Upload folder using huggingface_hub

README.md

@@ -12,37 +12,47 @@ model-index:
 <!-- 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. -->
 
+# core2
 
-# Model Card: LLama 2 - Version 7b (Embedding + Output + 1 Hidden Layer)
+This model is a fine-tuned version of [./core2](https://huggingface.co/./core2) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.7608
+- Accuracy: 0.4077
 
-## Overview
+## Model description
 
-- **Link to Training Progress:** [WandB Training Progress](https://wandb.ai/inteligenciaartificialcursos/huggingface/runs/pv74nzw8?workspace=user-inteligenciaartificialcursos)
+More information needed
 
-- **Model Name:** LLama 2 - Version 7b
+## Intended uses & limitations
 
-- **Total Parameters:** 446 million
+More information needed
 
-## Training Data
+## Training and evaluation data
 
-The model has been trained with the following sequence of datasets:
+More information needed
 
-1. **GPT-2 Data (DONE):** The initial training phase involves GPT-2 data and is currently in the finalization stage.
+## Training procedure
 
-2. **Wikipedia QA in Markdown (In Progress):** The model's training will continue with Wikipedia question-answering data in Markdown format.
+### Training hyperparameters
 
-3. **QA with Rhetoric (Future Stages):** The model will further be fine-tuned with question-answering data generated from various LLama models, incorporating rhetorical elements.
+The following hyperparameters were used during training:
+- learning_rate: 0.0001
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 32
+- total_train_batch_size: 32
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 1.0
 
-## Model Description
+### Training results
 
-The LLama 2 - Version 7b model is a powerful language model with a total of 446 million parameters. It utilizes embeddings, an output layer, and one hidden layer to perform a wide range of natural language processing tasks. The training is conducted in multiple stages, each focused on different datasets and objectives.
 
-## Disclaimer
 
-This model card provides an overview of the LLama 2 - Version 7b model, its training data, and intended use cases. Keep in mind that the model's performance may vary depending on the specific task or dataset. Users are encouraged to evaluate the model's suitability for their applications and exercise caution when using it in real-world scenarios.
+### Framework versions
 
-For any further inquiries or issues related to this model, please contact the model developers through the provided training progress link.
-
----
-
-Feel free to customize this Model Card further if you have additional details or specific use cases you'd like to highlight.
+- Transformers 4.34.0.dev0
+- Pytorch 2.0.1+cu117
+- Datasets 2.14.5
+- Tokenizers 0.13.3

all_results.json

@@ -1,15 +1,15 @@
 {
     "epoch": 1.0,
-    "eval_accuracy": 0.4413527624330325,
-    "eval_loss": 2.4912242889404297,
-    "eval_runtime": 1.7817,
+    "eval_accuracy": 0.4077155652549501,
+    "eval_loss": 2.7608211040496826,
+    "eval_runtime": 1.7213,
     "eval_samples": 129,
-    "eval_samples_per_second": 72.402,
-    "eval_steps_per_second": 9.541,
-    "perplexity": 12.076051650319789,
-    "train_loss": 2.5232514588412474,
-    "train_runtime": 5052.6511,
-    "train_samples": 175957,
-    "train_samples_per_second": 34.825,
-    "train_steps_per_second": 4.353
+    "eval_samples_per_second": 74.942,
+    "eval_steps_per_second": 9.876,
+    "perplexity": 15.81282159096841,
+    "train_loss": 2.7717256223521947,
+    "train_runtime": 3102.8464,
+    "train_samples": 119233,
+    "train_samples_per_second": 38.427,
+    "train_steps_per_second": 1.201
 }

eval_results.json

@@ -1,10 +1,10 @@
 {
     "epoch": 1.0,
-    "eval_accuracy": 0.4413527624330325,
-    "eval_loss": 2.4912242889404297,
-    "eval_runtime": 1.7817,
+    "eval_accuracy": 0.4077155652549501,
+    "eval_loss": 2.7608211040496826,
+    "eval_runtime": 1.7213,
     "eval_samples": 129,
-    "eval_samples_per_second": 72.402,
-    "eval_steps_per_second": 9.541,
-    "perplexity": 12.076051650319789
+    "eval_samples_per_second": 74.942,
+    "eval_steps_per_second": 9.876,
+    "perplexity": 15.81282159096841
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bb40c9bdf099a92651835537db6b802001ee4136b8b3ec0472fca0082aba72d7
+oid sha256:451f74d625df52f59037a15db1a90de07168091802624afed9235768562eba88
 size 929067029

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
-    "train_loss": 2.5232514588412474,
-    "train_runtime": 5052.6511,
-    "train_samples": 175957,
-    "train_samples_per_second": 34.825,
-    "train_steps_per_second": 4.353
+    "train_loss": 2.7717256223521947,
+    "train_runtime": 3102.8464,
+    "train_samples": 119233,
+    "train_samples_per_second": 38.427,
+    "train_steps_per_second": 1.201
 }

trainer_state.json

@@ -1,2662 +1,472 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.9999715839665373,
+  "epoch": 0.9999916130601427,
   "eval_steps": 500,
-  "global_step": 21994,
+  "global_step": 3726,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0,
-      "learning_rate": 9.977266527234701e-05,
-      "loss": 2.6296,
+      "epoch": 0.01,
+      "learning_rate": 9.86580783682233e-05,
+      "loss": 3.6768,
       "step": 50
     },
     {
-      "epoch": 0.0,
-      "learning_rate": 9.954533054469402e-05,
-      "loss": 2.6156,
+      "epoch": 0.03,
+      "learning_rate": 9.731615673644659e-05,
+      "loss": 3.2865,
       "step": 100
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 9.931799581704102e-05,
-      "loss": 2.6035,
+      "epoch": 0.04,
+      "learning_rate": 9.597423510466989e-05,
+      "loss": 3.1518,
       "step": 150
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 9.909066108938801e-05,
-      "loss": 2.5399,
+      "epoch": 0.05,
+      "learning_rate": 9.463231347289319e-05,
+      "loss": 3.1091,
       "step": 200
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 9.886332636173502e-05,
-      "loss": 2.5857,
+      "epoch": 0.07,
+      "learning_rate": 9.329039184111649e-05,
+      "loss": 3.0456,
       "step": 250
     },
     {
-      "epoch": 0.01,
-      "learning_rate": 9.863599163408202e-05,
-      "loss": 2.6078,
+      "epoch": 0.08,
+      "learning_rate": 9.194847020933978e-05,
+      "loss": 3.0357,
       "step": 300
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 9.840865690642903e-05,
-      "loss": 2.5931,
+      "epoch": 0.09,
+      "learning_rate": 9.060654857756307e-05,
+      "loss": 2.9846,
       "step": 350
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 9.818132217877604e-05,
-      "loss": 2.5919,
+      "epoch": 0.11,
+      "learning_rate": 8.926462694578636e-05,
+      "loss": 2.9834,
       "step": 400
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 9.795398745112304e-05,
-      "loss": 2.59,
+      "epoch": 0.12,
+      "learning_rate": 8.792270531400967e-05,
+      "loss": 2.9385,
       "step": 450
     },
     {
-      "epoch": 0.02,
-      "learning_rate": 9.772665272347005e-05,
-      "loss": 2.605,
+      "epoch": 0.13,
+      "learning_rate": 8.658078368223296e-05,
+      "loss": 2.8967,
       "step": 500
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.749931799581704e-05,
-      "loss": 2.6026,
+      "epoch": 0.15,
+      "learning_rate": 8.523886205045626e-05,
+      "loss": 2.886,
       "step": 550
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.727198326816404e-05,
-      "loss": 2.5839,
+      "epoch": 0.16,
+      "learning_rate": 8.389694041867955e-05,
+      "loss": 2.8671,
       "step": 600
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.704464854051105e-05,
-      "loss": 2.5862,
+      "epoch": 0.17,
+      "learning_rate": 8.255501878690284e-05,
+      "loss": 2.8547,
       "step": 650
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.681731381285806e-05,
-      "loss": 2.609,
+      "epoch": 0.19,
+      "learning_rate": 8.121309715512614e-05,
+      "loss": 2.8396,
       "step": 700
     },
     {
-      "epoch": 0.03,
-      "learning_rate": 9.658997908520506e-05,
-      "loss": 2.5759,
+      "epoch": 0.2,
+      "learning_rate": 7.987117552334944e-05,
+      "loss": 2.8531,
       "step": 750
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 9.636264435755207e-05,
-      "loss": 2.6046,
+      "epoch": 0.21,
+      "learning_rate": 7.852925389157274e-05,
+      "loss": 2.8196,
       "step": 800
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 9.613530962989907e-05,
-      "loss": 2.5811,
+      "epoch": 0.23,
+      "learning_rate": 7.718733225979603e-05,
+      "loss": 2.7921,
       "step": 850
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 9.590797490224606e-05,
-      "loss": 2.5797,
+      "epoch": 0.24,
+      "learning_rate": 7.584541062801933e-05,
+      "loss": 2.7596,
       "step": 900
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 9.568064017459307e-05,
-      "loss": 2.5867,
+      "epoch": 0.25,
+      "learning_rate": 7.450348899624262e-05,
+      "loss": 2.7918,
       "step": 950
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 9.545330544694008e-05,
-      "loss": 2.5927,
+      "epoch": 0.27,
+      "learning_rate": 7.316156736446593e-05,
+      "loss": 2.7553,
       "step": 1000
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 9.522597071928708e-05,
-      "loss": 2.568,
+      "epoch": 0.28,
+      "learning_rate": 7.181964573268921e-05,
+      "loss": 2.7914,
       "step": 1050
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 9.499863599163409e-05,
-      "loss": 2.6024,
+      "epoch": 0.3,
+      "learning_rate": 7.047772410091251e-05,
+      "loss": 2.7924,
       "step": 1100
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 9.477130126398109e-05,
-      "loss": 2.5936,
+      "epoch": 0.31,
+      "learning_rate": 6.91358024691358e-05,
+      "loss": 2.7823,
       "step": 1150
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 9.45439665363281e-05,
-      "loss": 2.605,
+      "epoch": 0.32,
+      "learning_rate": 6.77938808373591e-05,
+      "loss": 2.7437,
       "step": 1200
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 9.431663180867509e-05,
-      "loss": 2.5775,
+      "epoch": 0.34,
+      "learning_rate": 6.64519592055824e-05,
+      "loss": 2.7404,
       "step": 1250
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 9.40892970810221e-05,
-      "loss": 2.5752,
+      "epoch": 0.35,
+      "learning_rate": 6.51100375738057e-05,
+      "loss": 2.7318,
       "step": 1300
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 9.38619623533691e-05,
-      "loss": 2.5679,
+      "epoch": 0.36,
+      "learning_rate": 6.376811594202898e-05,
+      "loss": 2.708,
       "step": 1350
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 9.36346276257161e-05,
-      "loss": 2.5856,
+      "epoch": 0.38,
+      "learning_rate": 6.242619431025228e-05,
+      "loss": 2.7579,
       "step": 1400
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 9.340729289806311e-05,
-      "loss": 2.5787,
+      "epoch": 0.39,
+      "learning_rate": 6.108427267847558e-05,
+      "loss": 2.7037,
       "step": 1450
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 9.317995817041012e-05,
-      "loss": 2.5875,
+      "epoch": 0.4,
+      "learning_rate": 5.9742351046698876e-05,
+      "loss": 2.7326,
       "step": 1500
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 9.295262344275712e-05,
-      "loss": 2.5631,
+      "epoch": 0.42,
+      "learning_rate": 5.8400429414922176e-05,
+      "loss": 2.7252,
       "step": 1550
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 9.272528871510412e-05,
-      "loss": 2.583,
+      "epoch": 0.43,
+      "learning_rate": 5.705850778314546e-05,
+      "loss": 2.7263,
       "step": 1600
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 9.249795398745112e-05,
-      "loss": 2.5609,
+      "epoch": 0.44,
+      "learning_rate": 5.571658615136877e-05,
+      "loss": 2.6944,
       "step": 1650
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 9.227061925979813e-05,
-      "loss": 2.587,
+      "epoch": 0.46,
+      "learning_rate": 5.4374664519592054e-05,
+      "loss": 2.7292,
       "step": 1700
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 9.204328453214513e-05,
-      "loss": 2.5555,
+      "epoch": 0.47,
+      "learning_rate": 5.3032742887815354e-05,
+      "loss": 2.7156,
       "step": 1750
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 9.181594980449214e-05,
-      "loss": 2.5488,
+      "epoch": 0.48,
+      "learning_rate": 5.1690821256038647e-05,
+      "loss": 2.6852,
       "step": 1800
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 9.158861507683914e-05,
-      "loss": 2.5554,
+      "epoch": 0.5,
+      "learning_rate": 5.0348899624261946e-05,
+      "loss": 2.6922,
       "step": 1850
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 9.136128034918615e-05,
-      "loss": 2.5408,
+      "epoch": 0.51,
+      "learning_rate": 4.9006977992485246e-05,
+      "loss": 2.7217,
       "step": 1900
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 9.113394562153314e-05,
-      "loss": 2.582,
+      "epoch": 0.52,
+      "learning_rate": 4.766505636070854e-05,
+      "loss": 2.6923,
       "step": 1950
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 9.090661089388015e-05,
-      "loss": 2.5533,
+      "epoch": 0.54,
+      "learning_rate": 4.632313472893183e-05,
+      "loss": 2.6993,
       "step": 2000
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 9.067927616622715e-05,
-      "loss": 2.5432,
+      "epoch": 0.55,
+      "learning_rate": 4.498121309715513e-05,
+      "loss": 2.7156,
       "step": 2050
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 9.045194143857416e-05,
-      "loss": 2.5867,
+      "epoch": 0.56,
+      "learning_rate": 4.3639291465378424e-05,
+      "loss": 2.6932,
       "step": 2100
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 9.022460671092116e-05,
-      "loss": 2.5343,
+      "epoch": 0.58,
+      "learning_rate": 4.2297369833601716e-05,
+      "loss": 2.714,
       "step": 2150
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 8.999727198326817e-05,
-      "loss": 2.585,
+      "epoch": 0.59,
+      "learning_rate": 4.0955448201825016e-05,
+      "loss": 2.6921,
       "step": 2200
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 8.976993725561517e-05,
-      "loss": 2.5679,
+      "epoch": 0.6,
+      "learning_rate": 3.961352657004831e-05,
+      "loss": 2.6949,
       "step": 2250
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 8.954260252796217e-05,
-      "loss": 2.5515,
+      "epoch": 0.62,
+      "learning_rate": 3.82716049382716e-05,
+      "loss": 2.694,
       "step": 2300
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 8.931526780030917e-05,
-      "loss": 2.5713,
+      "epoch": 0.63,
+      "learning_rate": 3.69296833064949e-05,
+      "loss": 2.6781,
       "step": 2350
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 8.908793307265618e-05,
-      "loss": 2.5587,
+      "epoch": 0.64,
+      "learning_rate": 3.55877616747182e-05,
+      "loss": 2.6932,
       "step": 2400
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 8.886059834500318e-05,
-      "loss": 2.5774,
+      "epoch": 0.66,
+      "learning_rate": 3.4245840042941493e-05,
+      "loss": 2.6748,
       "step": 2450
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 8.863326361735019e-05,
-      "loss": 2.551,
+      "epoch": 0.67,
+      "learning_rate": 3.290391841116479e-05,
+      "loss": 2.6861,
       "step": 2500
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 8.84059288896972e-05,
-      "loss": 2.5685,
+      "epoch": 0.68,
+      "learning_rate": 3.1561996779388086e-05,
+      "loss": 2.6891,
       "step": 2550
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 8.81785941620442e-05,
-      "loss": 2.5707,
+      "epoch": 0.7,
+      "learning_rate": 3.0220075147611382e-05,
+      "loss": 2.7091,
       "step": 2600
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 8.795125943439119e-05,
-      "loss": 2.568,
+      "epoch": 0.71,
+      "learning_rate": 2.8878153515834678e-05,
+      "loss": 2.6993,
       "step": 2650
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 8.77239247067382e-05,
-      "loss": 2.5536,
+      "epoch": 0.72,
+      "learning_rate": 2.753623188405797e-05,
+      "loss": 2.6827,
       "step": 2700
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 8.74965899790852e-05,
-      "loss": 2.5406,
+      "epoch": 0.74,
+      "learning_rate": 2.6194310252281267e-05,
+      "loss": 2.6876,
       "step": 2750
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 8.726925525143221e-05,
-      "loss": 2.5572,
+      "epoch": 0.75,
+      "learning_rate": 2.4852388620504563e-05,
+      "loss": 2.6812,
       "step": 2800
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 8.704192052377921e-05,
-      "loss": 2.5749,
+      "epoch": 0.76,
+      "learning_rate": 2.351046698872786e-05,
+      "loss": 2.6936,
       "step": 2850
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 8.681458579612622e-05,
-      "loss": 2.5607,
+      "epoch": 0.78,
+      "learning_rate": 2.2168545356951156e-05,
+      "loss": 2.6579,
       "step": 2900
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 8.658725106847322e-05,
-      "loss": 2.5612,
+      "epoch": 0.79,
+      "learning_rate": 2.0826623725174452e-05,
+      "loss": 2.6786,
       "step": 2950
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 8.635991634082023e-05,
-      "loss": 2.5626,
+      "epoch": 0.81,
+      "learning_rate": 1.9484702093397748e-05,
+      "loss": 2.6974,
       "step": 3000
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 8.613258161316724e-05,
-      "loss": 2.5677,
+      "epoch": 0.82,
+      "learning_rate": 1.814278046162104e-05,
+      "loss": 2.6818,
       "step": 3050
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 8.590524688551423e-05,
-      "loss": 2.5072,
+      "epoch": 0.83,
+      "learning_rate": 1.6800858829844337e-05,
+      "loss": 2.6742,
       "step": 3100
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 8.567791215786123e-05,
-      "loss": 2.562,
+      "epoch": 0.85,
+      "learning_rate": 1.5458937198067633e-05,
+      "loss": 2.6748,
       "step": 3150
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 8.545057743020824e-05,
-      "loss": 2.5786,
+      "epoch": 0.86,
+      "learning_rate": 1.4117015566290927e-05,
+      "loss": 2.6771,
       "step": 3200
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 8.522324270255524e-05,
-      "loss": 2.5388,
+      "epoch": 0.87,
+      "learning_rate": 1.2775093934514227e-05,
+      "loss": 2.6895,
       "step": 3250
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 8.499590797490225e-05,
-      "loss": 2.538,
+      "epoch": 0.89,
+      "learning_rate": 1.143317230273752e-05,
+      "loss": 2.6663,
       "step": 3300
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 8.476857324724926e-05,
-      "loss": 2.5448,
+      "epoch": 0.9,
+      "learning_rate": 1.0091250670960816e-05,
+      "loss": 2.6485,
       "step": 3350
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 8.454123851959626e-05,
-      "loss": 2.5306,
+      "epoch": 0.91,
+      "learning_rate": 8.749329039184112e-06,
+      "loss": 2.6825,
       "step": 3400
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 8.431390379194327e-05,
-      "loss": 2.5647,
+      "epoch": 0.93,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 2.6832,
       "step": 3450
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 8.408656906429027e-05,
-      "loss": 2.5386,
+      "epoch": 0.94,
+      "learning_rate": 6.065485775630704e-06,
+      "loss": 2.7134,
       "step": 3500
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 8.385923433663728e-05,
-      "loss": 2.5376,
+      "epoch": 0.95,
+      "learning_rate": 4.723564143853999e-06,
+      "loss": 2.6631,
       "step": 3550
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 8.363189960898427e-05,
-      "loss": 2.535,
+      "epoch": 0.97,
+      "learning_rate": 3.3816425120772947e-06,
+      "loss": 2.6862,
       "step": 3600
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 8.340456488133128e-05,
-      "loss": 2.5476,
+      "epoch": 0.98,
+      "learning_rate": 2.0397208803005905e-06,
+      "loss": 2.6663,
       "step": 3650
     },
     {
-      "epoch": 0.17,
-      "learning_rate": 8.317723015367828e-05,
-      "loss": 2.5462,
+      "epoch": 0.99,
+      "learning_rate": 6.977992485238862e-07,
+      "loss": 2.6592,
       "step": 3700
     },
-    {
-      "epoch": 0.17,
-      "learning_rate": 8.294989542602529e-05,
-      "loss": 2.5795,
-      "step": 3750
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 8.272256069837229e-05,
-      "loss": 2.5425,
-      "step": 3800
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 8.24952259707193e-05,
-      "loss": 2.5663,
-      "step": 3850
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 8.22678912430663e-05,
-      "loss": 2.5376,
-      "step": 3900
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 8.204055651541331e-05,
-      "loss": 2.5695,
-      "step": 3950
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 8.18132217877603e-05,
-      "loss": 2.5112,
-      "step": 4000
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 8.15858870601073e-05,
-      "loss": 2.5255,
-      "step": 4050
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 8.135855233245431e-05,
-      "loss": 2.5448,
-      "step": 4100
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 8.113121760480132e-05,
-      "loss": 2.5483,
-      "step": 4150
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 8.090388287714832e-05,
-      "loss": 2.5319,
-      "step": 4200
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 8.067654814949533e-05,
-      "loss": 2.5655,
-      "step": 4250
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 8.044921342184233e-05,
-      "loss": 2.5399,
-      "step": 4300
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 8.022187869418933e-05,
-      "loss": 2.5485,
-      "step": 4350
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 7.999454396653633e-05,
-      "loss": 2.5196,
-      "step": 4400
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 7.976720923888334e-05,
-      "loss": 2.5554,
-      "step": 4450
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 7.953987451123034e-05,
-      "loss": 2.583,
-      "step": 4500
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 7.931253978357735e-05,
-      "loss": 2.5643,
-      "step": 4550
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 7.908520505592435e-05,
-      "loss": 2.5345,
-      "step": 4600
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 7.885787032827136e-05,
-      "loss": 2.5393,
-      "step": 4650
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 7.863053560061835e-05,
-      "loss": 2.5349,
-      "step": 4700
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 7.840320087296536e-05,
-      "loss": 2.54,
-      "step": 4750
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 7.817586614531236e-05,
-      "loss": 2.5526,
-      "step": 4800
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 7.794853141765937e-05,
-      "loss": 2.5419,
-      "step": 4850
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 7.772119669000637e-05,
-      "loss": 2.5122,
-      "step": 4900
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 7.749386196235338e-05,
-      "loss": 2.5247,
-      "step": 4950
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 7.726652723470039e-05,
-      "loss": 2.5516,
-      "step": 5000
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 7.703919250704738e-05,
-      "loss": 2.5321,
-      "step": 5050
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 7.681185777939438e-05,
-      "loss": 2.5453,
-      "step": 5100
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 7.658452305174139e-05,
-      "loss": 2.5453,
-      "step": 5150
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 7.63571883240884e-05,
-      "loss": 2.5522,
-      "step": 5200
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 7.61298535964354e-05,
-      "loss": 2.5417,
-      "step": 5250
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 7.59025188687824e-05,
-      "loss": 2.5241,
-      "step": 5300
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 7.567518414112941e-05,
-      "loss": 2.5574,
-      "step": 5350
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 7.54478494134764e-05,
-      "loss": 2.5127,
-      "step": 5400
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 7.522051468582341e-05,
-      "loss": 2.5346,
-      "step": 5450
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 7.499317995817041e-05,
-      "loss": 2.5164,
-      "step": 5500
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 7.476584523051742e-05,
-      "loss": 2.5571,
-      "step": 5550
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 7.453851050286442e-05,
-      "loss": 2.5455,
-      "step": 5600
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 7.431117577521143e-05,
-      "loss": 2.544,
-      "step": 5650
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 7.408384104755844e-05,
-      "loss": 2.5271,
-      "step": 5700
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 7.385650631990543e-05,
-      "loss": 2.525,
-      "step": 5750
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 7.362917159225243e-05,
-      "loss": 2.5278,
-      "step": 5800
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 7.340183686459944e-05,
-      "loss": 2.5161,
-      "step": 5850
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 7.317450213694644e-05,
-      "loss": 2.5296,
-      "step": 5900
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 7.294716740929345e-05,
-      "loss": 2.5454,
-      "step": 5950
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 7.271983268164046e-05,
-      "loss": 2.5319,
-      "step": 6000
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 7.249249795398746e-05,
-      "loss": 2.5282,
-      "step": 6050
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 7.226516322633445e-05,
-      "loss": 2.5359,
-      "step": 6100
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 7.203782849868146e-05,
-      "loss": 2.494,
-      "step": 6150
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 7.181049377102846e-05,
-      "loss": 2.5289,
-      "step": 6200
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 7.158315904337547e-05,
-      "loss": 2.4985,
-      "step": 6250
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 7.135582431572248e-05,
-      "loss": 2.5156,
-      "step": 6300
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 7.112848958806948e-05,
-      "loss": 2.53,
-      "step": 6350
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 7.090115486041649e-05,
-      "loss": 2.5157,
-      "step": 6400
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 7.067382013276348e-05,
-      "loss": 2.5303,
-      "step": 6450
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 7.044648540511048e-05,
-      "loss": 2.5286,
-      "step": 6500
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 7.021915067745749e-05,
-      "loss": 2.5039,
-      "step": 6550
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 6.99918159498045e-05,
-      "loss": 2.5161,
-      "step": 6600
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 6.97644812221515e-05,
-      "loss": 2.5105,
-      "step": 6650
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 6.95371464944985e-05,
-      "loss": 2.5151,
-      "step": 6700
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 6.930981176684551e-05,
-      "loss": 2.5425,
-      "step": 6750
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 6.90824770391925e-05,
-      "loss": 2.5357,
-      "step": 6800
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 6.885514231153951e-05,
-      "loss": 2.4989,
-      "step": 6850
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 6.862780758388652e-05,
-      "loss": 2.5413,
-      "step": 6900
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 6.840047285623352e-05,
-      "loss": 2.4909,
-      "step": 6950
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 6.817313812858053e-05,
-      "loss": 2.5177,
-      "step": 7000
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 6.794580340092753e-05,
-      "loss": 2.5107,
-      "step": 7050
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 6.771846867327454e-05,
-      "loss": 2.5343,
-      "step": 7100
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 6.749113394562153e-05,
-      "loss": 2.5247,
-      "step": 7150
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 6.726379921796854e-05,
-      "loss": 2.5202,
-      "step": 7200
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 6.703646449031554e-05,
-      "loss": 2.5156,
-      "step": 7250
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 6.680912976266255e-05,
-      "loss": 2.5431,
-      "step": 7300
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 6.658179503500955e-05,
-      "loss": 2.5221,
-      "step": 7350
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 6.635446030735656e-05,
-      "loss": 2.516,
-      "step": 7400
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 6.612712557970356e-05,
-      "loss": 2.5297,
-      "step": 7450
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 6.589979085205056e-05,
-      "loss": 2.5052,
-      "step": 7500
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 6.567245612439756e-05,
-      "loss": 2.4981,
-      "step": 7550
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.544512139674457e-05,
-      "loss": 2.5292,
-      "step": 7600
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.521778666909157e-05,
-      "loss": 2.4853,
-      "step": 7650
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.499045194143858e-05,
-      "loss": 2.5181,
-      "step": 7700
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.476311721378558e-05,
-      "loss": 2.5599,
-      "step": 7750
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.453578248613259e-05,
-      "loss": 2.5093,
-      "step": 7800
-    },
-    {
-      "epoch": 0.36,
-      "learning_rate": 6.430844775847958e-05,
-      "loss": 2.5449,
-      "step": 7850
-    },
-    {
-      "epoch": 0.36,
-      "learning_rate": 6.408111303082659e-05,
-      "loss": 2.5013,
-      "step": 7900
-    },
-    {
-      "epoch": 0.36,
-      "learning_rate": 6.385377830317359e-05,
-      "loss": 2.5366,
-      "step": 7950
-    },
-    {
-      "epoch": 0.36,
-      "learning_rate": 6.36264435755206e-05,
-      "loss": 2.49,
-      "step": 8000
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 6.33991088478676e-05,
-      "loss": 2.5496,
-      "step": 8050
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 6.317177412021461e-05,
-      "loss": 2.5071,
-      "step": 8100
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 6.294443939256161e-05,
-      "loss": 2.5374,
-      "step": 8150
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 6.27171046649086e-05,
-      "loss": 2.5284,
-      "step": 8200
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 6.248976993725561e-05,
-      "loss": 2.4792,
-      "step": 8250
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 6.226243520960262e-05,
-      "loss": 2.5061,
-      "step": 8300
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 6.203510048194962e-05,
-      "loss": 2.5183,
-      "step": 8350
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 6.180776575429663e-05,
-      "loss": 2.4886,
-      "step": 8400
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 6.158043102664363e-05,
-      "loss": 2.5116,
-      "step": 8450
-    },
-    {
-      "epoch": 0.39,
-      "learning_rate": 6.135309629899064e-05,
-      "loss": 2.5364,
-      "step": 8500
-    },
-    {
-      "epoch": 0.39,
-      "learning_rate": 6.112576157133763e-05,
-      "loss": 2.5205,
-      "step": 8550
-    },
-    {
-      "epoch": 0.39,
-      "learning_rate": 6.0898426843684644e-05,
-      "loss": 2.5125,
-      "step": 8600
-    },
-    {
-      "epoch": 0.39,
-      "learning_rate": 6.067109211603165e-05,
-      "loss": 2.5089,
-      "step": 8650
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 6.0443757388378655e-05,
-      "loss": 2.5088,
-      "step": 8700
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 6.021642266072566e-05,
-      "loss": 2.5264,
-      "step": 8750
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 5.9989087933072666e-05,
-      "loss": 2.5045,
-      "step": 8800
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 5.976175320541967e-05,
-      "loss": 2.5085,
-      "step": 8850
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 5.9534418477766663e-05,
-      "loss": 2.4801,
-      "step": 8900
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 5.930708375011367e-05,
-      "loss": 2.5017,
-      "step": 8950
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 5.9079749022460675e-05,
-      "loss": 2.5109,
-      "step": 9000
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 5.885241429480768e-05,
-      "loss": 2.5052,
-      "step": 9050
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 5.8625079567154686e-05,
-      "loss": 2.5139,
-      "step": 9100
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 5.839774483950169e-05,
-      "loss": 2.4941,
-      "step": 9150
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 5.817041011184868e-05,
-      "loss": 2.5137,
-      "step": 9200
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 5.794307538419569e-05,
-      "loss": 2.5101,
-      "step": 9250
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 5.7715740656542694e-05,
-      "loss": 2.5009,
-      "step": 9300
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 5.74884059288897e-05,
-      "loss": 2.5395,
-      "step": 9350
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 5.7261071201236706e-05,
-      "loss": 2.5108,
-      "step": 9400
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 5.703373647358371e-05,
-      "loss": 2.5238,
-      "step": 9450
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 5.680640174593072e-05,
-      "loss": 2.5037,
-      "step": 9500
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 5.657906701827771e-05,
-      "loss": 2.5038,
-      "step": 9550
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 5.6351732290624714e-05,
-      "loss": 2.5324,
-      "step": 9600
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 5.612439756297172e-05,
-      "loss": 2.5054,
-      "step": 9650
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 5.5897062835318725e-05,
-      "loss": 2.5119,
-      "step": 9700
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 5.566972810766573e-05,
-      "loss": 2.5214,
-      "step": 9750
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 5.5442393380012737e-05,
-      "loss": 2.5404,
-      "step": 9800
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 5.521505865235974e-05,
-      "loss": 2.516,
-      "step": 9850
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 5.4987723924706734e-05,
-      "loss": 2.5166,
-      "step": 9900
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 5.476038919705374e-05,
-      "loss": 2.4983,
-      "step": 9950
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 5.4533054469400745e-05,
-      "loss": 2.5101,
-      "step": 10000
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 5.430571974174775e-05,
-      "loss": 2.4998,
-      "step": 10050
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 5.4078385014094756e-05,
-      "loss": 2.5116,
-      "step": 10100
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 5.385105028644176e-05,
-      "loss": 2.5136,
-      "step": 10150
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 5.362371555878877e-05,
-      "loss": 2.5313,
-      "step": 10200
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 5.339638083113576e-05,
-      "loss": 2.4989,
-      "step": 10250
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 5.3169046103482765e-05,
-      "loss": 2.5062,
-      "step": 10300
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 5.294171137582977e-05,
-      "loss": 2.531,
-      "step": 10350
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 5.2714376648176776e-05,
-      "loss": 2.4975,
-      "step": 10400
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 5.248704192052378e-05,
-      "loss": 2.4922,
-      "step": 10450
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 5.225970719287079e-05,
-      "loss": 2.5128,
-      "step": 10500
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 5.203237246521779e-05,
-      "loss": 2.504,
-      "step": 10550
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 5.1805037737564785e-05,
-      "loss": 2.5093,
-      "step": 10600
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 5.157770300991179e-05,
-      "loss": 2.491,
-      "step": 10650
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 5.1350368282258796e-05,
-      "loss": 2.5008,
-      "step": 10700
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 5.11230335546058e-05,
-      "loss": 2.5103,
-      "step": 10750
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 5.089569882695281e-05,
-      "loss": 2.5167,
-      "step": 10800
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 5.066836409929981e-05,
-      "loss": 2.5062,
-      "step": 10850
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 5.044102937164682e-05,
-      "loss": 2.5135,
-      "step": 10900
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 5.021369464399382e-05,
-      "loss": 2.489,
-      "step": 10950
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 4.998635991634082e-05,
-      "loss": 2.5071,
-      "step": 11000
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 4.975902518868782e-05,
-      "loss": 2.5181,
-      "step": 11050
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 4.953169046103483e-05,
-      "loss": 2.4997,
-      "step": 11100
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 4.930435573338183e-05,
-      "loss": 2.5127,
-      "step": 11150
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 4.907702100572884e-05,
-      "loss": 2.4906,
-      "step": 11200
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 4.884968627807584e-05,
-      "loss": 2.5129,
-      "step": 11250
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 4.862235155042284e-05,
-      "loss": 2.5015,
-      "step": 11300
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 4.839501682276985e-05,
-      "loss": 2.5049,
-      "step": 11350
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 4.8167682095116854e-05,
-      "loss": 2.4971,
-      "step": 11400
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 4.794034736746386e-05,
-      "loss": 2.5177,
-      "step": 11450
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 4.771301263981086e-05,
-      "loss": 2.5056,
-      "step": 11500
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 4.7485677912157864e-05,
-      "loss": 2.4831,
-      "step": 11550
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 4.725834318450487e-05,
-      "loss": 2.4972,
-      "step": 11600
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 4.7031008456851875e-05,
-      "loss": 2.5103,
-      "step": 11650
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 4.680367372919888e-05,
-      "loss": 2.5083,
-      "step": 11700
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 4.657633900154588e-05,
-      "loss": 2.5027,
-      "step": 11750
-    },
-    {
-      "epoch": 0.54,
-      "learning_rate": 4.6349004273892885e-05,
-      "loss": 2.4846,
-      "step": 11800
-    },
-    {
-      "epoch": 0.54,
-      "learning_rate": 4.612166954623989e-05,
-      "loss": 2.5193,
-      "step": 11850
-    },
-    {
-      "epoch": 0.54,
-      "learning_rate": 4.589433481858689e-05,
-      "loss": 2.5123,
-      "step": 11900
-    },
-    {
-      "epoch": 0.54,
-      "learning_rate": 4.5667000090933895e-05,
-      "loss": 2.5219,
-      "step": 11950
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 4.54396653632809e-05,
-      "loss": 2.4979,
-      "step": 12000
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 4.5212330635627906e-05,
-      "loss": 2.4849,
-      "step": 12050
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 4.4984995907974905e-05,
-      "loss": 2.4783,
-      "step": 12100
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 4.475766118032191e-05,
-      "loss": 2.5035,
-      "step": 12150
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 4.4530326452668916e-05,
-      "loss": 2.4879,
-      "step": 12200
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 4.4302991725015914e-05,
-      "loss": 2.4972,
-      "step": 12250
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 4.407565699736292e-05,
-      "loss": 2.5043,
-      "step": 12300
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 4.3848322269709926e-05,
-      "loss": 2.491,
-      "step": 12350
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 4.362098754205693e-05,
-      "loss": 2.5032,
-      "step": 12400
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 4.339365281440393e-05,
-      "loss": 2.5227,
-      "step": 12450
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 4.3166318086750935e-05,
-      "loss": 2.5245,
-      "step": 12500
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 4.293898335909794e-05,
-      "loss": 2.4927,
-      "step": 12550
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 4.271164863144494e-05,
-      "loss": 2.5002,
-      "step": 12600
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 4.2484313903791945e-05,
-      "loss": 2.4997,
-      "step": 12650
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 4.225697917613895e-05,
-      "loss": 2.4939,
-      "step": 12700
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 4.2029644448485957e-05,
-      "loss": 2.5223,
-      "step": 12750
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 4.1802309720832955e-05,
-      "loss": 2.4963,
-      "step": 12800
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 4.157497499317996e-05,
-      "loss": 2.5334,
-      "step": 12850
-    },
-    {
-      "epoch": 0.59,
-      "learning_rate": 4.1347640265526966e-05,
-      "loss": 2.5085,
-      "step": 12900
-    },
-    {
-      "epoch": 0.59,
-      "learning_rate": 4.1120305537873965e-05,
-      "loss": 2.4901,
-      "step": 12950
-    },
-    {
-      "epoch": 0.59,
-      "learning_rate": 4.089297081022097e-05,
-      "loss": 2.5268,
-      "step": 13000
-    },
-    {
-      "epoch": 0.59,
-      "learning_rate": 4.0665636082567976e-05,
-      "loss": 2.5237,
-      "step": 13050
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 4.043830135491498e-05,
-      "loss": 2.4928,
-      "step": 13100
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 4.021096662726198e-05,
-      "loss": 2.4852,
-      "step": 13150
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 3.9983631899608986e-05,
-      "loss": 2.5139,
-      "step": 13200
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 3.975629717195599e-05,
-      "loss": 2.5336,
-      "step": 13250
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 3.952896244430299e-05,
-      "loss": 2.527,
-      "step": 13300
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 3.9301627716649996e-05,
-      "loss": 2.4877,
-      "step": 13350
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 3.9074292988997e-05,
-      "loss": 2.4992,
-      "step": 13400
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 3.884695826134401e-05,
-      "loss": 2.4909,
-      "step": 13450
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 3.8619623533691006e-05,
-      "loss": 2.4983,
-      "step": 13500
-    },
-    {
-      "epoch": 0.62,
-      "learning_rate": 3.839228880603801e-05,
-      "loss": 2.5146,
-      "step": 13550
-    },
-    {
-      "epoch": 0.62,
-      "learning_rate": 3.816495407838502e-05,
-      "loss": 2.5058,
-      "step": 13600
-    },
-    {
-      "epoch": 0.62,
-      "learning_rate": 3.7937619350732016e-05,
-      "loss": 2.4943,
-      "step": 13650
-    },
-    {
-      "epoch": 0.62,
-      "learning_rate": 3.771028462307902e-05,
-      "loss": 2.5002,
-      "step": 13700
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 3.748294989542603e-05,
-      "loss": 2.4918,
-      "step": 13750
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 3.7255615167773026e-05,
-      "loss": 2.4915,
-      "step": 13800
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 3.702828044012003e-05,
-      "loss": 2.5089,
-      "step": 13850
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 3.680094571246704e-05,
-      "loss": 2.5048,
-      "step": 13900
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 3.657361098481404e-05,
-      "loss": 2.5108,
-      "step": 13950
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 3.634627625716104e-05,
-      "loss": 2.4959,
-      "step": 14000
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 3.611894152950805e-05,
-      "loss": 2.5154,
-      "step": 14050
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 3.589160680185505e-05,
-      "loss": 2.5092,
-      "step": 14100
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 3.566427207420205e-05,
-      "loss": 2.5265,
-      "step": 14150
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 3.543693734654906e-05,
-      "loss": 2.4678,
-      "step": 14200
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 3.520960261889606e-05,
-      "loss": 2.5236,
-      "step": 14250
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 3.498226789124307e-05,
-      "loss": 2.5156,
-      "step": 14300
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 3.475493316359007e-05,
-      "loss": 2.508,
-      "step": 14350
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 3.452759843593707e-05,
-      "loss": 2.4949,
-      "step": 14400
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 3.430026370828408e-05,
-      "loss": 2.4898,
-      "step": 14450
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 3.4072928980631084e-05,
-      "loss": 2.5006,
-      "step": 14500
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 3.384559425297808e-05,
-      "loss": 2.4878,
-      "step": 14550
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 3.361825952532509e-05,
-      "loss": 2.5073,
-      "step": 14600
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 3.3390924797672094e-05,
-      "loss": 2.5176,
-      "step": 14650
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 3.31635900700191e-05,
-      "loss": 2.5078,
-      "step": 14700
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 3.2936255342366105e-05,
-      "loss": 2.5101,
-      "step": 14750
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 3.2708920614713103e-05,
-      "loss": 2.5076,
-      "step": 14800
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 3.248158588706011e-05,
-      "loss": 2.4916,
-      "step": 14850
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 3.2254251159407115e-05,
-      "loss": 2.4919,
-      "step": 14900
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 3.202691643175412e-05,
-      "loss": 2.5042,
-      "step": 14950
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 3.1799581704101126e-05,
-      "loss": 2.5191,
-      "step": 15000
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 3.1572246976448124e-05,
-      "loss": 2.5034,
-      "step": 15050
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 3.134491224879513e-05,
-      "loss": 2.4878,
-      "step": 15100
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 3.1117577521142136e-05,
-      "loss": 2.5072,
-      "step": 15150
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 3.0890242793489134e-05,
-      "loss": 2.506,
-      "step": 15200
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 3.066290806583614e-05,
-      "loss": 2.4885,
-      "step": 15250
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 3.0435573338183142e-05,
-      "loss": 2.488,
-      "step": 15300
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 3.0208238610530148e-05,
-      "loss": 2.4939,
-      "step": 15350
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 2.998090388287715e-05,
-      "loss": 2.5397,
-      "step": 15400
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 2.9753569155224152e-05,
-      "loss": 2.5131,
-      "step": 15450
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 2.9526234427571158e-05,
-      "loss": 2.5287,
-      "step": 15500
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 2.929889969991816e-05,
-      "loss": 2.4852,
-      "step": 15550
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 2.9071564972265165e-05,
-      "loss": 2.4941,
-      "step": 15600
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 2.884423024461217e-05,
-      "loss": 2.508,
-      "step": 15650
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 2.8616895516959173e-05,
-      "loss": 2.5011,
-      "step": 15700
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 2.8389560789306175e-05,
-      "loss": 2.5029,
-      "step": 15750
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 2.816222606165318e-05,
-      "loss": 2.4956,
-      "step": 15800
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 2.7934891334000186e-05,
-      "loss": 2.4998,
-      "step": 15850
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 2.7707556606347185e-05,
-      "loss": 2.4954,
-      "step": 15900
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 2.748022187869419e-05,
-      "loss": 2.5171,
-      "step": 15950
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 2.7252887151041196e-05,
-      "loss": 2.476,
-      "step": 16000
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 2.7025552423388202e-05,
-      "loss": 2.506,
-      "step": 16050
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 2.67982176957352e-05,
-      "loss": 2.5201,
-      "step": 16100
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 2.6570882968082206e-05,
-      "loss": 2.5205,
-      "step": 16150
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 2.6343548240429212e-05,
-      "loss": 2.4971,
-      "step": 16200
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 2.611621351277621e-05,
-      "loss": 2.5135,
-      "step": 16250
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 2.5888878785123216e-05,
-      "loss": 2.4894,
-      "step": 16300
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 2.5661544057470222e-05,
-      "loss": 2.5127,
-      "step": 16350
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 2.5434209329817227e-05,
-      "loss": 2.4999,
-      "step": 16400
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 2.5206874602164226e-05,
-      "loss": 2.5048,
-      "step": 16450
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 2.4979539874511232e-05,
-      "loss": 2.5208,
-      "step": 16500
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 2.4752205146858234e-05,
-      "loss": 2.5155,
-      "step": 16550
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 2.452487041920524e-05,
-      "loss": 2.5196,
-      "step": 16600
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 2.429753569155224e-05,
-      "loss": 2.5205,
-      "step": 16650
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 2.4070200963899247e-05,
-      "loss": 2.5083,
-      "step": 16700
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 2.384286623624625e-05,
-      "loss": 2.506,
-      "step": 16750
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 2.361553150859325e-05,
-      "loss": 2.5251,
-      "step": 16800
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 2.3388196780940257e-05,
-      "loss": 2.5124,
-      "step": 16850
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 2.316086205328726e-05,
-      "loss": 2.4869,
-      "step": 16900
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 2.2933527325634265e-05,
-      "loss": 2.5066,
-      "step": 16950
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 2.2706192597981267e-05,
-      "loss": 2.4888,
-      "step": 17000
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 2.2478857870328273e-05,
-      "loss": 2.5086,
-      "step": 17050
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 2.2251523142675275e-05,
-      "loss": 2.5449,
-      "step": 17100
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 2.202418841502228e-05,
-      "loss": 2.5186,
-      "step": 17150
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 2.1796853687369283e-05,
-      "loss": 2.482,
-      "step": 17200
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 2.1569518959716288e-05,
-      "loss": 2.4895,
-      "step": 17250
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 2.1342184232063294e-05,
-      "loss": 2.4988,
-      "step": 17300
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 2.1114849504410296e-05,
-      "loss": 2.5107,
-      "step": 17350
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 2.0887514776757298e-05,
-      "loss": 2.5257,
-      "step": 17400
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 2.0660180049104304e-05,
-      "loss": 2.5139,
-      "step": 17450
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 2.0432845321451306e-05,
-      "loss": 2.5364,
-      "step": 17500
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 2.020551059379831e-05,
-      "loss": 2.5242,
-      "step": 17550
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 1.9978175866145313e-05,
-      "loss": 2.482,
-      "step": 17600
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 1.975084113849232e-05,
-      "loss": 2.4981,
-      "step": 17650
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 1.952350641083932e-05,
-      "loss": 2.5049,
-      "step": 17700
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 1.9296171683186323e-05,
-      "loss": 2.5089,
-      "step": 17750
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 1.906883695553333e-05,
-      "loss": 2.4937,
-      "step": 17800
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 1.884150222788033e-05,
-      "loss": 2.4983,
-      "step": 17850
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 1.8614167500227337e-05,
-      "loss": 2.5152,
-      "step": 17900
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 1.838683277257434e-05,
-      "loss": 2.5198,
-      "step": 17950
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 1.8159498044921344e-05,
-      "loss": 2.5108,
-      "step": 18000
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 1.7932163317268347e-05,
-      "loss": 2.5362,
-      "step": 18050
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 1.770482858961535e-05,
-      "loss": 2.5186,
-      "step": 18100
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 1.7477493861962354e-05,
-      "loss": 2.5331,
-      "step": 18150
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 1.7250159134309357e-05,
-      "loss": 2.506,
-      "step": 18200
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 1.7022824406656362e-05,
-      "loss": 2.4932,
-      "step": 18250
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 1.6795489679003364e-05,
-      "loss": 2.4975,
-      "step": 18300
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 1.656815495135037e-05,
-      "loss": 2.4996,
-      "step": 18350
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 1.6340820223697372e-05,
-      "loss": 2.4987,
-      "step": 18400
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 1.6113485496044374e-05,
-      "loss": 2.5013,
-      "step": 18450
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 1.588615076839138e-05,
-      "loss": 2.4971,
-      "step": 18500
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 1.5658816040738382e-05,
-      "loss": 2.5349,
-      "step": 18550
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 1.5431481313085388e-05,
-      "loss": 2.5176,
-      "step": 18600
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 1.5204146585432391e-05,
-      "loss": 2.4829,
-      "step": 18650
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 1.4976811857779397e-05,
-      "loss": 2.5258,
-      "step": 18700
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 1.47494771301264e-05,
-      "loss": 2.5232,
-      "step": 18750
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 1.4522142402473401e-05,
-      "loss": 2.5032,
-      "step": 18800
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 1.4294807674820407e-05,
-      "loss": 2.5197,
-      "step": 18850
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 1.4067472947167409e-05,
-      "loss": 2.5035,
-      "step": 18900
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 1.3840138219514415e-05,
-      "loss": 2.5016,
-      "step": 18950
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 1.3612803491861417e-05,
-      "loss": 2.5291,
-      "step": 19000
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 1.3385468764208419e-05,
-      "loss": 2.5092,
-      "step": 19050
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 1.3158134036555425e-05,
-      "loss": 2.4956,
-      "step": 19100
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 1.2930799308902428e-05,
-      "loss": 2.4947,
-      "step": 19150
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 1.2703464581249432e-05,
-      "loss": 2.519,
-      "step": 19200
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 1.2476129853596436e-05,
-      "loss": 2.5452,
-      "step": 19250
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 1.224879512594344e-05,
-      "loss": 2.5248,
-      "step": 19300
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 1.2021460398290444e-05,
-      "loss": 2.5179,
-      "step": 19350
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 1.1794125670637448e-05,
-      "loss": 2.5047,
-      "step": 19400
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 1.1566790942984452e-05,
-      "loss": 2.5084,
-      "step": 19450
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 1.1339456215331456e-05,
-      "loss": 2.5044,
-      "step": 19500
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 1.1112121487678458e-05,
-      "loss": 2.4969,
-      "step": 19550
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 1.0884786760025462e-05,
-      "loss": 2.5069,
-      "step": 19600
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 1.0657452032372465e-05,
-      "loss": 2.4756,
-      "step": 19650
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 1.043011730471947e-05,
-      "loss": 2.489,
-      "step": 19700
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 1.0202782577066473e-05,
-      "loss": 2.4985,
-      "step": 19750
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 9.975447849413477e-06,
-      "loss": 2.5375,
-      "step": 19800
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 9.748113121760481e-06,
-      "loss": 2.4924,
-      "step": 19850
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 9.520778394107483e-06,
-      "loss": 2.4879,
-      "step": 19900
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 9.293443666454487e-06,
-      "loss": 2.5486,
-      "step": 19950
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 9.066108938801491e-06,
-      "loss": 2.5187,
-      "step": 20000
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 8.838774211148495e-06,
-      "loss": 2.5102,
-      "step": 20050
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 8.6114394834955e-06,
-      "loss": 2.4975,
-      "step": 20100
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 8.384104755842504e-06,
-      "loss": 2.5036,
-      "step": 20150
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 8.156770028189506e-06,
-      "loss": 2.5323,
-      "step": 20200
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 7.92943530053651e-06,
-      "loss": 2.5228,
-      "step": 20250
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 7.702100572883514e-06,
-      "loss": 2.5411,
-      "step": 20300
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 7.474765845230518e-06,
-      "loss": 2.4923,
-      "step": 20350
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 7.247431117577522e-06,
-      "loss": 2.5178,
-      "step": 20400
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 7.020096389924526e-06,
-      "loss": 2.5121,
-      "step": 20450
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 6.79276166227153e-06,
-      "loss": 2.5414,
-      "step": 20500
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 6.565426934618532e-06,
-      "loss": 2.5222,
-      "step": 20550
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 6.338092206965536e-06,
-      "loss": 2.5261,
-      "step": 20600
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 6.11075747931254e-06,
-      "loss": 2.5274,
-      "step": 20650
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 5.883422751659544e-06,
-      "loss": 2.4965,
-      "step": 20700
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 5.656088024006548e-06,
-      "loss": 2.5141,
-      "step": 20750
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 5.428753296353551e-06,
-      "loss": 2.5101,
-      "step": 20800
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 5.201418568700555e-06,
-      "loss": 2.5011,
-      "step": 20850
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 4.974083841047559e-06,
-      "loss": 2.5091,
-      "step": 20900
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 4.746749113394562e-06,
-      "loss": 2.5237,
-      "step": 20950
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 4.519414385741566e-06,
-      "loss": 2.4949,
-      "step": 21000
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 4.29207965808857e-06,
-      "loss": 2.503,
-      "step": 21050
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 4.0647449304355735e-06,
-      "loss": 2.5068,
-      "step": 21100
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 3.837410202782577e-06,
-      "loss": 2.4922,
-      "step": 21150
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 3.6100754751295813e-06,
-      "loss": 2.5199,
-      "step": 21200
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 3.382740747476585e-06,
-      "loss": 2.5058,
-      "step": 21250
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 3.155406019823588e-06,
-      "loss": 2.5294,
-      "step": 21300
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 2.928071292170592e-06,
-      "loss": 2.4969,
-      "step": 21350
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 2.700736564517596e-06,
-      "loss": 2.5419,
-      "step": 21400
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 2.4734018368645998e-06,
-      "loss": 2.5299,
-      "step": 21450
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 2.2460671092116032e-06,
-      "loss": 2.5275,
-      "step": 21500
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 2.0187323815586067e-06,
-      "loss": 2.4891,
-      "step": 21550
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 1.7913976539056108e-06,
-      "loss": 2.5108,
-      "step": 21600
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 1.5640629262526144e-06,
-      "loss": 2.5246,
-      "step": 21650
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 1.336728198599618e-06,
-      "loss": 2.5304,
-      "step": 21700
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 1.109393470946622e-06,
-      "loss": 2.5159,
-      "step": 21750
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 8.820587432936256e-07,
-      "loss": 2.5071,
-      "step": 21800
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 6.547240156406293e-07,
-      "loss": 2.5091,
-      "step": 21850
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 4.2738928798763303e-07,
-      "loss": 2.5386,
-      "step": 21900
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 2.0005456033463672e-07,
-      "loss": 2.5228,
-      "step": 21950
-    },
     {
       "epoch": 1.0,
-      "step": 21994,
-      "total_flos": 3.604860407937761e+17,
-      "train_loss": 2.5232514588412474,
-      "train_runtime": 5052.6511,
-      "train_samples_per_second": 34.825,
-      "train_steps_per_second": 4.353
+      "step": 3726,
+      "total_flos": 2.4427952859827405e+17,
+      "train_loss": 2.7717256223521947,
+      "train_runtime": 3102.8464,
+      "train_samples_per_second": 38.427,
+      "train_steps_per_second": 1.201
     }
   ],
   "logging_steps": 50,
-  "max_steps": 21994,
+  "max_steps": 3726,
   "num_train_epochs": 1,
-  "save_steps": 2500,
-  "total_flos": 3.604860407937761e+17,
+  "save_steps": 5000,
+  "total_flos": 2.4427952859827405e+17,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2458e0243aaab0a29fd49a56d9466fcdd0e8ef0f37d199202b23e025f597ffca
+oid sha256:ce83680fb3cb44202bb32f2eca1b7e1d9dd35259d3cb4ec2c75bd5a87a2385d3
 size 4027