      "step": 210
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 0.9512580633163452,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 1.3207,
+      "step": 211
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 0.92391437292099,
+      "learning_rate": 4.245620818756822e-06,
+      "loss": 1.3182,
+      "step": 212
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.947001576423645,
+      "learning_rate": 4.202872332136639e-06,
+      "loss": 1.299,
+      "step": 213
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 0.9795234799385071,
+      "learning_rate": 4.160183550649176e-06,
+      "loss": 1.3005,
+      "step": 214
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9202683568000793,
+      "learning_rate": 4.117557671698648e-06,
+      "loss": 1.2494,
+      "step": 215
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.927078127861023,
+      "learning_rate": 4.074997887977843e-06,
+      "loss": 1.4299,
+      "step": 216
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 1.0304216146469116,
+      "learning_rate": 4.032507387229002e-06,
+      "loss": 1.379,
+      "step": 217
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 0.9599056839942932,
+      "learning_rate": 3.9900893520050446e-06,
+      "loss": 1.3061,
+      "step": 218
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 1.0049461126327515,
+      "learning_rate": 3.9477469594311975e-06,
+      "loss": 1.3921,
+      "step": 219
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9615343809127808,
+      "learning_rate": 3.905483380967027e-06,
+      "loss": 1.3521,
+      "step": 220
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 0.900286853313446,
+      "learning_rate": 3.863301782168896e-06,
+      "loss": 1.3919,
+      "step": 221
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 0.9626889228820801,
+      "learning_rate": 3.821205322452863e-06,
+      "loss": 1.4446,
+      "step": 222
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 0.9698411822319031,
+      "learning_rate": 3.779197154858044e-06,
+      "loss": 1.4121,
+      "step": 223
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.9630154967308044,
+      "learning_rate": 3.7372804258104367e-06,
+      "loss": 1.3486,
+      "step": 224
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9816241264343262,
+      "learning_rate": 3.695458274887268e-06,
+      "loss": 1.3536,
+      "step": 225
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 0.9554859399795532,
+      "learning_rate": 3.6537338345818273e-06,
+      "loss": 1.3906,
+      "step": 226
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 0.9753232598304749,
+      "learning_rate": 3.6121102300688504e-06,
+      "loss": 1.4416,
+      "step": 227
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 1.023533821105957,
+      "learning_rate": 3.5705905789704296e-06,
+      "loss": 1.4409,
+      "step": 228
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 0.9600017070770264,
+      "learning_rate": 3.529177991122519e-06,
+      "loss": 1.4683,
+      "step": 229
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.9704511761665344,
+      "learning_rate": 3.487875568341995e-06,
+      "loss": 1.3358,
+      "step": 230
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 0.9900558590888977,
+      "learning_rate": 3.446686404194337e-06,
+      "loss": 1.441,
+      "step": 231
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 1.0063139200210571,
+      "learning_rate": 3.4056135837619077e-06,
+      "loss": 1.3806,
+      "step": 232
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 0.9783025979995728,
+      "learning_rate": 3.3646601834128924e-06,
+      "loss": 1.2891,
+      "step": 233
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 0.9990521669387817,
+      "learning_rate": 3.3238292705708675e-06,
+      "loss": 1.3966,
+      "step": 234
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9336013197898865,
+      "learning_rate": 3.2831239034850593e-06,
+      "loss": 1.3219,
+      "step": 235
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 1.0710481405258179,
+      "learning_rate": 3.2425471310012645e-06,
+      "loss": 1.5404,
+      "step": 236
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.9238635897636414,
+      "learning_rate": 3.2021019923335093e-06,
+      "loss": 1.2135,
+      "step": 237
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 0.9419047236442566,
+      "learning_rate": 3.1617915168363994e-06,
+      "loss": 1.2094,
+      "step": 238
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 0.9390449523925781,
+      "learning_rate": 3.121618723778225e-06,
+      "loss": 1.3233,
+      "step": 239
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9554815292358398,
+      "learning_rate": 3.081586622114809e-06,
+      "loss": 1.3909,
+      "step": 240
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 0.9486646056175232,
+      "learning_rate": 3.041698210264149e-06,
+      "loss": 1.3041,
+      "step": 241
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 0.9331883788108826,
+      "learning_rate": 3.001956475881822e-06,
+      "loss": 1.256,
+      "step": 242
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 0.9461354613304138,
+      "learning_rate": 2.962364395637216e-06,
+      "loss": 1.4119,
+      "step": 243
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 0.9538270831108093,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 1.238,
+      "step": 244
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9340187907218933,
+      "learning_rate": 2.8836410479708625e-06,
+      "loss": 1.3517,
+      "step": 245
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 1.0196490287780762,
+      "learning_rate": 2.84451567695456e-06,
+      "loss": 1.2968,
+      "step": 246
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 1.013819694519043,
+      "learning_rate": 2.805551752445222e-06,
+      "loss": 1.3458,
+      "step": 247
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.9020823240280151,
+      "learning_rate": 2.766752192854012e-06,
+      "loss": 1.424,
+      "step": 248
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.8842499852180481,
+      "learning_rate": 2.728119904281105e-06,
+      "loss": 1.2664,
+      "step": 249
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9340562224388123,
+      "learning_rate": 2.689657780298019e-06,
+      "loss": 1.3434,
+      "step": 250
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 1.005499005317688,
+      "learning_rate": 2.651368701730889e-06,
+      "loss": 1.2695,
+      "step": 251
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 0.9357620477676392,
+      "learning_rate": 2.6132555364446856e-06,
+      "loss": 1.1521,
+      "step": 252
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 0.9387351870536804,
+      "learning_rate": 2.5753211391284172e-06,
+      "loss": 1.2569,
+      "step": 253
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 0.9448952674865723,
+      "learning_rate": 2.537568351081311e-06,
+      "loss": 1.1739,
+      "step": 254
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.895693302154541,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 1.1113,
+      "step": 255
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.9334677457809448,
+      "learning_rate": 2.4626188997667224e-06,
+      "loss": 1.2154,
+      "step": 256
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 0.968984067440033,
+      "learning_rate": 2.425427850238565e-06,
+      "loss": 1.1986,
+      "step": 257
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.8969867825508118,
+      "learning_rate": 2.388429637037753e-06,
+      "loss": 1.1513,
+      "step": 258
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 0.9212546348571777,
+      "learning_rate": 2.3516270313430085e-06,
+      "loss": 1.1515,
+      "step": 259
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.9310314059257507,
+      "learning_rate": 2.3150227896819782e-06,
+      "loss": 1.1398,
+      "step": 260
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 0.8860301971435547,
+      "learning_rate": 2.278619653724781e-06,
+      "loss": 1.2116,
+      "step": 261
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 0.9003278613090515,
+      "learning_rate": 2.2424203500786473e-06,
+      "loss": 1.1741,
+      "step": 262
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 0.9402487874031067,
+      "learning_rate": 2.206427590083703e-06,
+      "loss": 1.2057,
+      "step": 263
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.8900340795516968,
+      "learning_rate": 2.170644069609876e-06,
+      "loss": 1.1549,
+      "step": 264
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.9224807620048523,
+      "learning_rate": 2.1350724688549906e-06,
+      "loss": 1.2437,
+      "step": 265
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 0.9042741060256958,
+      "learning_rate": 2.09971545214401e-06,
+      "loss": 1.1539,
+      "step": 266
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 0.9493343830108643,
+      "learning_rate": 2.0645756677294788e-06,
+      "loss": 1.186,
+      "step": 267
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 0.97528475522995,
+      "learning_rate": 2.029655747593169e-06,
+      "loss": 1.2699,
+      "step": 268
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 0.914726972579956,
+      "learning_rate": 1.9949583072489455e-06,
+      "loss": 1.1699,
+      "step": 269
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.9276081323623657,
+      "learning_rate": 1.9604859455468587e-06,
+      "loss": 1.2022,
+      "step": 270
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 0.9139848947525024,
+      "learning_rate": 1.926241244478496e-06,
+      "loss": 1.1245,
+      "step": 271
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.9088538885116577,
+      "learning_rate": 1.8922267689835806e-06,
+      "loss": 1.1375,
+      "step": 272
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 0.9474396705627441,
+      "learning_rate": 1.8584450667578656e-06,
+      "loss": 1.178,
+      "step": 273
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 0.9475589394569397,
+      "learning_rate": 1.8248986680623077e-06,
+      "loss": 1.1776,
+      "step": 274
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.9256978034973145,
+      "learning_rate": 1.7915900855335506e-06,
+      "loss": 1.0931,
+      "step": 275
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 0.9370779991149902,
+      "learning_rate": 1.7585218139957205e-06,
+      "loss": 1.1597,
+      "step": 276
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 1.0122766494750977,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 1.2446,
+      "step": 277
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 0.9397243857383728,
+      "learning_rate": 1.6931160930069789e-06,
+      "loss": 1.202,
+      "step": 278
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 0.9453902840614319,
+      "learning_rate": 1.6607835424667578e-06,
+      "loss": 1.1506,
+      "step": 279
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.9595220685005188,
+      "learning_rate": 1.6287011003719105e-06,
+      "loss": 1.093,
+      "step": 280
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 0.9128125905990601,
+      "learning_rate": 1.596871169708235e-06,
+      "loss": 1.2059,
+      "step": 281
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.9542051553726196,
+      "learning_rate": 1.5652961345483353e-06,
+      "loss": 1.1799,
+      "step": 282
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 0.9437090754508972,
+      "learning_rate": 1.5339783598730568e-06,
+      "loss": 1.1929,
+      "step": 283
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 0.9622453451156616,
+      "learning_rate": 1.5029201913943425e-06,
+      "loss": 1.2047,
+      "step": 284
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.9787655472755432,
+      "learning_rate": 1.4721239553795485e-06,
+      "loss": 1.2801,
+      "step": 285
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 0.9507718682289124,
+      "learning_rate": 1.4415919584771999e-06,
+      "loss": 1.1508,
+      "step": 286
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 0.9262567162513733,
+      "learning_rate": 1.4113264875442201e-06,
+      "loss": 1.171,
+      "step": 287
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.9207803606987,
+      "learning_rate": 1.3813298094746491e-06,
+      "loss": 1.0647,
+      "step": 288
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 0.9652720093727112,
+      "learning_rate": 1.35160417102985e-06,
+      "loss": 1.21,
+      "step": 289
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.9531212449073792,
+      "learning_rate": 1.3221517986702249e-06,
+      "loss": 1.3148,
+      "step": 290
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 0.9690611362457275,
+      "learning_rate": 1.292974898388456e-06,
+      "loss": 1.133,
+      "step": 291
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 0.933113157749176,
+      "learning_rate": 1.2640756555442684e-06,
+      "loss": 1.1668,
+      "step": 292
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 0.903457522392273,
+      "learning_rate": 1.235456234700756e-06,
+      "loss": 1.1914,
+      "step": 293
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.9584729075431824,
+      "learning_rate": 1.207118779462248e-06,
+      "loss": 1.1777,
+      "step": 294
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.0123586654663086,
+      "learning_rate": 1.1790654123137552e-06,
+      "loss": 1.1946,
+      "step": 295
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 1.0037939548492432,
+      "learning_rate": 1.1512982344619904e-06,
+      "loss": 1.2927,
+      "step": 296
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 0.9953064322471619,
+      "learning_rate": 1.1238193256779955e-06,
+      "loss": 1.3264,
+      "step": 297
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 1.0182257890701294,
+      "learning_rate": 1.0966307441413598e-06,
+      "loss": 1.1768,
+      "step": 298
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 0.9684757590293884,
+      "learning_rate": 1.0697345262860638e-06,
+      "loss": 1.1419,
+      "step": 299
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.9765399098396301,
+      "learning_rate": 1.0431326866479457e-06,
+      "loss": 1.1746,
+      "step": 300
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 0.9365405440330505,
+      "learning_rate": 1.01682721771382e-06,
+      "loss": 1.2397,
+      "step": 301
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 0.9802318811416626,
+      "learning_rate": 9.908200897722332e-07,
+      "loss": 1.1136,
+      "step": 302
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 1.0139752626419067,
+      "learning_rate": 9.6511325076589e-07,
+      "loss": 1.2358,
+      "step": 303
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.9535879492759705,
+      "learning_rate": 9.397086261457511e-07,
+      "loss": 1.1577,
+      "step": 304
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.9084758162498474,
+      "learning_rate": 9.146081187268185e-07,
+      "loss": 1.1313,
+      "step": 305
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 1.0524094104766846,
+      "learning_rate": 8.898136085456127e-07,
+      "loss": 1.2469,
+      "step": 306
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 0.9283274412155151,
+      "learning_rate": 8.65326952719357e-07,
+      "loss": 1.1897,
+      "step": 307
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 0.9858782291412354,
+      "learning_rate": 8.411499853068783e-07,
+      "loss": 1.2205,
+      "step": 308
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 0.961403489112854,
+      "learning_rate": 8.172845171712379e-07,
+      "loss": 1.1335,
+      "step": 309
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.9039857387542725,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 1.1261,
+      "step": 310
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 0.8671655654907227,
+      "learning_rate": 7.70495205391818e-07,
+      "loss": 1.1149,
+      "step": 311
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.9556377530097961,
+      "learning_rate": 7.475748662833615e-07,
+      "loss": 1.2633,
+      "step": 312
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 0.9426801204681396,
+      "learning_rate": 7.249730352599e-07,
+      "loss": 1.2781,
+      "step": 313
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 0.9554891586303711,
+      "learning_rate": 7.026914052062433e-07,
+      "loss": 1.1741,
+      "step": 314
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.948947012424469,
+      "learning_rate": 6.807316450240425e-07,
+      "loss": 1.183,
+      "step": 315
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 0.9793767929077148,
+      "learning_rate": 6.590953995067812e-07,
+      "loss": 1.2157,
+      "step": 316
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 0.9236152172088623,
+      "learning_rate": 6.377842892165892e-07,
+      "loss": 1.1956,
+      "step": 317
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.89536452293396,
+      "learning_rate": 6.167999103628569e-07,
+      "loss": 1.1124,
+      "step": 318
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 0.8991439938545227,
+      "learning_rate": 5.961438346826792e-07,
+      "loss": 1.2367,
+      "step": 319
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.9654480814933777,
+      "learning_rate": 5.758176093231294e-07,
+      "loss": 1.1769,
+      "step": 320
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 0.9762253165245056,
+      "learning_rate": 5.558227567253832e-07,
+      "loss": 1.1362,
+      "step": 321
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 0.9315025210380554,
+      "learning_rate": 5.361607745106817e-07,
+      "loss": 1.0455,
+      "step": 322
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 1.0394562482833862,
+      "learning_rate": 5.168331353681643e-07,
+      "loss": 1.2309,
+      "step": 323
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 0.9258297681808472,
+      "learning_rate": 4.97841286944557e-07,
+      "loss": 1.1187,
+      "step": 324
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.9597828388214111,
+      "learning_rate": 4.791866517357491e-07,
+      "loss": 1.2398,
+      "step": 325
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 0.9284036159515381,
+      "learning_rate": 4.608706269802471e-07,
+      "loss": 1.1547,
+      "step": 326
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 1.011851191520691,
+      "learning_rate": 4.428945845545168e-07,
+      "loss": 1.2571,
+      "step": 327
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.9457777142524719,
+      "learning_rate": 4.2525987087023433e-07,
+      "loss": 1.2648,
+      "step": 328
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 0.9738786220550537,
+      "learning_rate": 4.0796780677343606e-07,
+      "loss": 1.211,
+      "step": 329
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.9281070232391357,
+      "learning_rate": 3.910196874455896e-07,
+      "loss": 1.1852,
+      "step": 330
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 0.9425092339515686,
+      "learning_rate": 3.744167823065814e-07,
+      "loss": 1.2286,
+      "step": 331
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 0.9291054606437683,
+      "learning_rate": 3.581603349196372e-07,
+      "loss": 1.1621,
+      "step": 332
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 0.92279052734375,
+      "learning_rate": 3.4225156289818096e-07,
+      "loss": 1.2425,
+      "step": 333
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 0.9327970743179321,
+      "learning_rate": 3.26691657814634e-07,
+      "loss": 1.1193,
+      "step": 334
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.9604306817054749,
+      "learning_rate": 3.1148178511116624e-07,
+      "loss": 1.1053,
+      "step": 335
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 1.009303331375122,
+      "learning_rate": 2.966230840124007e-07,
+      "loss": 1.1874,
+      "step": 336
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 0.9507207274436951,
+      "learning_rate": 2.821166674400905e-07,
+      "loss": 1.1105,
+      "step": 337
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 0.9599238634109497,
+      "learning_rate": 2.6796362192975766e-07,
+      "loss": 1.1718,
+      "step": 338
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 0.8968420624732971,
+      "learning_rate": 2.5416500754931294e-07,
+      "loss": 1.0511,
+      "step": 339
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.9986500144004822,
+      "learning_rate": 2.407218578196524e-07,
+      "loss": 1.2698,
+      "step": 340
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 0.9403897523880005,
+      "learning_rate": 2.2763517963725169e-07,
+      "loss": 1.2182,
+      "step": 341
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.9974818825721741,
+      "learning_rate": 2.1490595319874574e-07,
+      "loss": 1.2383,
+      "step": 342
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 1.008009910583496,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 1.1606,
+      "step": 343
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 1.019026279449463,
+      "learning_rate": 1.905236424022633e-07,
+      "loss": 1.1782,
+      "step": 344
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.9780669808387756,
+      "learning_rate": 1.7887238428763553e-07,
+      "loss": 1.1847,
+      "step": 345
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 1.0105910301208496,
+      "learning_rate": 1.6758223026681507e-07,
+      "loss": 1.2057,
+      "step": 346
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 0.934592068195343,
+      "learning_rate": 1.5665402597616842e-07,
+      "loss": 1.0974,
+      "step": 347
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 0.9480429887771606,
+      "learning_rate": 1.4608858994190344e-07,
+      "loss": 1.169,
+      "step": 348
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 0.8923419713973999,
+      "learning_rate": 1.358867135187636e-07,
+      "loss": 1.1029,
+      "step": 349
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.0625927448272705,
+      "learning_rate": 1.2604916083075236e-07,
+      "loss": 1.1654,
+      "step": 350
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 0.9121294617652893,
+      "learning_rate": 1.1657666871390471e-07,
+      "loss": 1.0817,
+      "step": 351
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 1.002794861793518,
+      "learning_rate": 1.0746994666109234e-07,
+      "loss": 1.3023,
+      "step": 352
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 0.9711022973060608,
+      "learning_rate": 9.872967676888611e-08,
+      "loss": 1.1863,
+      "step": 353
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.9451983571052551,
+      "learning_rate": 9.035651368646647e-08,
+      "loss": 1.1599,
+      "step": 354
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.9576354026794434,
+      "learning_rate": 8.235108456658814e-08,
+      "loss": 1.2885,
+      "step": 355
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 0.9445213675498962,
+      "learning_rate": 7.471398901860772e-08,
+      "loss": 1.2252,
+      "step": 356
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 1.0621612071990967,
+      "learning_rate": 6.744579906357185e-08,
+      "loss": 1.2234,
+      "step": 357
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 0.920613169670105,
+      "learning_rate": 6.054705909137426e-08,
+      "loss": 1.1826,
+      "step": 358
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 1.0022947788238525,
+      "learning_rate": 5.401828581997948e-08,
+      "loss": 1.1835,
+      "step": 359
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.9064760208129883,
+      "learning_rate": 4.7859968256719344e-08,
+      "loss": 1.0867,
+      "step": 360
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 0.9370819926261902,
+      "learning_rate": 4.207256766166845e-08,
+      "loss": 1.1871,
+      "step": 361
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 1.0383437871932983,
+      "learning_rate": 3.665651751309451e-08,
+      "loss": 1.2517,
+      "step": 362
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 0.962099552154541,
+      "learning_rate": 3.16122234749916e-08,
+      "loss": 1.2392,
+      "step": 363
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 0.9519680142402649,
+      "learning_rate": 2.6940063366693303e-08,
+      "loss": 1.0573,
+      "step": 364
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.9429070353507996,
+      "learning_rate": 2.264038713457706e-08,
+      "loss": 1.1827,
+      "step": 365
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 0.9092634916305542,
+      "learning_rate": 1.8713516825851207e-08,
+      "loss": 1.1677,
+      "step": 366
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 0.9508703351020813,
+      "learning_rate": 1.51597465644332e-08,
+      "loss": 1.15,
+      "step": 367
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.9615942239761353,
+      "learning_rate": 1.1979342528922189e-08,
+      "loss": 1.067,
+      "step": 368
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 0.9065539240837097,
+      "learning_rate": 9.1725429326589e-09,
+      "loss": 1.1662,
+      "step": 369
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.9946824312210083,
+      "learning_rate": 6.739558005884883e-09,
+      "loss": 1.2158,
+      "step": 370
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 0.9945969581604004,
+      "learning_rate": 4.6805699799967744e-09,
+      "loss": 1.2243,
+      "step": 371
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 1.0351194143295288,
+      "learning_rate": 2.995733073895557e-09,
+      "loss": 1.2542,
+      "step": 372
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 0.9449812769889832,
+      "learning_rate": 1.6851734824380184e-09,
+      "loss": 1.1568,
+      "step": 373
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 0.9165475368499756,
+      "learning_rate": 7.48989366980979e-10,
+      "loss": 1.1133,
+      "step": 374
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.0071150064468384,
+      "learning_rate": 1.872508480332824e-10,
+      "loss": 1.1486,
+      "step": 375
+    },
+    {
+      "epoch": 3.0,
+      "step": 375,
+      "total_flos": 62777166004224.0,
+      "train_loss": 1.4248257904052735,
+      "train_runtime": 9087.8291,
+      "train_samples_per_second": 1.32,
+      "train_steps_per_second": 0.041
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 62777166004224.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9dba639380f4aa9ea0f3bdbd42cabf233ebb9248a707953be640961e956b1fa6
+size 8401