Upload folder using huggingface_hub

README.md

@@ -0,0 +1,57 @@
+---
+tags:
+- generated_from_trainer
+datasets:
+- /pfs/lustrep4/scratch/project_462000259/shared_datasets/modified_200/modified_200/
+model-index:
+- name: layer_16,17,18,19,20
+  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. -->
+
+# layer_16,17,18,19,20
+
+This model is a fine-tuned version of [/pfs/lustrep4/scratch/project_462000259/shared_models/pythia-2.8b-deduped-base/pythia-2.8b-deduped](https://huggingface.co//pfs/lustrep4/scratch/project_462000259/shared_models/pythia-2.8b-deduped-base/pythia-2.8b-deduped) on the /pfs/lustrep4/scratch/project_462000259/shared_datasets/modified_200/modified_200/ 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:
+- learning_rate: 5e-05
+- train_batch_size: 2
+- eval_batch_size: 2
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 64
+- total_train_batch_size: 128
+- total_eval_batch_size: 128
+- optimizer: Adam with betas=(0.9,0.99) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 4
+- mixed_precision_training: Native AMP
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.28.1
+- Pytorch 2.0.1+rocm5.4.2
+- Datasets 2.11.0
+- Tokenizers 0.13.3

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.11551530145209522,
+    "train_runtime": 3653.4642,
+    "train_samples": 463392,
+    "train_samples_per_second": 507.345,
+    "train_steps_per_second": 3.964
+}

config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "/pfs/lustrep4/scratch/project_462000259/shared_models/pythia-2.8b-deduped-base/pythia-2.8b-deduped",
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "hidden_act": "gelu",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 10240,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 0.25,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.1",
+  "use_cache": true,
+  "use_parallel_residual": true,
+  "vocab_size": 50304
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "transformers_version": "4.28.1"
+}

pytorch_model-00001-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7434d7b62c40a53d2ae799370e967ca5c1d593ffc6e6c3f8d9bddc53ecd9cc4
+size 10086521429

pytorch_model-00002-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b74dbf36afa36107d077372846e6ba5c6e1743414d0ca5616821d2f5ad29120c
+size 1148707683

pytorch_model.bin.index.json

@@ -0,0 +1,491 @@
+{
+  "metadata": {
+    "total_size": 11117614464.0
+  },
+  "weight_map": {
+    "embed_out.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.embed_in.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.final_layer_norm.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.final_layer_norm.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.0.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.15.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.16.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.17.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.18.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.19.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.20.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.21.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.22.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.23.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.24.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.25.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.26.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.27.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.28.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.29.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.30.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.30.attention.dense.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.attention.dense.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.30.attention.query_key_value.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.attention.query_key_value.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.30.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.30.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.30.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.30.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.30.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.31.attention.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.attention.dense.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.attention.dense.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.attention.masked_bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.attention.query_key_value.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.attention.query_key_value.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.attention.rotary_emb.inv_freq": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.input_layernorm.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.mlp.dense_4h_to_h.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.mlp.dense_4h_to_h.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.mlp.dense_h_to_4h.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.mlp.dense_h_to_4h.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.post_attention_layernorm.bias": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.31.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
+    "gpt_neox.layers.4.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.attention.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.attention.dense.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.attention.dense.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.attention.masked_bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.attention.query_key_value.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.attention.query_key_value.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.attention.rotary_emb.inv_freq": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.input_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.mlp.dense_4h_to_h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.mlp.dense_4h_to_h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.mlp.dense_h_to_4h.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.mlp.dense_h_to_4h.weight": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.post_attention_layernorm.bias": "pytorch_model-00001-of-00002.bin",
+    "gpt_neox.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "unk_token": "<|endoftext|>"
+}

tokenizer_config.json

@@ -0,0 +1,9 @@
+{
+  "add_prefix_space": false,
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 1000000000000000019884624838656,
+  "tokenizer_class": "GPTNeoXTokenizer",
+  "unk_token": "<|endoftext|>"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.11551530145209522,
+    "train_runtime": 3653.4642,
+    "train_samples": 463392,
+    "train_samples_per_second": 507.345,
+    "train_steps_per_second": 3.964
+}

trainer_state.json

@@ -0,0 +1,4441 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "global_step": 14484,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.993093922651934e-05,
+      "loss": 1.6286,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.986187845303867e-05,
+      "loss": 1.572,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9792817679558015e-05,
+      "loss": 1.5466,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.972375690607735e-05,
+      "loss": 1.5356,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.965469613259669e-05,
+      "loss": 1.5218,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.958563535911602e-05,
+      "loss": 1.509,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.951657458563536e-05,
+      "loss": 1.5031,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.94475138121547e-05,
+      "loss": 1.4833,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.937845303867403e-05,
+      "loss": 1.4731,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.930939226519337e-05,
+      "loss": 1.4706,
+      "step": 200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.924033149171271e-05,
+      "loss": 1.4689,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9171270718232046e-05,
+      "loss": 1.4577,
+      "step": 240
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.910220994475138e-05,
+      "loss": 1.4482,
+      "step": 260
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.903314917127072e-05,
+      "loss": 1.4498,
+      "step": 280
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.896408839779006e-05,
+      "loss": 1.4366,
+      "step": 300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8895027624309394e-05,
+      "loss": 1.4364,
+      "step": 320
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.882596685082873e-05,
+      "loss": 1.4343,
+      "step": 340
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.875690607734807e-05,
+      "loss": 1.4291,
+      "step": 360
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8687845303867406e-05,
+      "loss": 1.4243,
+      "step": 380
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.861878453038674e-05,
+      "loss": 1.4105,
+      "step": 400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.854972375690608e-05,
+      "loss": 1.4164,
+      "step": 420
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.848066298342542e-05,
+      "loss": 1.4194,
+      "step": 440
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8411602209944754e-05,
+      "loss": 1.409,
+      "step": 460
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.834254143646409e-05,
+      "loss": 1.4026,
+      "step": 480
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8273480662983425e-05,
+      "loss": 1.402,
+      "step": 500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.820441988950277e-05,
+      "loss": 1.389,
+      "step": 520
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.81353591160221e-05,
+      "loss": 1.3875,
+      "step": 540
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.806629834254144e-05,
+      "loss": 1.3841,
+      "step": 560
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.799723756906078e-05,
+      "loss": 1.3848,
+      "step": 580
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7928176795580114e-05,
+      "loss": 1.3783,
+      "step": 600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.785911602209945e-05,
+      "loss": 1.3799,
+      "step": 620
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7790055248618785e-05,
+      "loss": 1.3861,
+      "step": 640
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.772099447513813e-05,
+      "loss": 1.3769,
+      "step": 660
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7651933701657456e-05,
+      "loss": 1.3692,
+      "step": 680
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.75828729281768e-05,
+      "loss": 1.3654,
+      "step": 700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.751381215469613e-05,
+      "loss": 1.3571,
+      "step": 720
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.7444751381215475e-05,
+      "loss": 1.3712,
+      "step": 740
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.737569060773481e-05,
+      "loss": 1.366,
+      "step": 760
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.7306629834254145e-05,
+      "loss": 1.364,
+      "step": 780
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.723756906077349e-05,
+      "loss": 1.3624,
+      "step": 800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.7168508287292816e-05,
+      "loss": 1.3595,
+      "step": 820
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.709944751381216e-05,
+      "loss": 1.353,
+      "step": 840
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.703038674033149e-05,
+      "loss": 1.3629,
+      "step": 860
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6961325966850835e-05,
+      "loss": 1.3558,
+      "step": 880
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6892265193370164e-05,
+      "loss": 1.3502,
+      "step": 900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6823204419889506e-05,
+      "loss": 1.3427,
+      "step": 920
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.675414364640884e-05,
+      "loss": 1.3498,
+      "step": 940
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.6685082872928176e-05,
+      "loss": 1.349,
+      "step": 960
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.661602209944751e-05,
+      "loss": 1.3479,
+      "step": 980
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.6546961325966854e-05,
+      "loss": 1.3461,
+      "step": 1000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.647790055248619e-05,
+      "loss": 1.3465,
+      "step": 1020
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.6408839779005524e-05,
+      "loss": 1.3351,
+      "step": 1040
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.6339779005524866e-05,
+      "loss": 1.3473,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.62707182320442e-05,
+      "loss": 1.3362,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.6201657458563544e-05,
+      "loss": 1.3402,
+      "step": 1100
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.613259668508287e-05,
+      "loss": 1.3186,
+      "step": 1120
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.6063535911602214e-05,
+      "loss": 1.3262,
+      "step": 1140
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.599447513812155e-05,
+      "loss": 1.3295,
+      "step": 1160
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.5925414364640885e-05,
+      "loss": 1.3291,
+      "step": 1180
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.585635359116022e-05,
+      "loss": 1.3247,
+      "step": 1200
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.578729281767956e-05,
+      "loss": 1.322,
+      "step": 1220
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.57182320441989e-05,
+      "loss": 1.3239,
+      "step": 1240
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.564917127071823e-05,
+      "loss": 1.3207,
+      "step": 1260
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.5580110497237574e-05,
+      "loss": 1.3191,
+      "step": 1280
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.551104972375691e-05,
+      "loss": 1.3187,
+      "step": 1300
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.5441988950276245e-05,
+      "loss": 1.3204,
+      "step": 1320
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.537292817679558e-05,
+      "loss": 1.3124,
+      "step": 1340
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.530386740331492e-05,
+      "loss": 1.3156,
+      "step": 1360
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.523480662983426e-05,
+      "loss": 1.317,
+      "step": 1380
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.516574585635359e-05,
+      "loss": 1.31,
+      "step": 1400
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.509668508287293e-05,
+      "loss": 1.3061,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.502762430939227e-05,
+      "loss": 1.3165,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.4958563535911605e-05,
+      "loss": 1.311,
+      "step": 1460
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.488950276243094e-05,
+      "loss": 1.3079,
+      "step": 1480
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.4820441988950276e-05,
+      "loss": 1.3107,
+      "step": 1500
+    },
+    {
+      "epoch": 0.41,
+      "eval_validation_accuracy": 0.05525197850512946,
+      "eval_validation_loss": 1.3119829893112183,
+      "eval_validation_runtime": 3780.1869,
+      "eval_validation_samples_per_second": 0.661,
+      "eval_validation_steps_per_second": 0.021,
+      "step": 1500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.475138121546962e-05,
+      "loss": 1.3129,
+      "step": 1520
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.468232044198895e-05,
+      "loss": 1.3044,
+      "step": 1540
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.461325966850829e-05,
+      "loss": 1.2942,
+      "step": 1560
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.454419889502763e-05,
+      "loss": 1.3053,
+      "step": 1580
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.447513812154696e-05,
+      "loss": 1.2971,
+      "step": 1600
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.44060773480663e-05,
+      "loss": 1.3038,
+      "step": 1620
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4337016574585636e-05,
+      "loss": 1.2996,
+      "step": 1640
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.426795580110498e-05,
+      "loss": 1.3082,
+      "step": 1660
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.419889502762431e-05,
+      "loss": 1.3103,
+      "step": 1680
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.412983425414365e-05,
+      "loss": 1.2928,
+      "step": 1700
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4060773480662984e-05,
+      "loss": 1.2934,
+      "step": 1720
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.399171270718232e-05,
+      "loss": 1.2965,
+      "step": 1740
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.392265193370166e-05,
+      "loss": 1.2982,
+      "step": 1760
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3853591160221e-05,
+      "loss": 1.3003,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.378453038674034e-05,
+      "loss": 1.2953,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.371546961325967e-05,
+      "loss": 1.2859,
+      "step": 1820
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.364640883977901e-05,
+      "loss": 1.2955,
+      "step": 1840
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3577348066298344e-05,
+      "loss": 1.2882,
+      "step": 1860
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3508287292817687e-05,
+      "loss": 1.2887,
+      "step": 1880
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3439226519337015e-05,
+      "loss": 1.2954,
+      "step": 1900
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.337016574585636e-05,
+      "loss": 1.288,
+      "step": 1920
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.330110497237569e-05,
+      "loss": 1.2837,
+      "step": 1940
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.323204419889503e-05,
+      "loss": 1.292,
+      "step": 1960
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.316298342541436e-05,
+      "loss": 1.2881,
+      "step": 1980
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.3093922651933705e-05,
+      "loss": 1.28,
+      "step": 2000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.302486187845304e-05,
+      "loss": 1.281,
+      "step": 2020
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2955801104972375e-05,
+      "loss": 1.2828,
+      "step": 2040
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.288674033149172e-05,
+      "loss": 1.2713,
+      "step": 2060
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.281767955801105e-05,
+      "loss": 1.2842,
+      "step": 2080
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.274861878453039e-05,
+      "loss": 1.2824,
+      "step": 2100
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.267955801104972e-05,
+      "loss": 1.2718,
+      "step": 2120
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.2610497237569065e-05,
+      "loss": 1.2842,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.25414364640884e-05,
+      "loss": 1.2747,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.2472375690607736e-05,
+      "loss": 1.2799,
+      "step": 2180
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.240331491712707e-05,
+      "loss": 1.2883,
+      "step": 2200
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.233425414364641e-05,
+      "loss": 1.2746,
+      "step": 2220
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.226519337016575e-05,
+      "loss": 1.2774,
+      "step": 2240
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.2196132596685084e-05,
+      "loss": 1.287,
+      "step": 2260
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2127071823204426e-05,
+      "loss": 1.2745,
+      "step": 2280
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.205801104972376e-05,
+      "loss": 1.2845,
+      "step": 2300
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.1988950276243096e-05,
+      "loss": 1.2748,
+      "step": 2320
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.191988950276243e-05,
+      "loss": 1.2711,
+      "step": 2340
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.1850828729281773e-05,
+      "loss": 1.2691,
+      "step": 2360
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.17817679558011e-05,
+      "loss": 1.268,
+      "step": 2380
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.1712707182320444e-05,
+      "loss": 1.2763,
+      "step": 2400
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.164364640883978e-05,
+      "loss": 1.2769,
+      "step": 2420
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.157458563535912e-05,
+      "loss": 1.2665,
+      "step": 2440
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.150552486187845e-05,
+      "loss": 1.2725,
+      "step": 2460
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.143646408839779e-05,
+      "loss": 1.2749,
+      "step": 2480
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.136740331491713e-05,
+      "loss": 1.2704,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.129834254143646e-05,
+      "loss": 1.2669,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.1229281767955804e-05,
+      "loss": 1.2752,
+      "step": 2540
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.116022099447514e-05,
+      "loss": 1.2743,
+      "step": 2560
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.109116022099448e-05,
+      "loss": 1.2689,
+      "step": 2580
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.102209944751381e-05,
+      "loss": 1.2614,
+      "step": 2600
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.095303867403315e-05,
+      "loss": 1.2592,
+      "step": 2620
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.088397790055249e-05,
+      "loss": 1.2824,
+      "step": 2640
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.081491712707183e-05,
+      "loss": 1.2602,
+      "step": 2660
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.074585635359116e-05,
+      "loss": 1.2661,
+      "step": 2680
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.06767955801105e-05,
+      "loss": 1.2769,
+      "step": 2700
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.0607734806629835e-05,
+      "loss": 1.267,
+      "step": 2720
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.053867403314917e-05,
+      "loss": 1.2535,
+      "step": 2740
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.046961325966851e-05,
+      "loss": 1.2607,
+      "step": 2760
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.040055248618785e-05,
+      "loss": 1.2744,
+      "step": 2780
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.033149171270719e-05,
+      "loss": 1.2683,
+      "step": 2800
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.026243093922652e-05,
+      "loss": 1.2698,
+      "step": 2820
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.019337016574586e-05,
+      "loss": 1.265,
+      "step": 2840
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.0124309392265196e-05,
+      "loss": 1.2629,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.005524861878453e-05,
+      "loss": 1.263,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9986187845303866e-05,
+      "loss": 1.2647,
+      "step": 2900
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.991712707182321e-05,
+      "loss": 1.2611,
+      "step": 2920
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.9848066298342544e-05,
+      "loss": 1.2563,
+      "step": 2940
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.977900552486188e-05,
+      "loss": 1.2645,
+      "step": 2960
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.9709944751381214e-05,
+      "loss": 1.2573,
+      "step": 2980
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.9640883977900556e-05,
+      "loss": 1.2556,
+      "step": 3000
+    },
+    {
+      "epoch": 0.83,
+      "eval_validation_accuracy": 0.055986907669760624,
+      "eval_validation_loss": 1.253327488899231,
+      "eval_validation_runtime": 3780.1229,
+      "eval_validation_samples_per_second": 0.661,
+      "eval_validation_steps_per_second": 0.021,
+      "step": 3000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.957182320441989e-05,
+      "loss": 1.2525,
+      "step": 3020
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.950276243093923e-05,
+      "loss": 1.2495,
+      "step": 3040
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.943370165745857e-05,
+      "loss": 1.2644,
+      "step": 3060
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.9364640883977904e-05,
+      "loss": 1.2484,
+      "step": 3080
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.929558011049724e-05,
+      "loss": 1.2495,
+      "step": 3100
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.9226519337016574e-05,
+      "loss": 1.2556,
+      "step": 3120
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.9157458563535917e-05,
+      "loss": 1.245,
+      "step": 3140
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.9088397790055245e-05,
+      "loss": 1.2492,
+      "step": 3160
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.901933701657459e-05,
+      "loss": 1.2577,
+      "step": 3180
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.895027624309392e-05,
+      "loss": 1.2465,
+      "step": 3200
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.8881215469613264e-05,
+      "loss": 1.2557,
+      "step": 3220
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.88121546961326e-05,
+      "loss": 1.2489,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.8743093922651935e-05,
+      "loss": 1.2469,
+      "step": 3260
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.867403314917128e-05,
+      "loss": 1.2528,
+      "step": 3280
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.860497237569061e-05,
+      "loss": 1.2473,
+      "step": 3300
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.853591160220995e-05,
+      "loss": 1.2441,
+      "step": 3320
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.846685082872928e-05,
+      "loss": 1.2528,
+      "step": 3340
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.8397790055248625e-05,
+      "loss": 1.2469,
+      "step": 3360
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.832872928176795e-05,
+      "loss": 1.2475,
+      "step": 3380
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.8259668508287295e-05,
+      "loss": 1.2375,
+      "step": 3400
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.819060773480663e-05,
+      "loss": 1.2488,
+      "step": 3420
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.812154696132597e-05,
+      "loss": 1.2446,
+      "step": 3440
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.80524861878453e-05,
+      "loss": 1.2467,
+      "step": 3460
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.798342541436464e-05,
+      "loss": 1.2435,
+      "step": 3480
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.791436464088398e-05,
+      "loss": 1.2501,
+      "step": 3500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.7845303867403314e-05,
+      "loss": 1.242,
+      "step": 3520
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.7776243093922656e-05,
+      "loss": 1.2479,
+      "step": 3540
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.770718232044199e-05,
+      "loss": 1.2417,
+      "step": 3560
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.763812154696133e-05,
+      "loss": 1.242,
+      "step": 3580
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.756906077348066e-05,
+      "loss": 1.2353,
+      "step": 3600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.2334,
+      "step": 3620
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.743093922651934e-05,
+      "loss": 1.2161,
+      "step": 3640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7361878453038674e-05,
+      "loss": 1.2262,
+      "step": 3660
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.729281767955801e-05,
+      "loss": 1.2179,
+      "step": 3680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.722375690607735e-05,
+      "loss": 1.2216,
+      "step": 3700
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.7154696132596687e-05,
+      "loss": 1.2183,
+      "step": 3720
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.708563535911602e-05,
+      "loss": 1.217,
+      "step": 3740
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.7016574585635364e-05,
+      "loss": 1.224,
+      "step": 3760
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.69475138121547e-05,
+      "loss": 1.2182,
+      "step": 3780
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.6878453038674034e-05,
+      "loss": 1.2123,
+      "step": 3800
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.680939226519337e-05,
+      "loss": 1.2197,
+      "step": 3820
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.674033149171271e-05,
+      "loss": 1.2207,
+      "step": 3840
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.667127071823205e-05,
+      "loss": 1.2152,
+      "step": 3860
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.660220994475138e-05,
+      "loss": 1.2238,
+      "step": 3880
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.653314917127072e-05,
+      "loss": 1.2162,
+      "step": 3900
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.646408839779006e-05,
+      "loss": 1.2287,
+      "step": 3920
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.639502762430939e-05,
+      "loss": 1.2152,
+      "step": 3940
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.632596685082873e-05,
+      "loss": 1.211,
+      "step": 3960
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.6256906077348065e-05,
+      "loss": 1.2093,
+      "step": 3980
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.618784530386741e-05,
+      "loss": 1.2094,
+      "step": 4000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.611878453038674e-05,
+      "loss": 1.217,
+      "step": 4020
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.604972375690608e-05,
+      "loss": 1.2176,
+      "step": 4040
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.598066298342542e-05,
+      "loss": 1.2196,
+      "step": 4060
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.5911602209944755e-05,
+      "loss": 1.2137,
+      "step": 4080
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.584254143646409e-05,
+      "loss": 1.2122,
+      "step": 4100
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.5773480662983426e-05,
+      "loss": 1.2145,
+      "step": 4120
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.570441988950277e-05,
+      "loss": 1.2162,
+      "step": 4140
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5635359116022096e-05,
+      "loss": 1.2118,
+      "step": 4160
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.556629834254144e-05,
+      "loss": 1.2215,
+      "step": 4180
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.5497237569060774e-05,
+      "loss": 1.208,
+      "step": 4200
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.5428176795580116e-05,
+      "loss": 1.2134,
+      "step": 4220
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.535911602209945e-05,
+      "loss": 1.2083,
+      "step": 4240
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.5290055248618786e-05,
+      "loss": 1.2196,
+      "step": 4260
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.522099447513813e-05,
+      "loss": 1.2215,
+      "step": 4280
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.515193370165746e-05,
+      "loss": 1.2124,
+      "step": 4300
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.50828729281768e-05,
+      "loss": 1.2111,
+      "step": 4320
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.5013812154696134e-05,
+      "loss": 1.2192,
+      "step": 4340
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.4944751381215476e-05,
+      "loss": 1.2142,
+      "step": 4360
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.4875690607734804e-05,
+      "loss": 1.2157,
+      "step": 4380
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.4806629834254147e-05,
+      "loss": 1.2193,
+      "step": 4400
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.473756906077348e-05,
+      "loss": 1.2172,
+      "step": 4420
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.466850828729282e-05,
+      "loss": 1.2117,
+      "step": 4440
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.459944751381215e-05,
+      "loss": 1.208,
+      "step": 4460
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.4530386740331494e-05,
+      "loss": 1.2115,
+      "step": 4480
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.446132596685083e-05,
+      "loss": 1.2119,
+      "step": 4500
+    },
+    {
+      "epoch": 1.24,
+      "eval_validation_accuracy": 0.05641172447484123,
+      "eval_validation_loss": 1.2233214378356934,
+      "eval_validation_runtime": 3856.0588,
+      "eval_validation_samples_per_second": 0.648,
+      "eval_validation_steps_per_second": 0.02,
+      "step": 4500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.4392265193370165e-05,
+      "loss": 1.2095,
+      "step": 4520
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.432320441988951e-05,
+      "loss": 1.2017,
+      "step": 4540
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.425414364640884e-05,
+      "loss": 1.2074,
+      "step": 4560
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.418508287292818e-05,
+      "loss": 1.2135,
+      "step": 4580
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.411602209944751e-05,
+      "loss": 1.2094,
+      "step": 4600
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.4046961325966855e-05,
+      "loss": 1.2164,
+      "step": 4620
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.397790055248619e-05,
+      "loss": 1.2172,
+      "step": 4640
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.3908839779005525e-05,
+      "loss": 1.2053,
+      "step": 4660
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.383977900552486e-05,
+      "loss": 1.21,
+      "step": 4680
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.37707182320442e-05,
+      "loss": 1.2026,
+      "step": 4700
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.370165745856354e-05,
+      "loss": 1.2163,
+      "step": 4720
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.363259668508287e-05,
+      "loss": 1.2049,
+      "step": 4740
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.3563535911602215e-05,
+      "loss": 1.1979,
+      "step": 4760
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.349447513812155e-05,
+      "loss": 1.21,
+      "step": 4780
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.3425414364640886e-05,
+      "loss": 1.2098,
+      "step": 4800
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.335635359116022e-05,
+      "loss": 1.2078,
+      "step": 4820
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.328729281767956e-05,
+      "loss": 1.1995,
+      "step": 4840
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.32182320441989e-05,
+      "loss": 1.2005,
+      "step": 4860
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.3149171270718233e-05,
+      "loss": 1.2011,
+      "step": 4880
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.308011049723757e-05,
+      "loss": 1.1911,
+      "step": 4900
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.301104972375691e-05,
+      "loss": 1.2037,
+      "step": 4920
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.294198895027624e-05,
+      "loss": 1.2029,
+      "step": 4940
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.287292817679558e-05,
+      "loss": 1.2084,
+      "step": 4960
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.2803867403314917e-05,
+      "loss": 1.2004,
+      "step": 4980
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.273480662983426e-05,
+      "loss": 1.1973,
+      "step": 5000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.2665745856353594e-05,
+      "loss": 1.2009,
+      "step": 5020
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.259668508287293e-05,
+      "loss": 1.1993,
+      "step": 5040
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.252762430939227e-05,
+      "loss": 1.2096,
+      "step": 5060
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.24585635359116e-05,
+      "loss": 1.2046,
+      "step": 5080
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.238950276243094e-05,
+      "loss": 1.2098,
+      "step": 5100
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.232044198895028e-05,
+      "loss": 1.1993,
+      "step": 5120
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.225138121546962e-05,
+      "loss": 1.2049,
+      "step": 5140
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.218232044198895e-05,
+      "loss": 1.2031,
+      "step": 5160
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.211325966850829e-05,
+      "loss": 1.2015,
+      "step": 5180
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.2044198895027625e-05,
+      "loss": 1.2065,
+      "step": 5200
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.197513812154696e-05,
+      "loss": 1.2032,
+      "step": 5220
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.19060773480663e-05,
+      "loss": 1.2011,
+      "step": 5240
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.183701657458564e-05,
+      "loss": 1.2005,
+      "step": 5260
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.176795580110498e-05,
+      "loss": 1.1961,
+      "step": 5280
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.169889502762431e-05,
+      "loss": 1.2032,
+      "step": 5300
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.162983425414365e-05,
+      "loss": 1.206,
+      "step": 5320
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.1560773480662985e-05,
+      "loss": 1.2145,
+      "step": 5340
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.149171270718232e-05,
+      "loss": 1.2015,
+      "step": 5360
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.1422651933701656e-05,
+      "loss": 1.1948,
+      "step": 5380
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.1353591160221e-05,
+      "loss": 1.2013,
+      "step": 5400
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.128453038674033e-05,
+      "loss": 1.1922,
+      "step": 5420
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.121546961325967e-05,
+      "loss": 1.2024,
+      "step": 5440
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.1146408839779004e-05,
+      "loss": 1.1992,
+      "step": 5460
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.1077348066298346e-05,
+      "loss": 1.1949,
+      "step": 5480
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.100828729281768e-05,
+      "loss": 1.1939,
+      "step": 5500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.0939226519337016e-05,
+      "loss": 1.1998,
+      "step": 5520
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.087016574585636e-05,
+      "loss": 1.199,
+      "step": 5540
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.0801104972375693e-05,
+      "loss": 1.2133,
+      "step": 5560
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.073204419889503e-05,
+      "loss": 1.21,
+      "step": 5580
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.0662983425414364e-05,
+      "loss": 1.2024,
+      "step": 5600
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.0593922651933706e-05,
+      "loss": 1.1914,
+      "step": 5620
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.052486187845304e-05,
+      "loss": 1.1936,
+      "step": 5640
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.0455801104972377e-05,
+      "loss": 1.2003,
+      "step": 5660
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.0386740331491715e-05,
+      "loss": 1.1875,
+      "step": 5680
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.0317679558011054e-05,
+      "loss": 1.1873,
+      "step": 5700
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.0248618784530386e-05,
+      "loss": 1.1945,
+      "step": 5720
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.0179558011049724e-05,
+      "loss": 1.1955,
+      "step": 5740
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.0110497237569063e-05,
+      "loss": 1.2036,
+      "step": 5760
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.00414364640884e-05,
+      "loss": 1.2039,
+      "step": 5780
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.9972375690607734e-05,
+      "loss": 1.2,
+      "step": 5800
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.9903314917127072e-05,
+      "loss": 1.203,
+      "step": 5820
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.983425414364641e-05,
+      "loss": 1.1998,
+      "step": 5840
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.9765193370165746e-05,
+      "loss": 1.1991,
+      "step": 5860
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.9696132596685085e-05,
+      "loss": 1.1915,
+      "step": 5880
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.9627071823204423e-05,
+      "loss": 1.1987,
+      "step": 5900
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.9558011049723762e-05,
+      "loss": 1.1969,
+      "step": 5920
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.9488950276243094e-05,
+      "loss": 1.1976,
+      "step": 5940
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.9419889502762433e-05,
+      "loss": 1.1999,
+      "step": 5960
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.935082872928177e-05,
+      "loss": 1.1921,
+      "step": 5980
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.9281767955801103e-05,
+      "loss": 1.1876,
+      "step": 6000
+    },
+    {
+      "epoch": 1.66,
+      "eval_validation_accuracy": 0.05660693698094773,
+      "eval_validation_loss": 1.2047861814498901,
+      "eval_validation_runtime": 3828.1304,
+      "eval_validation_samples_per_second": 0.653,
+      "eval_validation_steps_per_second": 0.021,
+      "step": 6000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.9212707182320442e-05,
+      "loss": 1.2027,
+      "step": 6020
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.914364640883978e-05,
+      "loss": 1.1934,
+      "step": 6040
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.907458563535912e-05,
+      "loss": 1.1842,
+      "step": 6060
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.900552486187845e-05,
+      "loss": 1.1984,
+      "step": 6080
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.893646408839779e-05,
+      "loss": 1.1872,
+      "step": 6100
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.8867403314917128e-05,
+      "loss": 1.2003,
+      "step": 6120
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.8798342541436463e-05,
+      "loss": 1.1861,
+      "step": 6140
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.8729281767955802e-05,
+      "loss": 1.2005,
+      "step": 6160
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.866022099447514e-05,
+      "loss": 1.1966,
+      "step": 6180
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.859116022099448e-05,
+      "loss": 1.1858,
+      "step": 6200
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.852209944751381e-05,
+      "loss": 1.1919,
+      "step": 6220
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.845303867403315e-05,
+      "loss": 1.1892,
+      "step": 6240
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.838397790055249e-05,
+      "loss": 1.197,
+      "step": 6260
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.8314917127071827e-05,
+      "loss": 1.1915,
+      "step": 6280
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.824585635359116e-05,
+      "loss": 1.1971,
+      "step": 6300
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.8176795580110498e-05,
+      "loss": 1.194,
+      "step": 6320
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.8107734806629836e-05,
+      "loss": 1.1912,
+      "step": 6340
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.8038674033149172e-05,
+      "loss": 1.1908,
+      "step": 6360
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.796961325966851e-05,
+      "loss": 1.1834,
+      "step": 6380
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.790055248618785e-05,
+      "loss": 1.1931,
+      "step": 6400
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.7831491712707188e-05,
+      "loss": 1.192,
+      "step": 6420
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.776243093922652e-05,
+      "loss": 1.1866,
+      "step": 6440
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.7693370165745858e-05,
+      "loss": 1.1933,
+      "step": 6460
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.7624309392265197e-05,
+      "loss": 1.185,
+      "step": 6480
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.755524861878453e-05,
+      "loss": 1.1893,
+      "step": 6500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.7486187845303867e-05,
+      "loss": 1.1897,
+      "step": 6520
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.7417127071823206e-05,
+      "loss": 1.1904,
+      "step": 6540
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.7348066298342545e-05,
+      "loss": 1.189,
+      "step": 6560
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.7279005524861877e-05,
+      "loss": 1.1932,
+      "step": 6580
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.7209944751381215e-05,
+      "loss": 1.1825,
+      "step": 6600
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.7140883977900554e-05,
+      "loss": 1.1913,
+      "step": 6620
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.707182320441989e-05,
+      "loss": 1.1889,
+      "step": 6640
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.7002762430939228e-05,
+      "loss": 1.183,
+      "step": 6660
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.6933701657458566e-05,
+      "loss": 1.1879,
+      "step": 6680
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.6864640883977905e-05,
+      "loss": 1.1879,
+      "step": 6700
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.6795580110497237e-05,
+      "loss": 1.1987,
+      "step": 6720
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.6726519337016576e-05,
+      "loss": 1.1826,
+      "step": 6740
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.6657458563535914e-05,
+      "loss": 1.1854,
+      "step": 6760
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.6588397790055246e-05,
+      "loss": 1.1916,
+      "step": 6780
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.6519337016574585e-05,
+      "loss": 1.1898,
+      "step": 6800
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.6450276243093923e-05,
+      "loss": 1.1857,
+      "step": 6820
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.6381215469613262e-05,
+      "loss": 1.1856,
+      "step": 6840
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.6312154696132597e-05,
+      "loss": 1.1822,
+      "step": 6860
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.6243093922651936e-05,
+      "loss": 1.1986,
+      "step": 6880
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.6174033149171275e-05,
+      "loss": 1.1821,
+      "step": 6900
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.6104972375690613e-05,
+      "loss": 1.1934,
+      "step": 6920
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.6035911602209945e-05,
+      "loss": 1.1831,
+      "step": 6940
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.5966850828729284e-05,
+      "loss": 1.188,
+      "step": 6960
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.5897790055248622e-05,
+      "loss": 1.1808,
+      "step": 6980
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.5828729281767954e-05,
+      "loss": 1.1883,
+      "step": 7000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.5759668508287293e-05,
+      "loss": 1.1858,
+      "step": 7020
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.569060773480663e-05,
+      "loss": 1.1983,
+      "step": 7040
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.562154696132597e-05,
+      "loss": 1.1753,
+      "step": 7060
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.5552486187845302e-05,
+      "loss": 1.187,
+      "step": 7080
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.548342541436464e-05,
+      "loss": 1.1784,
+      "step": 7100
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.541436464088398e-05,
+      "loss": 1.188,
+      "step": 7120
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.5345303867403315e-05,
+      "loss": 1.187,
+      "step": 7140
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.5276243093922653e-05,
+      "loss": 1.1794,
+      "step": 7160
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.5207182320441992e-05,
+      "loss": 1.1819,
+      "step": 7180
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.513812154696133e-05,
+      "loss": 1.1839,
+      "step": 7200
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.5069060773480663e-05,
+      "loss": 1.1828,
+      "step": 7220
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.5e-05,
+      "loss": 1.1921,
+      "step": 7240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4930939226519336e-05,
+      "loss": 1.1571,
+      "step": 7260
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.4861878453038675e-05,
+      "loss": 1.1573,
+      "step": 7280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.479281767955801e-05,
+      "loss": 1.1609,
+      "step": 7300
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.472375690607735e-05,
+      "loss": 1.1587,
+      "step": 7320
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.4654696132596684e-05,
+      "loss": 1.1698,
+      "step": 7340
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.4585635359116023e-05,
+      "loss": 1.1653,
+      "step": 7360
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.451657458563536e-05,
+      "loss": 1.1633,
+      "step": 7380
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.4447513812154697e-05,
+      "loss": 1.1612,
+      "step": 7400
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.4378453038674036e-05,
+      "loss": 1.1526,
+      "step": 7420
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.430939226519337e-05,
+      "loss": 1.1628,
+      "step": 7440
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.424033149171271e-05,
+      "loss": 1.1695,
+      "step": 7460
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.4171270718232045e-05,
+      "loss": 1.1663,
+      "step": 7480
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.4102209944751383e-05,
+      "loss": 1.1601,
+      "step": 7500
+    },
+    {
+      "epoch": 2.07,
+      "eval_validation_accuracy": 0.05679042501221299,
+      "eval_validation_loss": 1.1941546201705933,
+      "eval_validation_runtime": 3780.6761,
+      "eval_validation_samples_per_second": 0.661,
+      "eval_validation_steps_per_second": 0.021,
+      "step": 7500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.403314917127072e-05,
+      "loss": 1.1613,
+      "step": 7520
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.3964088397790057e-05,
+      "loss": 1.1609,
+      "step": 7540
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.3895027624309393e-05,
+      "loss": 1.1597,
+      "step": 7560
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.3825966850828728e-05,
+      "loss": 1.1561,
+      "step": 7580
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.3756906077348066e-05,
+      "loss": 1.154,
+      "step": 7600
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.3687845303867405e-05,
+      "loss": 1.1555,
+      "step": 7620
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.3618784530386744e-05,
+      "loss": 1.163,
+      "step": 7640
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.354972375690608e-05,
+      "loss": 1.1677,
+      "step": 7660
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.3480662983425418e-05,
+      "loss": 1.1662,
+      "step": 7680
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.3411602209944753e-05,
+      "loss": 1.1638,
+      "step": 7700
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.3342541436464088e-05,
+      "loss": 1.1604,
+      "step": 7720
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.3273480662983427e-05,
+      "loss": 1.1516,
+      "step": 7740
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.3204419889502762e-05,
+      "loss": 1.1546,
+      "step": 7760
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.31353591160221e-05,
+      "loss": 1.1565,
+      "step": 7780
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.3066298342541436e-05,
+      "loss": 1.1568,
+      "step": 7800
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.2997237569060775e-05,
+      "loss": 1.1534,
+      "step": 7820
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.292817679558011e-05,
+      "loss": 1.1669,
+      "step": 7840
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.285911602209945e-05,
+      "loss": 1.1558,
+      "step": 7860
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.2790055248618787e-05,
+      "loss": 1.1617,
+      "step": 7880
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.2720994475138122e-05,
+      "loss": 1.1566,
+      "step": 7900
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.265193370165746e-05,
+      "loss": 1.1576,
+      "step": 7920
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.2582872928176796e-05,
+      "loss": 1.1492,
+      "step": 7940
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.2513812154696135e-05,
+      "loss": 1.1662,
+      "step": 7960
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.244475138121547e-05,
+      "loss": 1.1589,
+      "step": 7980
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.237569060773481e-05,
+      "loss": 1.157,
+      "step": 8000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.2306629834254144e-05,
+      "loss": 1.1606,
+      "step": 8020
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.223756906077348e-05,
+      "loss": 1.1461,
+      "step": 8040
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.2168508287292818e-05,
+      "loss": 1.157,
+      "step": 8060
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.2099447513812153e-05,
+      "loss": 1.1612,
+      "step": 8080
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.2030386740331492e-05,
+      "loss": 1.1624,
+      "step": 8100
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.196132596685083e-05,
+      "loss": 1.1586,
+      "step": 8120
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.189226519337017e-05,
+      "loss": 1.1559,
+      "step": 8140
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.1823204419889505e-05,
+      "loss": 1.1603,
+      "step": 8160
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.1754143646408843e-05,
+      "loss": 1.1636,
+      "step": 8180
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.168508287292818e-05,
+      "loss": 1.1554,
+      "step": 8200
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.1616022099447514e-05,
+      "loss": 1.1569,
+      "step": 8220
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.1546961325966852e-05,
+      "loss": 1.1712,
+      "step": 8240
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.1477900552486188e-05,
+      "loss": 1.1532,
+      "step": 8260
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.1408839779005526e-05,
+      "loss": 1.1664,
+      "step": 8280
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.133977900552486e-05,
+      "loss": 1.1619,
+      "step": 8300
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.12707182320442e-05,
+      "loss": 1.158,
+      "step": 8320
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.1201657458563536e-05,
+      "loss": 1.153,
+      "step": 8340
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.1132596685082874e-05,
+      "loss": 1.1608,
+      "step": 8360
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.1063535911602213e-05,
+      "loss": 1.1548,
+      "step": 8380
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.0994475138121548e-05,
+      "loss": 1.1484,
+      "step": 8400
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.0925414364640887e-05,
+      "loss": 1.1576,
+      "step": 8420
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.0856353591160222e-05,
+      "loss": 1.1623,
+      "step": 8440
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.078729281767956e-05,
+      "loss": 1.1555,
+      "step": 8460
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.0718232044198896e-05,
+      "loss": 1.1571,
+      "step": 8480
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.064917127071823e-05,
+      "loss": 1.1557,
+      "step": 8500
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.058011049723757e-05,
+      "loss": 1.1554,
+      "step": 8520
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.0511049723756905e-05,
+      "loss": 1.1554,
+      "step": 8540
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.0441988950276244e-05,
+      "loss": 1.1605,
+      "step": 8560
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.037292817679558e-05,
+      "loss": 1.1535,
+      "step": 8580
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.0303867403314918e-05,
+      "loss": 1.1585,
+      "step": 8600
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.0234806629834256e-05,
+      "loss": 1.162,
+      "step": 8620
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0165745856353595e-05,
+      "loss": 1.1588,
+      "step": 8640
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.009668508287293e-05,
+      "loss": 1.1568,
+      "step": 8660
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0027624309392265e-05,
+      "loss": 1.1548,
+      "step": 8680
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.9958563535911604e-05,
+      "loss": 1.1521,
+      "step": 8700
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.988950276243094e-05,
+      "loss": 1.1612,
+      "step": 8720
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9820441988950278e-05,
+      "loss": 1.1592,
+      "step": 8740
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9751381215469613e-05,
+      "loss": 1.1571,
+      "step": 8760
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.9682320441988952e-05,
+      "loss": 1.1572,
+      "step": 8780
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.9613259668508287e-05,
+      "loss": 1.1507,
+      "step": 8800
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9544198895027623e-05,
+      "loss": 1.1569,
+      "step": 8820
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.947513812154696e-05,
+      "loss": 1.1626,
+      "step": 8840
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.94060773480663e-05,
+      "loss": 1.1565,
+      "step": 8860
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.933701657458564e-05,
+      "loss": 1.1551,
+      "step": 8880
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9267955801104974e-05,
+      "loss": 1.1718,
+      "step": 8900
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9198895027624312e-05,
+      "loss": 1.1512,
+      "step": 8920
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.913328729281768e-05,
+      "loss": 1.1479,
+      "step": 8940
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9064226519337015e-05,
+      "loss": 1.1606,
+      "step": 8960
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.8995165745856354e-05,
+      "loss": 1.1566,
+      "step": 8980
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8926104972375693e-05,
+      "loss": 1.1648,
+      "step": 9000
+    },
+    {
+      "epoch": 2.49,
+      "eval_validation_accuracy": 0.05684982901807523,
+      "eval_validation_loss": 1.1861985921859741,
+      "eval_validation_runtime": 3774.7742,
+      "eval_validation_samples_per_second": 0.662,
+      "eval_validation_steps_per_second": 0.021,
+      "step": 9000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8857044198895028e-05,
+      "loss": 1.1584,
+      "step": 9020
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8787983425414367e-05,
+      "loss": 1.1512,
+      "step": 9040
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8718922651933702e-05,
+      "loss": 1.1547,
+      "step": 9060
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.864986187845304e-05,
+      "loss": 1.1496,
+      "step": 9080
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8580801104972376e-05,
+      "loss": 1.1563,
+      "step": 9100
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8511740331491715e-05,
+      "loss": 1.1484,
+      "step": 9120
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.844267955801105e-05,
+      "loss": 1.154,
+      "step": 9140
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.837361878453039e-05,
+      "loss": 1.1548,
+      "step": 9160
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.8304558011049724e-05,
+      "loss": 1.146,
+      "step": 9180
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.8235497237569062e-05,
+      "loss": 1.1508,
+      "step": 9200
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.8166436464088398e-05,
+      "loss": 1.157,
+      "step": 9220
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.8097375690607736e-05,
+      "loss": 1.1549,
+      "step": 9240
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.8028314917127075e-05,
+      "loss": 1.1618,
+      "step": 9260
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.795925414364641e-05,
+      "loss": 1.1557,
+      "step": 9280
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.789019337016575e-05,
+      "loss": 1.1556,
+      "step": 9300
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7821132596685084e-05,
+      "loss": 1.1507,
+      "step": 9320
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.7752071823204423e-05,
+      "loss": 1.1554,
+      "step": 9340
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.7683011049723758e-05,
+      "loss": 1.1641,
+      "step": 9360
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.7613950276243093e-05,
+      "loss": 1.1503,
+      "step": 9380
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.7544889502762432e-05,
+      "loss": 1.1563,
+      "step": 9400
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.7475828729281767e-05,
+      "loss": 1.1497,
+      "step": 9420
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.7406767955801106e-05,
+      "loss": 1.1498,
+      "step": 9440
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.733770718232044e-05,
+      "loss": 1.1507,
+      "step": 9460
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.726864640883978e-05,
+      "loss": 1.1581,
+      "step": 9480
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.719958563535912e-05,
+      "loss": 1.1477,
+      "step": 9500
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.7130524861878454e-05,
+      "loss": 1.1549,
+      "step": 9520
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.7061464088397792e-05,
+      "loss": 1.1607,
+      "step": 9540
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.6992403314917128e-05,
+      "loss": 1.1493,
+      "step": 9560
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.6923342541436466e-05,
+      "loss": 1.1531,
+      "step": 9580
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.68542817679558e-05,
+      "loss": 1.1539,
+      "step": 9600
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.678522099447514e-05,
+      "loss": 1.1494,
+      "step": 9620
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.6716160220994475e-05,
+      "loss": 1.1517,
+      "step": 9640
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.6647099447513814e-05,
+      "loss": 1.1538,
+      "step": 9660
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.657803867403315e-05,
+      "loss": 1.1565,
+      "step": 9680
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.6508977900552485e-05,
+      "loss": 1.161,
+      "step": 9700
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.6439917127071823e-05,
+      "loss": 1.1495,
+      "step": 9720
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.6370856353591162e-05,
+      "loss": 1.1471,
+      "step": 9740
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.6301795580110497e-05,
+      "loss": 1.1496,
+      "step": 9760
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.6232734806629836e-05,
+      "loss": 1.1469,
+      "step": 9780
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.6163674033149174e-05,
+      "loss": 1.1547,
+      "step": 9800
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.609461325966851e-05,
+      "loss": 1.1518,
+      "step": 9820
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.6025552486187845e-05,
+      "loss": 1.1603,
+      "step": 9840
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.5956491712707184e-05,
+      "loss": 1.1584,
+      "step": 9860
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.588743093922652e-05,
+      "loss": 1.1569,
+      "step": 9880
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.5818370165745858e-05,
+      "loss": 1.1575,
+      "step": 9900
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.5749309392265193e-05,
+      "loss": 1.153,
+      "step": 9920
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.568024861878453e-05,
+      "loss": 1.1521,
+      "step": 9940
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.5611187845303867e-05,
+      "loss": 1.1549,
+      "step": 9960
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5542127071823205e-05,
+      "loss": 1.1564,
+      "step": 9980
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.547306629834254e-05,
+      "loss": 1.154,
+      "step": 10000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.540400552486188e-05,
+      "loss": 1.1477,
+      "step": 10020
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5334944751381218e-05,
+      "loss": 1.1532,
+      "step": 10040
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.5265883977900553e-05,
+      "loss": 1.1548,
+      "step": 10060
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.5196823204419892e-05,
+      "loss": 1.1468,
+      "step": 10080
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.5127762430939227e-05,
+      "loss": 1.1535,
+      "step": 10100
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.5058701657458566e-05,
+      "loss": 1.161,
+      "step": 10120
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.4989640883977901e-05,
+      "loss": 1.1419,
+      "step": 10140
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4920580110497236e-05,
+      "loss": 1.156,
+      "step": 10160
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4851519337016575e-05,
+      "loss": 1.1511,
+      "step": 10180
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.4782458563535912e-05,
+      "loss": 1.1498,
+      "step": 10200
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.471339779005525e-05,
+      "loss": 1.1512,
+      "step": 10220
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4644337016574586e-05,
+      "loss": 1.1495,
+      "step": 10240
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4575276243093924e-05,
+      "loss": 1.1499,
+      "step": 10260
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.450621546961326e-05,
+      "loss": 1.1561,
+      "step": 10280
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.4437154696132598e-05,
+      "loss": 1.1448,
+      "step": 10300
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.4368093922651935e-05,
+      "loss": 1.1534,
+      "step": 10320
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.429903314917127e-05,
+      "loss": 1.1473,
+      "step": 10340
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.422997237569061e-05,
+      "loss": 1.1448,
+      "step": 10360
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.4160911602209944e-05,
+      "loss": 1.1549,
+      "step": 10380
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.4091850828729283e-05,
+      "loss": 1.1463,
+      "step": 10400
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.4022790055248618e-05,
+      "loss": 1.1513,
+      "step": 10420
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.3953729281767957e-05,
+      "loss": 1.1527,
+      "step": 10440
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.3884668508287294e-05,
+      "loss": 1.1476,
+      "step": 10460
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.381560773480663e-05,
+      "loss": 1.1474,
+      "step": 10480
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.3746546961325968e-05,
+      "loss": 1.1451,
+      "step": 10500
+    },
+    {
+      "epoch": 2.9,
+      "eval_validation_accuracy": 0.0568980947728383,
+      "eval_validation_loss": 1.1791410446166992,
+      "eval_validation_runtime": 4192.8215,
+      "eval_validation_samples_per_second": 0.596,
+      "eval_validation_steps_per_second": 0.019,
+      "step": 10500
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.3677486187845303e-05,
+      "loss": 1.1604,
+      "step": 10520
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.3608425414364642e-05,
+      "loss": 1.1568,
+      "step": 10540
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3539364640883979e-05,
+      "loss": 1.1506,
+      "step": 10560
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3470303867403317e-05,
+      "loss": 1.1509,
+      "step": 10580
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3401243093922653e-05,
+      "loss": 1.1569,
+      "step": 10600
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3332182320441988e-05,
+      "loss": 1.1592,
+      "step": 10620
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.3263121546961327e-05,
+      "loss": 1.1508,
+      "step": 10640
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.3194060773480662e-05,
+      "loss": 1.1521,
+      "step": 10660
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3125e-05,
+      "loss": 1.1576,
+      "step": 10680
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.3055939226519338e-05,
+      "loss": 1.151,
+      "step": 10700
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.2986878453038676e-05,
+      "loss": 1.1466,
+      "step": 10720
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.2917817679558011e-05,
+      "loss": 1.1481,
+      "step": 10740
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.284875690607735e-05,
+      "loss": 1.1483,
+      "step": 10760
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.2779696132596685e-05,
+      "loss": 1.1484,
+      "step": 10780
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.2710635359116022e-05,
+      "loss": 1.1515,
+      "step": 10800
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.2641574585635361e-05,
+      "loss": 1.149,
+      "step": 10820
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.2572513812154696e-05,
+      "loss": 1.1494,
+      "step": 10840
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.2503453038674035e-05,
+      "loss": 1.1484,
+      "step": 10860
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.243439226519337e-05,
+      "loss": 1.1214,
+      "step": 10880
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.2365331491712707e-05,
+      "loss": 1.1231,
+      "step": 10900
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2296270718232044e-05,
+      "loss": 1.1282,
+      "step": 10920
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.2227209944751383e-05,
+      "loss": 1.1277,
+      "step": 10940
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.215814917127072e-05,
+      "loss": 1.1402,
+      "step": 10960
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.2089088397790057e-05,
+      "loss": 1.1381,
+      "step": 10980
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.2020027624309394e-05,
+      "loss": 1.1265,
+      "step": 11000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.1950966850828729e-05,
+      "loss": 1.1281,
+      "step": 11020
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.1881906077348066e-05,
+      "loss": 1.1291,
+      "step": 11040
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1812845303867404e-05,
+      "loss": 1.1359,
+      "step": 11060
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.1743784530386741e-05,
+      "loss": 1.1254,
+      "step": 11080
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1674723756906078e-05,
+      "loss": 1.1372,
+      "step": 11100
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.1605662983425415e-05,
+      "loss": 1.1399,
+      "step": 11120
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.1536602209944752e-05,
+      "loss": 1.1284,
+      "step": 11140
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.146754143646409e-05,
+      "loss": 1.1254,
+      "step": 11160
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.1398480662983426e-05,
+      "loss": 1.1288,
+      "step": 11180
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.1329419889502763e-05,
+      "loss": 1.1285,
+      "step": 11200
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.12603591160221e-05,
+      "loss": 1.1259,
+      "step": 11220
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.1191298342541437e-05,
+      "loss": 1.1295,
+      "step": 11240
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.1122237569060774e-05,
+      "loss": 1.1193,
+      "step": 11260
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.1053176795580111e-05,
+      "loss": 1.1344,
+      "step": 11280
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.0984116022099448e-05,
+      "loss": 1.1316,
+      "step": 11300
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0915055248618785e-05,
+      "loss": 1.1275,
+      "step": 11320
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0845994475138122e-05,
+      "loss": 1.1334,
+      "step": 11340
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.0776933701657459e-05,
+      "loss": 1.1347,
+      "step": 11360
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.0707872928176796e-05,
+      "loss": 1.1372,
+      "step": 11380
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.0638812154696133e-05,
+      "loss": 1.1316,
+      "step": 11400
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.056975138121547e-05,
+      "loss": 1.1325,
+      "step": 11420
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.0500690607734807e-05,
+      "loss": 1.1321,
+      "step": 11440
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0431629834254145e-05,
+      "loss": 1.1351,
+      "step": 11460
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0362569060773482e-05,
+      "loss": 1.1345,
+      "step": 11480
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0293508287292817e-05,
+      "loss": 1.1245,
+      "step": 11500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0224447513812154e-05,
+      "loss": 1.1249,
+      "step": 11520
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.0155386740331491e-05,
+      "loss": 1.1274,
+      "step": 11540
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.008632596685083e-05,
+      "loss": 1.1315,
+      "step": 11560
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0017265193370167e-05,
+      "loss": 1.1261,
+      "step": 11580
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 9.948204419889504e-06,
+      "loss": 1.1276,
+      "step": 11600
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 9.879143646408841e-06,
+      "loss": 1.1312,
+      "step": 11620
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.810082872928178e-06,
+      "loss": 1.1353,
+      "step": 11640
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.741022099447513e-06,
+      "loss": 1.1243,
+      "step": 11660
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 9.671961325966852e-06,
+      "loss": 1.1343,
+      "step": 11680
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 9.602900552486189e-06,
+      "loss": 1.1238,
+      "step": 11700
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.533839779005526e-06,
+      "loss": 1.1315,
+      "step": 11720
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.464779005524863e-06,
+      "loss": 1.1358,
+      "step": 11740
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.3957182320442e-06,
+      "loss": 1.1229,
+      "step": 11760
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.326657458563537e-06,
+      "loss": 1.1281,
+      "step": 11780
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.257596685082874e-06,
+      "loss": 1.1318,
+      "step": 11800
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.18853591160221e-06,
+      "loss": 1.1232,
+      "step": 11820
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 9.119475138121547e-06,
+      "loss": 1.1234,
+      "step": 11840
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 9.050414364640884e-06,
+      "loss": 1.1334,
+      "step": 11860
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 8.981353591160221e-06,
+      "loss": 1.1309,
+      "step": 11880
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.912292817679558e-06,
+      "loss": 1.1335,
+      "step": 11900
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.843232044198895e-06,
+      "loss": 1.1322,
+      "step": 11920
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.774171270718232e-06,
+      "loss": 1.1206,
+      "step": 11940
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.70511049723757e-06,
+      "loss": 1.1333,
+      "step": 11960
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.636049723756906e-06,
+      "loss": 1.1361,
+      "step": 11980
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.566988950276243e-06,
+      "loss": 1.1267,
+      "step": 12000
+    },
+    {
+      "epoch": 3.31,
+      "eval_validation_accuracy": 0.0569209574987787,
+      "eval_validation_loss": 1.1777355670928955,
+      "eval_validation_runtime": 3858.1389,
+      "eval_validation_samples_per_second": 0.648,
+      "eval_validation_steps_per_second": 0.02,
+      "step": 12000
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 8.512841756420879e-06,
+      "loss": 1.119,
+      "step": 12020
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.447252140292737e-06,
+      "loss": 1.1313,
+      "step": 12040
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.37821043910522e-06,
+      "loss": 1.1329,
+      "step": 12060
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.309168737917703e-06,
+      "loss": 1.1308,
+      "step": 12080
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.240127036730186e-06,
+      "loss": 1.1366,
+      "step": 12100
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 8.171085335542667e-06,
+      "loss": 1.1307,
+      "step": 12120
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 8.102043634355152e-06,
+      "loss": 1.1324,
+      "step": 12140
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 8.033001933167633e-06,
+      "loss": 1.1218,
+      "step": 12160
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 7.963960231980116e-06,
+      "loss": 1.1272,
+      "step": 12180
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 7.894918530792599e-06,
+      "loss": 1.1233,
+      "step": 12200
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 7.825876829605082e-06,
+      "loss": 1.1307,
+      "step": 12220
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 7.756835128417565e-06,
+      "loss": 1.1303,
+      "step": 12240
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 7.687793427230048e-06,
+      "loss": 1.1308,
+      "step": 12260
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 7.618751726042531e-06,
+      "loss": 1.1223,
+      "step": 12280
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 7.549710024855013e-06,
+      "loss": 1.1238,
+      "step": 12300
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 7.480668323667496e-06,
+      "loss": 1.1288,
+      "step": 12320
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 7.411626622479978e-06,
+      "loss": 1.1397,
+      "step": 12340
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 7.3425849212924615e-06,
+      "loss": 1.1284,
+      "step": 12360
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 7.2735432201049436e-06,
+      "loss": 1.1339,
+      "step": 12380
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 7.2045015189174265e-06,
+      "loss": 1.1286,
+      "step": 12400
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 7.1354598177299086e-06,
+      "loss": 1.1287,
+      "step": 12420
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 7.066418116542392e-06,
+      "loss": 1.1277,
+      "step": 12440
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 6.997376415354874e-06,
+      "loss": 1.1333,
+      "step": 12460
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 6.928334714167357e-06,
+      "loss": 1.1332,
+      "step": 12480
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 6.859293012979841e-06,
+      "loss": 1.1307,
+      "step": 12500
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 6.790251311792322e-06,
+      "loss": 1.1345,
+      "step": 12520
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 6.721209610604806e-06,
+      "loss": 1.1289,
+      "step": 12540
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 6.652167909417288e-06,
+      "loss": 1.1281,
+      "step": 12560
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 6.583126208229771e-06,
+      "loss": 1.1329,
+      "step": 12580
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 6.514084507042253e-06,
+      "loss": 1.1175,
+      "step": 12600
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 6.445042805854737e-06,
+      "loss": 1.1262,
+      "step": 12620
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 6.376001104667219e-06,
+      "loss": 1.1296,
+      "step": 12640
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 6.306959403479702e-06,
+      "loss": 1.1193,
+      "step": 12660
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 6.237917702292185e-06,
+      "loss": 1.1256,
+      "step": 12680
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 6.168876001104668e-06,
+      "loss": 1.1308,
+      "step": 12700
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 6.09983429991715e-06,
+      "loss": 1.1341,
+      "step": 12720
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 6.030792598729633e-06,
+      "loss": 1.1321,
+      "step": 12740
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5.961750897542116e-06,
+      "loss": 1.128,
+      "step": 12760
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.892709196354598e-06,
+      "loss": 1.1271,
+      "step": 12780
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.823667495167081e-06,
+      "loss": 1.1327,
+      "step": 12800
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5.754625793979564e-06,
+      "loss": 1.1255,
+      "step": 12820
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.685584092792047e-06,
+      "loss": 1.1373,
+      "step": 12840
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.61654239160453e-06,
+      "loss": 1.1212,
+      "step": 12860
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.5475006904170125e-06,
+      "loss": 1.1303,
+      "step": 12880
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.4784589892294955e-06,
+      "loss": 1.1241,
+      "step": 12900
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.4094172880419775e-06,
+      "loss": 1.1265,
+      "step": 12920
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.3403755868544605e-06,
+      "loss": 1.1099,
+      "step": 12940
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.271333885666943e-06,
+      "loss": 1.128,
+      "step": 12960
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.2022921844794255e-06,
+      "loss": 1.117,
+      "step": 12980
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.133250483291908e-06,
+      "loss": 1.1224,
+      "step": 13000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.064208782104391e-06,
+      "loss": 1.1249,
+      "step": 13020
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 4.995167080916873e-06,
+      "loss": 1.1282,
+      "step": 13040
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 4.926125379729356e-06,
+      "loss": 1.1242,
+      "step": 13060
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 4.85708367854184e-06,
+      "loss": 1.1241,
+      "step": 13080
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 4.788041977354322e-06,
+      "loss": 1.1212,
+      "step": 13100
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 4.719000276166805e-06,
+      "loss": 1.1248,
+      "step": 13120
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.649958574979288e-06,
+      "loss": 1.1305,
+      "step": 13140
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.58091687379177e-06,
+      "loss": 1.1375,
+      "step": 13160
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 4.511875172604253e-06,
+      "loss": 1.1184,
+      "step": 13180
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 4.442833471416736e-06,
+      "loss": 1.1294,
+      "step": 13200
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 4.373791770229219e-06,
+      "loss": 1.1304,
+      "step": 13220
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.304750069041701e-06,
+      "loss": 1.1247,
+      "step": 13240
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.235708367854184e-06,
+      "loss": 1.1344,
+      "step": 13260
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.1284,
+      "step": 13280
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 4.097624965479149e-06,
+      "loss": 1.1289,
+      "step": 13300
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 4.028583264291632e-06,
+      "loss": 1.1179,
+      "step": 13320
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.959541563104116e-06,
+      "loss": 1.1256,
+      "step": 13340
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.890499861916598e-06,
+      "loss": 1.1258,
+      "step": 13360
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.821458160729081e-06,
+      "loss": 1.1317,
+      "step": 13380
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.7524164595415636e-06,
+      "loss": 1.1232,
+      "step": 13400
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.683374758354046e-06,
+      "loss": 1.1293,
+      "step": 13420
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.6143330571665286e-06,
+      "loss": 1.1266,
+      "step": 13440
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.5452913559790115e-06,
+      "loss": 1.1317,
+      "step": 13460
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.476249654791494e-06,
+      "loss": 1.1345,
+      "step": 13480
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.407207953603977e-06,
+      "loss": 1.1195,
+      "step": 13500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.3381662524164595e-06,
+      "loss": 1.135,
+      "step": 13520
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.269124551228942e-06,
+      "loss": 1.1401,
+      "step": 13540
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.200082850041425e-06,
+      "loss": 1.1335,
+      "step": 13560
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.1310411488539083e-06,
+      "loss": 1.1305,
+      "step": 13580
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.0619994476663903e-06,
+      "loss": 1.1276,
+      "step": 13600
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 2.9929577464788733e-06,
+      "loss": 1.1257,
+      "step": 13620
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 2.923916045291356e-06,
+      "loss": 1.1228,
+      "step": 13640
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 2.854874344103839e-06,
+      "loss": 1.1342,
+      "step": 13660
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.7858326429163216e-06,
+      "loss": 1.1366,
+      "step": 13680
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.716790941728804e-06,
+      "loss": 1.132,
+      "step": 13700
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 2.647749240541287e-06,
+      "loss": 1.1308,
+      "step": 13720
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 2.5787075393537696e-06,
+      "loss": 1.1297,
+      "step": 13740
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 2.5096658381662525e-06,
+      "loss": 1.1241,
+      "step": 13760
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.4406241369787354e-06,
+      "loss": 1.1288,
+      "step": 13780
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.371582435791218e-06,
+      "loss": 1.1226,
+      "step": 13800
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 2.302540734603701e-06,
+      "loss": 1.1224,
+      "step": 13820
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 2.2334990334161834e-06,
+      "loss": 1.1243,
+      "step": 13840
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 2.164457332228666e-06,
+      "loss": 1.1295,
+      "step": 13860
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 2.0954156310411492e-06,
+      "loss": 1.1227,
+      "step": 13880
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 2.0263739298536317e-06,
+      "loss": 1.1185,
+      "step": 13900
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.9573322286661142e-06,
+      "loss": 1.1259,
+      "step": 13920
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.8882905274785972e-06,
+      "loss": 1.1294,
+      "step": 13940
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.8192488262910799e-06,
+      "loss": 1.1304,
+      "step": 13960
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.7502071251035624e-06,
+      "loss": 1.1263,
+      "step": 13980
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.6811654239160455e-06,
+      "loss": 1.133,
+      "step": 14000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.6121237227285282e-06,
+      "loss": 1.122,
+      "step": 14020
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.5465341066003867e-06,
+      "loss": 1.1327,
+      "step": 14040
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.4774924054128694e-06,
+      "loss": 1.1308,
+      "step": 14060
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.4084507042253521e-06,
+      "loss": 1.1351,
+      "step": 14080
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.3394090030378348e-06,
+      "loss": 1.1237,
+      "step": 14100
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.2703673018503175e-06,
+      "loss": 1.1318,
+      "step": 14120
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.2013256006628005e-06,
+      "loss": 1.1226,
+      "step": 14140
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.1322838994752832e-06,
+      "loss": 1.127,
+      "step": 14160
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.0632421982877657e-06,
+      "loss": 1.1215,
+      "step": 14180
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.942004971002486e-07,
+      "loss": 1.1305,
+      "step": 14200
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.251587959127313e-07,
+      "loss": 1.1265,
+      "step": 14220
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 8.56117094725214e-07,
+      "loss": 1.12,
+      "step": 14240
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 7.870753935376969e-07,
+      "loss": 1.1219,
+      "step": 14260
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 7.180336923501795e-07,
+      "loss": 1.1257,
+      "step": 14280
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.489919911626623e-07,
+      "loss": 1.1321,
+      "step": 14300
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.79950289975145e-07,
+      "loss": 1.1264,
+      "step": 14320
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.109085887876277e-07,
+      "loss": 1.1272,
+      "step": 14340
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.418668876001105e-07,
+      "loss": 1.1233,
+      "step": 14360
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 3.7282518641259323e-07,
+      "loss": 1.1326,
+      "step": 14380
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.03783485225076e-07,
+      "loss": 1.1286,
+      "step": 14400
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 2.347417840375587e-07,
+      "loss": 1.1198,
+      "step": 14420
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.657000828500414e-07,
+      "loss": 1.1193,
+      "step": 14440
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.665838166252417e-08,
+      "loss": 1.1329,
+      "step": 14460
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 2.7616680475006907e-08,
+      "loss": 1.127,
+      "step": 14480
+    },
+    {
+      "epoch": 4.0,
+      "step": 14484,
+      "total_flos": 6.028721702271818e+19,
+      "train_loss": 0.11551530145209522,
+      "train_runtime": 3653.4642,
+      "train_samples_per_second": 507.345,
+      "train_steps_per_second": 3.964
+    }
+  ],
+  "max_steps": 14484,
+  "num_train_epochs": 4,
+  "total_flos": 6.028721702271818e+19,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3f099f5182e50add4876993ef34bd991b3d18a989b842b2e0885ca042525e5e
+size 4027