Initial release v1.0

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+checkpoint-3708/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,61 @@
----
-license: apache-2.0
----
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen3-8B-Base
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: seed-42
+  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. -->
+
+# seed-42
+
+This model is a fine-tuned version of [/share/wulijun/caimz/model/base-model/models--Qwen--Qwen3-8B-Base/snapshots/49e3418fbbbca6ecbdf9608b4d22e5a407081db4](https://huggingface.co//share/wulijun/caimz/model/base-model/models--Qwen--Qwen3-8B-Base/snapshots/49e3418fbbbca6ecbdf9608b4d22e5a407081db4) on the overall-v6b-100k 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: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 8
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 32
+- total_eval_batch_size: 64
+- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 3.0
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.55.0
+- Pytorch 2.6.0+cu124
+- Datasets 3.2.0
+- Tokenizers 0.21.0

README.md.backup

@@ -0,0 +1,61 @@
+---
+library_name: transformers
+license: other
+base_model: /share/wulijun/caimz/model/base-model/models--Qwen--Qwen3-8B-Base/snapshots/49e3418fbbbca6ecbdf9608b4d22e5a407081db4
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: seed-42
+  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. -->
+
+# seed-42
+
+This model is a fine-tuned version of [/share/wulijun/caimz/model/base-model/models--Qwen--Qwen3-8B-Base/snapshots/49e3418fbbbca6ecbdf9608b4d22e5a407081db4](https://huggingface.co//share/wulijun/caimz/model/base-model/models--Qwen--Qwen3-8B-Base/snapshots/49e3418fbbbca6ecbdf9608b4d22e5a407081db4) on the overall-v6b-100k 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: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 8
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 32
+- total_eval_batch_size: 64
+- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 3.0
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.55.0
+- Pytorch 2.6.0+cu124
+- Datasets 3.2.0
+- Tokenizers 0.21.0

added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 7.26626862104576e+18,
+    "train_loss": 0.24862830043482034,
+    "train_runtime": 257107.0797,
+    "train_samples_per_second": 0.461,
+    "train_steps_per_second": 0.014
+}

chat_template.jinja

@@ -0,0 +1,4 @@
+{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ 'System: ' + system_message + '<|endoftext|>' + '
+' }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ 'Human: ' + content + '<|endoftext|>' + '
+Assistant:' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + '
+' }}{% endif %}{% endfor %}

checkpoint-3708/added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoint-3708/chat_template.jinja

@@ -0,0 +1,4 @@
+{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ 'System: ' + system_message + '<|endoftext|>' + '
+' }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ 'Human: ' + content + '<|endoftext|>' + '
+Assistant:' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + '
+' }}{% endif %}{% endfor %}

checkpoint-3708/config.json

@@ -0,0 +1,68 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 12288,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 36,
+  "model_type": "qwen3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.55.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

checkpoint-3708/generation_config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048,
+  "transformers_version": "4.55.0"
+}

checkpoint-3708/model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53096ac7bb0e3a0ec6cecc4d902d62f520e74ac30e63b423214283a814c0776f
+size 4902257696

checkpoint-3708/model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e02590c760b06bef8b03c37ab5156352f2a1c7f5862d1aed747ea0a1a5bc6f4d
+size 4915960368

checkpoint-3708/model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca5be3fa1663f2990e1bdc3cc029546d69e03538f02e9faa9a049b93261cb3a5
+size 4983068496

checkpoint-3708/model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99a72c2cad443584179996d1d7fa5d512319232128449148213f31e469b660e7
+size 1580230264

checkpoint-3708/model.safetensors.index.json

@@ -0,0 +1,407 @@
+{
+  "metadata": {
+    "total_parameters": 308224,
+    "total_size": 16381470720
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00004-of-00004.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.k_norm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.q_norm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.norm.weight": "model-00004-of-00004.safetensors"
+  }
+}

checkpoint-3708/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-3708/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

checkpoint-3708/tokenizer_config.json

@@ -0,0 +1,240 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-3708/trainer_state.json

@@ -0,0 +1,2624 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3708,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008090614886731391,
+      "grad_norm": 22.52252141115457,
+      "learning_rate": 1.2129380053908356e-06,
+      "loss": 2.5582,
+      "step": 10
+    },
+    {
+      "epoch": 0.016181229773462782,
+      "grad_norm": 11.441648898342923,
+      "learning_rate": 2.560646900269542e-06,
+      "loss": 2.2025,
+      "step": 20
+    },
+    {
+      "epoch": 0.024271844660194174,
+      "grad_norm": 5.370882200215843,
+      "learning_rate": 3.908355795148248e-06,
+      "loss": 1.668,
+      "step": 30
+    },
+    {
+      "epoch": 0.032362459546925564,
+      "grad_norm": 2.0115313434456827,
+      "learning_rate": 5.256064690026954e-06,
+      "loss": 1.1266,
+      "step": 40
+    },
+    {
+      "epoch": 0.040453074433656956,
+      "grad_norm": 1.2036774860036918,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.7799,
+      "step": 50
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 0.8106871294631399,
+      "learning_rate": 7.951482479784367e-06,
+      "loss": 0.5907,
+      "step": 60
+    },
+    {
+      "epoch": 0.05663430420711974,
+      "grad_norm": 0.6572037767143685,
+      "learning_rate": 9.299191374663074e-06,
+      "loss": 0.4977,
+      "step": 70
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 0.9743925361281945,
+      "learning_rate": 1.0646900269541779e-05,
+      "loss": 0.4728,
+      "step": 80
+    },
+    {
+      "epoch": 0.07281553398058252,
+      "grad_norm": 0.5302813737105591,
+      "learning_rate": 1.1994609164420486e-05,
+      "loss": 0.446,
+      "step": 90
+    },
+    {
+      "epoch": 0.08090614886731391,
+      "grad_norm": 0.5504892707748702,
+      "learning_rate": 1.3342318059299191e-05,
+      "loss": 0.4164,
+      "step": 100
+    },
+    {
+      "epoch": 0.0889967637540453,
+      "grad_norm": 0.46842577857609835,
+      "learning_rate": 1.4690026954177898e-05,
+      "loss": 0.412,
+      "step": 110
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 0.5228315101871648,
+      "learning_rate": 1.6037735849056604e-05,
+      "loss": 0.3975,
+      "step": 120
+    },
+    {
+      "epoch": 0.10517799352750809,
+      "grad_norm": 0.5623706583567197,
+      "learning_rate": 1.738544474393531e-05,
+      "loss": 0.3857,
+      "step": 130
+    },
+    {
+      "epoch": 0.11326860841423948,
+      "grad_norm": 0.3506071062876399,
+      "learning_rate": 1.8733153638814018e-05,
+      "loss": 0.3837,
+      "step": 140
+    },
+    {
+      "epoch": 0.12135922330097088,
+      "grad_norm": 0.491171903870991,
+      "learning_rate": 2.0080862533692725e-05,
+      "loss": 0.3721,
+      "step": 150
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 0.5505284472221431,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.3734,
+      "step": 160
+    },
+    {
+      "epoch": 0.13754045307443366,
+      "grad_norm": 0.8767975628519997,
+      "learning_rate": 2.2776280323450135e-05,
+      "loss": 0.365,
+      "step": 170
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 0.5343891120951508,
+      "learning_rate": 2.4123989218328842e-05,
+      "loss": 0.3631,
+      "step": 180
+    },
+    {
+      "epoch": 0.15372168284789645,
+      "grad_norm": 0.47266629407427385,
+      "learning_rate": 2.547169811320755e-05,
+      "loss": 0.3509,
+      "step": 190
+    },
+    {
+      "epoch": 0.16181229773462782,
+      "grad_norm": 0.49039579917599835,
+      "learning_rate": 2.6819407008086256e-05,
+      "loss": 0.3474,
+      "step": 200
+    },
+    {
+      "epoch": 0.16990291262135923,
+      "grad_norm": 0.5095673082261557,
+      "learning_rate": 2.8167115902964963e-05,
+      "loss": 0.3439,
+      "step": 210
+    },
+    {
+      "epoch": 0.1779935275080906,
+      "grad_norm": 0.7559655324445135,
+      "learning_rate": 2.9514824797843667e-05,
+      "loss": 0.3477,
+      "step": 220
+    },
+    {
+      "epoch": 0.18608414239482202,
+      "grad_norm": 0.717837164870727,
+      "learning_rate": 3.086253369272237e-05,
+      "loss": 0.3424,
+      "step": 230
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 0.44154340671189124,
+      "learning_rate": 3.221024258760108e-05,
+      "loss": 0.3432,
+      "step": 240
+    },
+    {
+      "epoch": 0.2022653721682848,
+      "grad_norm": 0.645302977385022,
+      "learning_rate": 3.355795148247979e-05,
+      "loss": 0.3378,
+      "step": 250
+    },
+    {
+      "epoch": 0.21035598705501618,
+      "grad_norm": 0.45392402289172107,
+      "learning_rate": 3.490566037735849e-05,
+      "loss": 0.34,
+      "step": 260
+    },
+    {
+      "epoch": 0.21844660194174756,
+      "grad_norm": 0.7064296610826926,
+      "learning_rate": 3.62533692722372e-05,
+      "loss": 0.339,
+      "step": 270
+    },
+    {
+      "epoch": 0.22653721682847897,
+      "grad_norm": 0.5611292380408083,
+      "learning_rate": 3.76010781671159e-05,
+      "loss": 0.3284,
+      "step": 280
+    },
+    {
+      "epoch": 0.23462783171521034,
+      "grad_norm": 0.5448155574311832,
+      "learning_rate": 3.894878706199461e-05,
+      "loss": 0.3291,
+      "step": 290
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 0.43795871099809547,
+      "learning_rate": 4.0296495956873316e-05,
+      "loss": 0.3206,
+      "step": 300
+    },
+    {
+      "epoch": 0.25080906148867316,
+      "grad_norm": 0.5584253534316864,
+      "learning_rate": 4.164420485175202e-05,
+      "loss": 0.3276,
+      "step": 310
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 0.5617857932874714,
+      "learning_rate": 4.299191374663073e-05,
+      "loss": 0.3258,
+      "step": 320
+    },
+    {
+      "epoch": 0.2669902912621359,
+      "grad_norm": 0.5441488728977509,
+      "learning_rate": 4.433962264150944e-05,
+      "loss": 0.3227,
+      "step": 330
+    },
+    {
+      "epoch": 0.2750809061488673,
+      "grad_norm": 0.8386999082473845,
+      "learning_rate": 4.5687331536388144e-05,
+      "loss": 0.3196,
+      "step": 340
+    },
+    {
+      "epoch": 0.28317152103559873,
+      "grad_norm": 0.43608786652562637,
+      "learning_rate": 4.703504043126685e-05,
+      "loss": 0.3191,
+      "step": 350
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.4937683097487802,
+      "learning_rate": 4.838274932614555e-05,
+      "loss": 0.3255,
+      "step": 360
+    },
+    {
+      "epoch": 0.2993527508090615,
+      "grad_norm": 0.5061197818365496,
+      "learning_rate": 4.973045822102426e-05,
+      "loss": 0.3119,
+      "step": 370
+    },
+    {
+      "epoch": 0.3074433656957929,
+      "grad_norm": 0.4744335618999143,
+      "learning_rate": 4.9999290952604396e-05,
+      "loss": 0.3191,
+      "step": 380
+    },
+    {
+      "epoch": 0.3155339805825243,
+      "grad_norm": 0.5037259460580787,
+      "learning_rate": 4.9996410516491115e-05,
+      "loss": 0.3098,
+      "step": 390
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 0.5710713036434062,
+      "learning_rate": 4.9991314631296585e-05,
+      "loss": 0.3096,
+      "step": 400
+    },
+    {
+      "epoch": 0.33171521035598706,
+      "grad_norm": 0.33451540591505297,
+      "learning_rate": 4.9984003748672604e-05,
+      "loss": 0.3095,
+      "step": 410
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 0.4340252206831366,
+      "learning_rate": 4.997447851658774e-05,
+      "loss": 0.3076,
+      "step": 420
+    },
+    {
+      "epoch": 0.3478964401294498,
+      "grad_norm": 0.3838413262491291,
+      "learning_rate": 4.9962739779269887e-05,
+      "loss": 0.3067,
+      "step": 430
+    },
+    {
+      "epoch": 0.3559870550161812,
+      "grad_norm": 0.42248192127450584,
+      "learning_rate": 4.9948788577131414e-05,
+      "loss": 0.2989,
+      "step": 440
+    },
+    {
+      "epoch": 0.3640776699029126,
+      "grad_norm": 0.4026141748836561,
+      "learning_rate": 4.993262614667696e-05,
+      "loss": 0.3018,
+      "step": 450
+    },
+    {
+      "epoch": 0.37216828478964403,
+      "grad_norm": 0.4295230581487718,
+      "learning_rate": 4.9914253920393884e-05,
+      "loss": 0.3048,
+      "step": 460
+    },
+    {
+      "epoch": 0.3802588996763754,
+      "grad_norm": 0.3384646839950234,
+      "learning_rate": 4.9893673526625265e-05,
+      "loss": 0.3025,
+      "step": 470
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.41348113959511545,
+      "learning_rate": 4.987088678942555e-05,
+      "loss": 0.2997,
+      "step": 480
+    },
+    {
+      "epoch": 0.3964401294498382,
+      "grad_norm": 0.3673430436712548,
+      "learning_rate": 4.984589572839897e-05,
+      "loss": 0.2981,
+      "step": 490
+    },
+    {
+      "epoch": 0.4045307443365696,
+      "grad_norm": 0.3637541700718482,
+      "learning_rate": 4.9818702558520485e-05,
+      "loss": 0.2994,
+      "step": 500
+    },
+    {
+      "epoch": 0.41262135922330095,
+      "grad_norm": 0.3667744606459147,
+      "learning_rate": 4.978930968993946e-05,
+      "loss": 0.3027,
+      "step": 510
+    },
+    {
+      "epoch": 0.42071197411003236,
+      "grad_norm": 0.3783856834139639,
+      "learning_rate": 4.9757719727766085e-05,
+      "loss": 0.2995,
+      "step": 520
+    },
+    {
+      "epoch": 0.42880258899676377,
+      "grad_norm": 0.39358597558975966,
+      "learning_rate": 4.972393547184046e-05,
+      "loss": 0.2968,
+      "step": 530
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.3961355130202433,
+      "learning_rate": 4.968795991648446e-05,
+      "loss": 0.2951,
+      "step": 540
+    },
+    {
+      "epoch": 0.4449838187702265,
+      "grad_norm": 0.37157203976493997,
+      "learning_rate": 4.9649796250236344e-05,
+      "loss": 0.2842,
+      "step": 550
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 0.2840234936379272,
+      "learning_rate": 4.960944785556814e-05,
+      "loss": 0.2862,
+      "step": 560
+    },
+    {
+      "epoch": 0.46116504854368934,
+      "grad_norm": 0.320746304504261,
+      "learning_rate": 4.956691830858585e-05,
+      "loss": 0.2925,
+      "step": 570
+    },
+    {
+      "epoch": 0.4692556634304207,
+      "grad_norm": 0.3627941299734378,
+      "learning_rate": 4.952221137871252e-05,
+      "loss": 0.2884,
+      "step": 580
+    },
+    {
+      "epoch": 0.4773462783171521,
+      "grad_norm": 0.44860484414052815,
+      "learning_rate": 4.947533102835413e-05,
+      "loss": 0.2937,
+      "step": 590
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.3404390735893878,
+      "learning_rate": 4.942628141254843e-05,
+      "loss": 0.2888,
+      "step": 600
+    },
+    {
+      "epoch": 0.4935275080906149,
+      "grad_norm": 0.3346590749564513,
+      "learning_rate": 4.937506687859666e-05,
+      "loss": 0.2839,
+      "step": 610
+    },
+    {
+      "epoch": 0.5016181229773463,
+      "grad_norm": 0.3540605558733981,
+      "learning_rate": 4.932169196567824e-05,
+      "loss": 0.2926,
+      "step": 620
+    },
+    {
+      "epoch": 0.5097087378640777,
+      "grad_norm": 0.3225115085834403,
+      "learning_rate": 4.9266161404448454e-05,
+      "loss": 0.2914,
+      "step": 630
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 0.333757773957392,
+      "learning_rate": 4.920848011661919e-05,
+      "loss": 0.2844,
+      "step": 640
+    },
+    {
+      "epoch": 0.5258899676375405,
+      "grad_norm": 0.3691189283201263,
+      "learning_rate": 4.914865321452274e-05,
+      "loss": 0.2838,
+      "step": 650
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 0.41635131408315823,
+      "learning_rate": 4.908668600065862e-05,
+      "loss": 0.2853,
+      "step": 660
+    },
+    {
+      "epoch": 0.5420711974110033,
+      "grad_norm": 0.3161547689784877,
+      "learning_rate": 4.90225839672237e-05,
+      "loss": 0.2844,
+      "step": 670
+    },
+    {
+      "epoch": 0.5501618122977346,
+      "grad_norm": 0.32264889950016085,
+      "learning_rate": 4.8956352795625325e-05,
+      "loss": 0.2912,
+      "step": 680
+    },
+    {
+      "epoch": 0.558252427184466,
+      "grad_norm": 0.3354534030777558,
+      "learning_rate": 4.8887998355977886e-05,
+      "loss": 0.2826,
+      "step": 690
+    },
+    {
+      "epoch": 0.5663430420711975,
+      "grad_norm": 0.3409096712904377,
+      "learning_rate": 4.881752670658244e-05,
+      "loss": 0.2799,
+      "step": 700
+    },
+    {
+      "epoch": 0.5744336569579288,
+      "grad_norm": 0.3012793393888447,
+      "learning_rate": 4.87449440933898e-05,
+      "loss": 0.2713,
+      "step": 710
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.3127062042429517,
+      "learning_rate": 4.867025694944698e-05,
+      "loss": 0.2813,
+      "step": 720
+    },
+    {
+      "epoch": 0.5906148867313916,
+      "grad_norm": 0.3269471051340985,
+      "learning_rate": 4.859347189432699e-05,
+      "loss": 0.28,
+      "step": 730
+    },
+    {
+      "epoch": 0.598705501618123,
+      "grad_norm": 0.32543025148595867,
+      "learning_rate": 4.8514595733542144e-05,
+      "loss": 0.2709,
+      "step": 740
+    },
+    {
+      "epoch": 0.6067961165048543,
+      "grad_norm": 0.3147551999550132,
+      "learning_rate": 4.8433635457940915e-05,
+      "loss": 0.2814,
+      "step": 750
+    },
+    {
+      "epoch": 0.6148867313915858,
+      "grad_norm": 0.26606594884988966,
+      "learning_rate": 4.8350598243088283e-05,
+      "loss": 0.2776,
+      "step": 760
+    },
+    {
+      "epoch": 0.6229773462783171,
+      "grad_norm": 0.3007757906909445,
+      "learning_rate": 4.8265491448629804e-05,
+      "loss": 0.2767,
+      "step": 770
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.3795450058305756,
+      "learning_rate": 4.817832261763928e-05,
+      "loss": 0.2772,
+      "step": 780
+    },
+    {
+      "epoch": 0.63915857605178,
+      "grad_norm": 0.3106723870622729,
+      "learning_rate": 4.8089099475950257e-05,
+      "loss": 0.276,
+      "step": 790
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 0.34373131058866446,
+      "learning_rate": 4.7997829931471225e-05,
+      "loss": 0.2765,
+      "step": 800
+    },
+    {
+      "epoch": 0.6553398058252428,
+      "grad_norm": 0.2846997869981113,
+      "learning_rate": 4.7904522073484786e-05,
+      "loss": 0.2778,
+      "step": 810
+    },
+    {
+      "epoch": 0.6634304207119741,
+      "grad_norm": 0.31767113534168995,
+      "learning_rate": 4.780918417193065e-05,
+      "loss": 0.2775,
+      "step": 820
+    },
+    {
+      "epoch": 0.6715210355987055,
+      "grad_norm": 0.31169855988969003,
+      "learning_rate": 4.7711824676672726e-05,
+      "loss": 0.2733,
+      "step": 830
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.3175387528791534,
+      "learning_rate": 4.76124522167501e-05,
+      "loss": 0.2746,
+      "step": 840
+    },
+    {
+      "epoch": 0.6877022653721683,
+      "grad_norm": 0.2975263968714564,
+      "learning_rate": 4.751107559961238e-05,
+      "loss": 0.2745,
+      "step": 850
+    },
+    {
+      "epoch": 0.6957928802588996,
+      "grad_norm": 0.29446496624170493,
+      "learning_rate": 4.740770381033894e-05,
+      "loss": 0.275,
+      "step": 860
+    },
+    {
+      "epoch": 0.7038834951456311,
+      "grad_norm": 0.2973639432877628,
+      "learning_rate": 4.730234601084268e-05,
+      "loss": 0.2716,
+      "step": 870
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 0.2497924694108393,
+      "learning_rate": 4.719501153905793e-05,
+      "loss": 0.2673,
+      "step": 880
+    },
+    {
+      "epoch": 0.7200647249190939,
+      "grad_norm": 0.2948558193473281,
+      "learning_rate": 4.7085709908112866e-05,
+      "loss": 0.2731,
+      "step": 890
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.2459558041018532,
+      "learning_rate": 4.6974450805486305e-05,
+      "loss": 0.2639,
+      "step": 900
+    },
+    {
+      "epoch": 0.7362459546925566,
+      "grad_norm": 0.2559758327802977,
+      "learning_rate": 4.686124409214917e-05,
+      "loss": 0.2715,
+      "step": 910
+    },
+    {
+      "epoch": 0.7443365695792881,
+      "grad_norm": 0.2811398346458149,
+      "learning_rate": 4.674609980169042e-05,
+      "loss": 0.2712,
+      "step": 920
+    },
+    {
+      "epoch": 0.7524271844660194,
+      "grad_norm": 0.2686221572337129,
+      "learning_rate": 4.662902813942784e-05,
+      "loss": 0.272,
+      "step": 930
+    },
+    {
+      "epoch": 0.7605177993527508,
+      "grad_norm": 0.2755997369035635,
+      "learning_rate": 4.651003948150349e-05,
+      "loss": 0.2689,
+      "step": 940
+    },
+    {
+      "epoch": 0.7686084142394822,
+      "grad_norm": 0.2672412475643565,
+      "learning_rate": 4.638914437396408e-05,
+      "loss": 0.2693,
+      "step": 950
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.23259525487399887,
+      "learning_rate": 4.626635353182626e-05,
+      "loss": 0.2659,
+      "step": 960
+    },
+    {
+      "epoch": 0.7847896440129449,
+      "grad_norm": 0.26440852988259955,
+      "learning_rate": 4.614167783812694e-05,
+      "loss": 0.2753,
+      "step": 970
+    },
+    {
+      "epoch": 0.7928802588996764,
+      "grad_norm": 0.2786648778182051,
+      "learning_rate": 4.601512834295874e-05,
+      "loss": 0.272,
+      "step": 980
+    },
+    {
+      "epoch": 0.8009708737864077,
+      "grad_norm": 0.2475631464266887,
+      "learning_rate": 4.588671626249057e-05,
+      "loss": 0.2631,
+      "step": 990
+    },
+    {
+      "epoch": 0.8090614886731392,
+      "grad_norm": 0.23380841398688618,
+      "learning_rate": 4.5756452977973585e-05,
+      "loss": 0.2609,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8171521035598706,
+      "grad_norm": 0.27565297705248776,
+      "learning_rate": 4.56243500347324e-05,
+      "loss": 0.2651,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.29059051987658346,
+      "learning_rate": 4.549041914114188e-05,
+      "loss": 0.2687,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.29594618957533503,
+      "learning_rate": 4.535467216758936e-05,
+      "loss": 0.2666,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 0.3931170619188329,
+      "learning_rate": 4.5217121145422616e-05,
+      "loss": 0.2631,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8495145631067961,
+      "grad_norm": 0.2857742902635426,
+      "learning_rate": 4.5077778265883477e-05,
+      "loss": 0.2632,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8576051779935275,
+      "grad_norm": 0.24489808715354155,
+      "learning_rate": 4.4936655879027336e-05,
+      "loss": 0.2634,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8656957928802589,
+      "grad_norm": 0.26376201661023657,
+      "learning_rate": 4.479376649262855e-05,
+      "loss": 0.266,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.25183442403581696,
+      "learning_rate": 4.464912277107185e-05,
+      "loss": 0.2658,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8818770226537217,
+      "grad_norm": 0.2613869002025676,
+      "learning_rate": 4.450273753422992e-05,
+      "loss": 0.2653,
+      "step": 1090
+    },
+    {
+      "epoch": 0.889967637540453,
+      "grad_norm": 0.29304343706553704,
+      "learning_rate": 4.435462375632711e-05,
+      "loss": 0.2651,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8980582524271845,
+      "grad_norm": 0.263122059677539,
+      "learning_rate": 4.420479456478957e-05,
+      "loss": 0.2621,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 0.24882087207880094,
+      "learning_rate": 4.405326323908178e-05,
+      "loss": 0.2635,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9142394822006472,
+      "grad_norm": 0.2744368878036377,
+      "learning_rate": 4.390004320952947e-05,
+      "loss": 0.2639,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.25617953359336987,
+      "learning_rate": 4.374514805612942e-05,
+      "loss": 0.2667,
+      "step": 1140
+    },
+    {
+      "epoch": 0.93042071197411,
+      "grad_norm": 0.26969276785611856,
+      "learning_rate": 4.358859150734576e-05,
+      "loss": 0.2592,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9385113268608414,
+      "grad_norm": 0.2703864111603607,
+      "learning_rate": 4.343038743889324e-05,
+      "loss": 0.2602,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9466019417475728,
+      "grad_norm": 0.3115866006041078,
+      "learning_rate": 4.3270549872507415e-05,
+      "loss": 0.26,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9546925566343042,
+      "grad_norm": 0.2611547498092126,
+      "learning_rate": 4.3109092974701895e-05,
+      "loss": 0.2591,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9627831715210357,
+      "grad_norm": 0.26931477864354814,
+      "learning_rate": 4.2946031055512733e-05,
+      "loss": 0.26,
+      "step": 1190
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.2578523252517678,
+      "learning_rate": 4.2781378567230145e-05,
+      "loss": 0.2633,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9789644012944984,
+      "grad_norm": 0.27451378253592124,
+      "learning_rate": 4.2615150103117576e-05,
+      "loss": 0.2598,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9870550161812298,
+      "grad_norm": 0.2609329028667746,
+      "learning_rate": 4.24473603961183e-05,
+      "loss": 0.2587,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9951456310679612,
+      "grad_norm": 0.2280958023284859,
+      "learning_rate": 4.227802431754961e-05,
+      "loss": 0.2589,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0032362459546926,
+      "grad_norm": 0.30070050651075947,
+      "learning_rate": 4.210715687578481e-05,
+      "loss": 0.2468,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0113268608414239,
+      "grad_norm": 0.26616355843357287,
+      "learning_rate": 4.193477321492293e-05,
+      "loss": 0.2219,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.2713492665233822,
+      "learning_rate": 4.176088861344657e-05,
+      "loss": 0.2251,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0275080906148868,
+      "grad_norm": 0.21687794735432822,
+      "learning_rate": 4.158551848286773e-05,
+      "loss": 0.2178,
+      "step": 1270
+    },
+    {
+      "epoch": 1.035598705501618,
+      "grad_norm": 0.2539615253588423,
+      "learning_rate": 4.140867836636189e-05,
+      "loss": 0.2289,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0436893203883495,
+      "grad_norm": 0.24657642734847438,
+      "learning_rate": 4.1230383937390374e-05,
+      "loss": 0.2258,
+      "step": 1290
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 0.2609125461360237,
+      "learning_rate": 4.1050650998311215e-05,
+      "loss": 0.2216,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0598705501618122,
+      "grad_norm": 0.3630380994135679,
+      "learning_rate": 4.086949547897862e-05,
+      "loss": 0.2225,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.2504138070534478,
+      "learning_rate": 4.068693343533103e-05,
+      "loss": 0.2227,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0760517799352751,
+      "grad_norm": 0.3126812459563885,
+      "learning_rate": 4.050298104796812e-05,
+      "loss": 0.2268,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0841423948220066,
+      "grad_norm": 0.2257523743221347,
+      "learning_rate": 4.0317654620716704e-05,
+      "loss": 0.2306,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0922330097087378,
+      "grad_norm": 0.21960504136006673,
+      "learning_rate": 4.013097057918566e-05,
+      "loss": 0.2254,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1003236245954693,
+      "grad_norm": 0.250009270408735,
+      "learning_rate": 3.9942945469310175e-05,
+      "loss": 0.2279,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1084142394822007,
+      "grad_norm": 0.23480544241701537,
+      "learning_rate": 3.9753595955885264e-05,
+      "loss": 0.2256,
+      "step": 1370
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.23720387815460467,
+      "learning_rate": 3.9562938821088705e-05,
+      "loss": 0.2235,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1245954692556634,
+      "grad_norm": 0.23437981665878366,
+      "learning_rate": 3.9370990962993695e-05,
+      "loss": 0.2237,
+      "step": 1390
+    },
+    {
+      "epoch": 1.132686084142395,
+      "grad_norm": 0.2286079680729574,
+      "learning_rate": 3.9177769394071086e-05,
+      "loss": 0.2215,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1407766990291262,
+      "grad_norm": 0.22687828650289346,
+      "learning_rate": 3.898329123968163e-05,
+      "loss": 0.2254,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1488673139158576,
+      "grad_norm": 0.2224235012119624,
+      "learning_rate": 3.87875737365581e-05,
+      "loss": 0.2184,
+      "step": 1420
+    },
+    {
+      "epoch": 1.156957928802589,
+      "grad_norm": 0.20195576636612733,
+      "learning_rate": 3.8590634231277626e-05,
+      "loss": 0.2176,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.20327918718818466,
+      "learning_rate": 3.8392490178724184e-05,
+      "loss": 0.2207,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1731391585760518,
+      "grad_norm": 0.21744881881616449,
+      "learning_rate": 3.8193159140541645e-05,
+      "loss": 0.223,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 0.30658821143823883,
+      "learning_rate": 3.7992658783577215e-05,
+      "loss": 0.2264,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1893203883495145,
+      "grad_norm": 0.25174532799295857,
+      "learning_rate": 3.779100687831563e-05,
+      "loss": 0.2216,
+      "step": 1470
+    },
+    {
+      "epoch": 1.197411003236246,
+      "grad_norm": 0.2561981869189091,
+      "learning_rate": 3.758822129730415e-05,
+      "loss": 0.2201,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2055016181229774,
+      "grad_norm": 0.22464952614110853,
+      "learning_rate": 3.738432001356851e-05,
+      "loss": 0.2239,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.2100829197882277,
+      "learning_rate": 3.7179321099019916e-05,
+      "loss": 0.2281,
+      "step": 1500
+    },
+    {
+      "epoch": 1.22168284789644,
+      "grad_norm": 0.2469541669228544,
+      "learning_rate": 3.6973242722853365e-05,
+      "loss": 0.225,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2297734627831716,
+      "grad_norm": 0.23312096116267617,
+      "learning_rate": 3.6766103149937295e-05,
+      "loss": 0.2246,
+      "step": 1520
+    },
+    {
+      "epoch": 1.237864077669903,
+      "grad_norm": 0.20677868772581298,
+      "learning_rate": 3.655792073919471e-05,
+      "loss": 0.2249,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 0.2563183131089248,
+      "learning_rate": 3.634871394197607e-05,
+      "loss": 0.2263,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2540453074433657,
+      "grad_norm": 0.22128405881374966,
+      "learning_rate": 3.6138501300423934e-05,
+      "loss": 0.2211,
+      "step": 1550
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.21006117612744632,
+      "learning_rate": 3.592730144582948e-05,
+      "loss": 0.2243,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2702265372168284,
+      "grad_norm": 0.1975344463355082,
+      "learning_rate": 3.571513309698131e-05,
+      "loss": 0.2202,
+      "step": 1570
+    },
+    {
+      "epoch": 1.27831715210356,
+      "grad_norm": 0.20997495813930359,
+      "learning_rate": 3.5502015058506335e-05,
+      "loss": 0.2229,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2864077669902914,
+      "grad_norm": 0.2332593631600526,
+      "learning_rate": 3.528796621920307e-05,
+      "loss": 0.219,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2944983818770226,
+      "grad_norm": 0.22815002833576079,
+      "learning_rate": 3.50730055503676e-05,
+      "loss": 0.222,
+      "step": 1600
+    },
+    {
+      "epoch": 1.302588996763754,
+      "grad_norm": 0.20437252443820447,
+      "learning_rate": 3.485715210411204e-05,
+      "loss": 0.2262,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.2395647563898874,
+      "learning_rate": 3.4640425011676034e-05,
+      "loss": 0.2226,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3187702265372168,
+      "grad_norm": 0.22072548502378347,
+      "learning_rate": 3.442284348173106e-05,
+      "loss": 0.2254,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3268608414239482,
+      "grad_norm": 0.21606415951025024,
+      "learning_rate": 3.420442679867796e-05,
+      "loss": 0.2244,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3349514563106797,
+      "grad_norm": 0.27863651532858746,
+      "learning_rate": 3.398519432093782e-05,
+      "loss": 0.2227,
+      "step": 1650
+    },
+    {
+      "epoch": 1.343042071197411,
+      "grad_norm": 0.2202632134109216,
+      "learning_rate": 3.376516547923614e-05,
+      "loss": 0.2195,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3511326860841424,
+      "grad_norm": 0.21524751059743225,
+      "learning_rate": 3.3544359774880714e-05,
+      "loss": 0.2203,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.20473236269563994,
+      "learning_rate": 3.3322796778033204e-05,
+      "loss": 0.2151,
+      "step": 1680
+    },
+    {
+      "epoch": 1.367313915857605,
+      "grad_norm": 0.21469153751522554,
+      "learning_rate": 3.3100496125974624e-05,
+      "loss": 0.216,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 0.19954679234440933,
+      "learning_rate": 3.2877477521364895e-05,
+      "loss": 0.2162,
+      "step": 1700
+    },
+    {
+      "epoch": 1.383495145631068,
+      "grad_norm": 0.18861748166722786,
+      "learning_rate": 3.2653760730496555e-05,
+      "loss": 0.2187,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3915857605177995,
+      "grad_norm": 0.20341053182705085,
+      "learning_rate": 3.242936558154285e-05,
+      "loss": 0.225,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3996763754045307,
+      "grad_norm": 0.19372792735024014,
+      "learning_rate": 3.2204311962800426e-05,
+      "loss": 0.2239,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.27663774087598847,
+      "learning_rate": 3.197861982092651e-05,
+      "loss": 0.2201,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4158576051779934,
+      "grad_norm": 0.20101300631018146,
+      "learning_rate": 3.175230915917108e-05,
+      "loss": 0.2234,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4239482200647249,
+      "grad_norm": 0.20722987740988344,
+      "learning_rate": 3.152540003560398e-05,
+      "loss": 0.219,
+      "step": 1760
+    },
+    {
+      "epoch": 1.4320388349514563,
+      "grad_norm": 0.23987446750061972,
+      "learning_rate": 3.129791256133712e-05,
+      "loss": 0.2158,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 0.2072184752121984,
+      "learning_rate": 3.106986689874204e-05,
+      "loss": 0.2208,
+      "step": 1780
+    },
+    {
+      "epoch": 1.448220064724919,
+      "grad_norm": 0.19280709770923388,
+      "learning_rate": 3.0841283259662875e-05,
+      "loss": 0.2227,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.20536031119127787,
+      "learning_rate": 3.0612181903625014e-05,
+      "loss": 0.222,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4644012944983817,
+      "grad_norm": 0.22626011728625453,
+      "learning_rate": 3.0382583136039444e-05,
+      "loss": 0.2174,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4724919093851132,
+      "grad_norm": 0.2105569387760891,
+      "learning_rate": 3.015250730640308e-05,
+      "loss": 0.2167,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4805825242718447,
+      "grad_norm": 0.20632917153940236,
+      "learning_rate": 2.9921974806495178e-05,
+      "loss": 0.2141,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4886731391585761,
+      "grad_norm": 0.2021951348802143,
+      "learning_rate": 2.969100606856998e-05,
+      "loss": 0.2189,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4967637540453074,
+      "grad_norm": 0.21580708092456813,
+      "learning_rate": 2.9459621563545825e-05,
+      "loss": 0.2236,
+      "step": 1850
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.19252241914699963,
+      "learning_rate": 2.9227841799190775e-05,
+      "loss": 0.2182,
+      "step": 1860
+    },
+    {
+      "epoch": 1.51294498381877,
+      "grad_norm": 0.28670687006718837,
+      "learning_rate": 2.8995687318304975e-05,
+      "loss": 0.2224,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5210355987055015,
+      "grad_norm": 0.19972242558140968,
+      "learning_rate": 2.8763178696899995e-05,
+      "loss": 0.219,
+      "step": 1880
+    },
+    {
+      "epoch": 1.529126213592233,
+      "grad_norm": 0.21361706883180026,
+      "learning_rate": 2.853033654237507e-05,
+      "loss": 0.2207,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5372168284789645,
+      "grad_norm": 0.22498931655736576,
+      "learning_rate": 2.8297181491690756e-05,
+      "loss": 0.2191,
+      "step": 1900
+    },
+    {
+      "epoch": 1.545307443365696,
+      "grad_norm": 0.23189215817184283,
+      "learning_rate": 2.8063734209539773e-05,
+      "loss": 0.2192,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 0.22214336417666555,
+      "learning_rate": 2.783001538651554e-05,
+      "loss": 0.2196,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5614886731391586,
+      "grad_norm": 0.22032252634525124,
+      "learning_rate": 2.7596045737278336e-05,
+      "loss": 0.2202,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5695792880258899,
+      "grad_norm": 0.19861461201811947,
+      "learning_rate": 2.7361845998719315e-05,
+      "loss": 0.2224,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5776699029126213,
+      "grad_norm": 0.20017500987570222,
+      "learning_rate": 2.7127436928122612e-05,
+      "loss": 0.2196,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5857605177993528,
+      "grad_norm": 0.2065059828986781,
+      "learning_rate": 2.6892839301325623e-05,
+      "loss": 0.2134,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5938511326860842,
+      "grad_norm": 0.20590200374163903,
+      "learning_rate": 2.6658073910877603e-05,
+      "loss": 0.2167,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.1855510838919491,
+      "learning_rate": 2.6423161564196803e-05,
+      "loss": 0.2161,
+      "step": 1980
+    },
+    {
+      "epoch": 1.610032362459547,
+      "grad_norm": 0.23296838329760938,
+      "learning_rate": 2.6188123081726306e-05,
+      "loss": 0.2174,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6181229773462782,
+      "grad_norm": 0.21045668674547632,
+      "learning_rate": 2.5952979295088714e-05,
+      "loss": 0.2149,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6262135922330097,
+      "grad_norm": 0.19447307206087575,
+      "learning_rate": 2.57177510452398e-05,
+      "loss": 0.2148,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6343042071197411,
+      "grad_norm": 0.23090130774168546,
+      "learning_rate": 2.5482459180621377e-05,
+      "loss": 0.2191,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6423948220064726,
+      "grad_norm": 0.18913378090947097,
+      "learning_rate": 2.524712455531347e-05,
+      "loss": 0.2136,
+      "step": 2030
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.20240602934546922,
+      "learning_rate": 2.501176802718599e-05,
+      "loss": 0.2153,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6585760517799353,
+      "grad_norm": 0.18200383138065016,
+      "learning_rate": 2.4776410456050165e-05,
+      "loss": 0.2149,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.1919414220183083,
+      "learning_rate": 2.4541072701809624e-05,
+      "loss": 0.2187,
+      "step": 2060
+    },
+    {
+      "epoch": 1.674757281553398,
+      "grad_norm": 0.194424204771118,
+      "learning_rate": 2.4305775622611627e-05,
+      "loss": 0.2156,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6828478964401294,
+      "grad_norm": 0.1986457800050313,
+      "learning_rate": 2.4070540072998372e-05,
+      "loss": 0.2186,
+      "step": 2080
+    },
+    {
+      "epoch": 1.690938511326861,
+      "grad_norm": 0.20702148931842815,
+      "learning_rate": 2.3835386902058637e-05,
+      "loss": 0.2145,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.1933456613050236,
+      "learning_rate": 2.360033695157995e-05,
+      "loss": 0.213,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7071197411003236,
+      "grad_norm": 0.1902051798469252,
+      "learning_rate": 2.3365411054201315e-05,
+      "loss": 0.2154,
+      "step": 2110
+    },
+    {
+      "epoch": 1.715210355987055,
+      "grad_norm": 0.2128582736922773,
+      "learning_rate": 2.3130630031566818e-05,
+      "loss": 0.2209,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7233009708737863,
+      "grad_norm": 0.16887432451943601,
+      "learning_rate": 2.2896014692480226e-05,
+      "loss": 0.2156,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7313915857605178,
+      "grad_norm": 0.20654290160014185,
+      "learning_rate": 2.266158583106063e-05,
+      "loss": 0.2156,
+      "step": 2140
+    },
+    {
+      "epoch": 1.7394822006472492,
+      "grad_norm": 0.1889181780602885,
+      "learning_rate": 2.2427364224899502e-05,
+      "loss": 0.2152,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.20357991974988612,
+      "learning_rate": 2.2193370633219115e-05,
+      "loss": 0.2137,
+      "step": 2160
+    },
+    {
+      "epoch": 1.755663430420712,
+      "grad_norm": 0.18958212912498537,
+      "learning_rate": 2.1959625795032664e-05,
+      "loss": 0.2113,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7637540453074434,
+      "grad_norm": 0.18671823093768689,
+      "learning_rate": 2.1726150427306182e-05,
+      "loss": 0.2146,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7718446601941746,
+      "grad_norm": 0.18915530672104663,
+      "learning_rate": 2.1492965223122305e-05,
+      "loss": 0.2132,
+      "step": 2190
+    },
+    {
+      "epoch": 1.779935275080906,
+      "grad_norm": 0.23999090664815534,
+      "learning_rate": 2.126009084984629e-05,
+      "loss": 0.2124,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7880258899676376,
+      "grad_norm": 0.1958651936225041,
+      "learning_rate": 2.102754794729426e-05,
+      "loss": 0.2151,
+      "step": 2210
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.1977750621585522,
+      "learning_rate": 2.079535712590382e-05,
+      "loss": 0.2107,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8042071197411005,
+      "grad_norm": 0.17805905963116433,
+      "learning_rate": 2.056353896490742e-05,
+      "loss": 0.2091,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8122977346278317,
+      "grad_norm": 0.17973603360437174,
+      "learning_rate": 2.0332114010508334e-05,
+      "loss": 0.2114,
+      "step": 2240
+    },
+    {
+      "epoch": 1.820388349514563,
+      "grad_norm": 0.18242429802007595,
+      "learning_rate": 2.010110277405966e-05,
+      "loss": 0.213,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8284789644012944,
+      "grad_norm": 0.18356014531964865,
+      "learning_rate": 1.9870525730246424e-05,
+      "loss": 0.2142,
+      "step": 2260
+    },
+    {
+      "epoch": 1.8365695792880259,
+      "grad_norm": 0.18174264807445242,
+      "learning_rate": 1.9640403315270824e-05,
+      "loss": 0.212,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.19024588215963859,
+      "learning_rate": 1.9410755925041006e-05,
+      "loss": 0.2107,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8527508090614888,
+      "grad_norm": 0.19119554332148833,
+      "learning_rate": 1.918160391336335e-05,
+      "loss": 0.2117,
+      "step": 2290
+    },
+    {
+      "epoch": 1.86084142394822,
+      "grad_norm": 0.19462064918607552,
+      "learning_rate": 1.8952967590138472e-05,
+      "loss": 0.2099,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8689320388349513,
+      "grad_norm": 0.1800377494816668,
+      "learning_rate": 1.8724867219561203e-05,
+      "loss": 0.2154,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8770226537216828,
+      "grad_norm": 0.20601322758987628,
+      "learning_rate": 1.8497323018324476e-05,
+      "loss": 0.2145,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8851132686084142,
+      "grad_norm": 0.1930709584493409,
+      "learning_rate": 1.8270355153827598e-05,
+      "loss": 0.2077,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.1911370351967725,
+      "learning_rate": 1.804398374238872e-05,
+      "loss": 0.2102,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9012944983818771,
+      "grad_norm": 0.18726153065650458,
+      "learning_rate": 1.781822884746196e-05,
+      "loss": 0.2098,
+      "step": 2350
+    },
+    {
+      "epoch": 1.9093851132686084,
+      "grad_norm": 0.1786359702194776,
+      "learning_rate": 1.7593110477859153e-05,
+      "loss": 0.2114,
+      "step": 2360
+    },
+    {
+      "epoch": 1.9174757281553398,
+      "grad_norm": 0.17464418924684114,
+      "learning_rate": 1.736864858597645e-05,
+      "loss": 0.2125,
+      "step": 2370
+    },
+    {
+      "epoch": 1.925566343042071,
+      "grad_norm": 0.19553554568438702,
+      "learning_rate": 1.7144863066025955e-05,
+      "loss": 0.2101,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9336569579288025,
+      "grad_norm": 0.1919410672013163,
+      "learning_rate": 1.692177375227242e-05,
+      "loss": 0.2148,
+      "step": 2390
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.19222701771364425,
+      "learning_rate": 1.669940041727538e-05,
+      "loss": 0.2135,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9498381877022655,
+      "grad_norm": 0.19090816514102554,
+      "learning_rate": 1.6477762770136707e-05,
+      "loss": 0.2107,
+      "step": 2410
+    },
+    {
+      "epoch": 1.9579288025889967,
+      "grad_norm": 0.196303834522943,
+      "learning_rate": 1.625688045475371e-05,
+      "loss": 0.2154,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9660194174757282,
+      "grad_norm": 0.2021430652987766,
+      "learning_rate": 1.603677304807815e-05,
+      "loss": 0.2085,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9741100323624594,
+      "grad_norm": 0.19099797924736667,
+      "learning_rate": 1.5817460058381088e-05,
+      "loss": 0.2089,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9822006472491909,
+      "grad_norm": 0.19913570866606684,
+      "learning_rate": 1.5598960923523842e-05,
+      "loss": 0.2088,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.18039032373193892,
+      "learning_rate": 1.5381295009235262e-05,
+      "loss": 0.2033,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9983818770226538,
+      "grad_norm": 0.17656611003487743,
+      "learning_rate": 1.5164481607395238e-05,
+      "loss": 0.2073,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0064724919093853,
+      "grad_norm": 0.2203078879292808,
+      "learning_rate": 1.4948539934324923e-05,
+      "loss": 0.184,
+      "step": 2480
+    },
+    {
+      "epoch": 2.0145631067961167,
+      "grad_norm": 0.19377235611735016,
+      "learning_rate": 1.4733489129083534e-05,
+      "loss": 0.1729,
+      "step": 2490
+    },
+    {
+      "epoch": 2.0226537216828477,
+      "grad_norm": 0.19353245562742485,
+      "learning_rate": 1.4519348251772058e-05,
+      "loss": 0.1732,
+      "step": 2500
+    },
+    {
+      "epoch": 2.030744336569579,
+      "grad_norm": 0.19647629833009353,
+      "learning_rate": 1.4306136281843962e-05,
+      "loss": 0.1712,
+      "step": 2510
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.16744835404717143,
+      "learning_rate": 1.4093872116422979e-05,
+      "loss": 0.1703,
+      "step": 2520
+    },
+    {
+      "epoch": 2.046925566343042,
+      "grad_norm": 0.17703272841109144,
+      "learning_rate": 1.3882574568628315e-05,
+      "loss": 0.1706,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0550161812297736,
+      "grad_norm": 0.2024362245821369,
+      "learning_rate": 1.3672262365907163e-05,
+      "loss": 0.1695,
+      "step": 2540
+    },
+    {
+      "epoch": 2.063106796116505,
+      "grad_norm": 0.1767034203926602,
+      "learning_rate": 1.3462954148374899e-05,
+      "loss": 0.1701,
+      "step": 2550
+    },
+    {
+      "epoch": 2.071197411003236,
+      "grad_norm": 0.18197977986545294,
+      "learning_rate": 1.3254668467163029e-05,
+      "loss": 0.1722,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0792880258899675,
+      "grad_norm": 0.17760389056330828,
+      "learning_rate": 1.3047423782774937e-05,
+      "loss": 0.1696,
+      "step": 2570
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.19203143690929964,
+      "learning_rate": 1.2841238463449743e-05,
+      "loss": 0.1712,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0954692556634305,
+      "grad_norm": 0.17183295745816207,
+      "learning_rate": 1.2636130783534319e-05,
+      "loss": 0.1711,
+      "step": 2590
+    },
+    {
+      "epoch": 2.103559870550162,
+      "grad_norm": 0.171911071755841,
+      "learning_rate": 1.2432118921863604e-05,
+      "loss": 0.1661,
+      "step": 2600
+    },
+    {
+      "epoch": 2.1116504854368934,
+      "grad_norm": 0.16535945547449518,
+      "learning_rate": 1.2229220960149431e-05,
+      "loss": 0.1721,
+      "step": 2610
+    },
+    {
+      "epoch": 2.1197411003236244,
+      "grad_norm": 0.15768148348111047,
+      "learning_rate": 1.2027454881377889e-05,
+      "loss": 0.1681,
+      "step": 2620
+    },
+    {
+      "epoch": 2.127831715210356,
+      "grad_norm": 0.16958702285946278,
+      "learning_rate": 1.1826838568215526e-05,
+      "loss": 0.1693,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.16417738734331427,
+      "learning_rate": 1.1627389801424351e-05,
+      "loss": 0.1674,
+      "step": 2640
+    },
+    {
+      "epoch": 2.144012944983819,
+      "grad_norm": 0.1875141618399819,
+      "learning_rate": 1.1429126258285946e-05,
+      "loss": 0.1743,
+      "step": 2650
+    },
+    {
+      "epoch": 2.1521035598705502,
+      "grad_norm": 0.1840991587252681,
+      "learning_rate": 1.1232065511034696e-05,
+      "loss": 0.1749,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1601941747572817,
+      "grad_norm": 0.18075731721273597,
+      "learning_rate": 1.1036225025300357e-05,
+      "loss": 0.1758,
+      "step": 2670
+    },
+    {
+      "epoch": 2.168284789644013,
+      "grad_norm": 0.1751267297742484,
+      "learning_rate": 1.0841622158560085e-05,
+      "loss": 0.1762,
+      "step": 2680
+    },
+    {
+      "epoch": 2.176375404530744,
+      "grad_norm": 0.17441997867016168,
+      "learning_rate": 1.0648274158599994e-05,
+      "loss": 0.1747,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.16061544075925469,
+      "learning_rate": 1.0456198161986489e-05,
+      "loss": 0.171,
+      "step": 2700
+    },
+    {
+      "epoch": 2.192556634304207,
+      "grad_norm": 0.17229966007961503,
+      "learning_rate": 1.0265411192547462e-05,
+      "loss": 0.1712,
+      "step": 2710
+    },
+    {
+      "epoch": 2.2006472491909386,
+      "grad_norm": 0.17576497117173268,
+      "learning_rate": 1.0075930159863416e-05,
+      "loss": 0.1713,
+      "step": 2720
+    },
+    {
+      "epoch": 2.20873786407767,
+      "grad_norm": 0.17654170464263277,
+      "learning_rate": 9.887771857768796e-06,
+      "loss": 0.1716,
+      "step": 2730
+    },
+    {
+      "epoch": 2.2168284789644015,
+      "grad_norm": 0.15949424974193477,
+      "learning_rate": 9.700952962863513e-06,
+      "loss": 0.1724,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2249190938511325,
+      "grad_norm": 0.1797493219887846,
+      "learning_rate": 9.515490033034893e-06,
+      "loss": 0.1747,
+      "step": 2750
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.1640136227884059,
+      "learning_rate": 9.331399505990168e-06,
+      "loss": 0.17,
+      "step": 2760
+    },
+    {
+      "epoch": 2.2411003236245954,
+      "grad_norm": 0.1761964552588544,
+      "learning_rate": 9.148697697799533e-06,
+      "loss": 0.1744,
+      "step": 2770
+    },
+    {
+      "epoch": 2.249190938511327,
+      "grad_norm": 0.16581441786521817,
+      "learning_rate": 8.967400801450105e-06,
+      "loss": 0.1683,
+      "step": 2780
+    },
+    {
+      "epoch": 2.2572815533980584,
+      "grad_norm": 0.15986397423733795,
+      "learning_rate": 8.787524885410678e-06,
+      "loss": 0.1684,
+      "step": 2790
+    },
+    {
+      "epoch": 2.26537216828479,
+      "grad_norm": 0.16854533205632355,
+      "learning_rate": 8.60908589220758e-06,
+      "loss": 0.1695,
+      "step": 2800
+    },
+    {
+      "epoch": 2.273462783171521,
+      "grad_norm": 0.16192651639080963,
+      "learning_rate": 8.432099637011693e-06,
+      "loss": 0.1686,
+      "step": 2810
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.15876123318459937,
+      "learning_rate": 8.256581806236704e-06,
+      "loss": 0.1727,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2896440129449838,
+      "grad_norm": 0.15953467663865786,
+      "learning_rate": 8.082547956148873e-06,
+      "loss": 0.1702,
+      "step": 2830
+    },
+    {
+      "epoch": 2.2977346278317152,
+      "grad_norm": 0.15986140875482005,
+      "learning_rate": 7.91001351148819e-06,
+      "loss": 0.1688,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3058252427184467,
+      "grad_norm": 0.16937360962913361,
+      "learning_rate": 7.738993764101324e-06,
+      "loss": 0.1687,
+      "step": 2850
+    },
+    {
+      "epoch": 2.313915857605178,
+      "grad_norm": 0.1624566036577771,
+      "learning_rate": 7.569503871586292e-06,
+      "loss": 0.1709,
+      "step": 2860
+    },
+    {
+      "epoch": 2.3220064724919096,
+      "grad_norm": 0.17174966186956542,
+      "learning_rate": 7.401558855949004e-06,
+      "loss": 0.168,
+      "step": 2870
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.1619257670204373,
+      "learning_rate": 7.235173602271875e-06,
+      "loss": 0.1697,
+      "step": 2880
+    },
+    {
+      "epoch": 2.338187702265372,
+      "grad_norm": 0.1659355702344565,
+      "learning_rate": 7.070362857394538e-06,
+      "loss": 0.1712,
+      "step": 2890
+    },
+    {
+      "epoch": 2.3462783171521036,
+      "grad_norm": 0.16020534539972092,
+      "learning_rate": 6.907141228606831e-06,
+      "loss": 0.165,
+      "step": 2900
+    },
+    {
+      "epoch": 2.354368932038835,
+      "grad_norm": 0.16872762980492875,
+      "learning_rate": 6.745523182354147e-06,
+      "loss": 0.1717,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3624595469255665,
+      "grad_norm": 0.15648717008884194,
+      "learning_rate": 6.585523042955233e-06,
+      "loss": 0.1668,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3705501618122975,
+      "grad_norm": 0.16263064803747318,
+      "learning_rate": 6.427154991332665e-06,
+      "loss": 0.1721,
+      "step": 2930
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.15947499101323748,
+      "learning_rate": 6.2704330637559315e-06,
+      "loss": 0.1709,
+      "step": 2940
+    },
+    {
+      "epoch": 2.3867313915857604,
+      "grad_norm": 0.16083163188925498,
+      "learning_rate": 6.115371150597413e-06,
+      "loss": 0.1702,
+      "step": 2950
+    },
+    {
+      "epoch": 2.394822006472492,
+      "grad_norm": 0.16542590461200707,
+      "learning_rate": 5.961982995101301e-06,
+      "loss": 0.1729,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4029126213592233,
+      "grad_norm": 0.16420989704643807,
+      "learning_rate": 5.810282192165442e-06,
+      "loss": 0.1722,
+      "step": 2970
+    },
+    {
+      "epoch": 2.411003236245955,
+      "grad_norm": 0.1551104913079497,
+      "learning_rate": 5.660282187136507e-06,
+      "loss": 0.1689,
+      "step": 2980
+    },
+    {
+      "epoch": 2.4190938511326863,
+      "grad_norm": 0.15940095271603197,
+      "learning_rate": 5.511996274618253e-06,
+      "loss": 0.1702,
+      "step": 2990
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.15033675616819664,
+      "learning_rate": 5.365437597293238e-06,
+      "loss": 0.1702,
+      "step": 3000
+    },
+    {
+      "epoch": 2.4352750809061487,
+      "grad_norm": 0.15579865123668477,
+      "learning_rate": 5.220619144757996e-06,
+      "loss": 0.1693,
+      "step": 3010
+    },
+    {
+      "epoch": 2.44336569579288,
+      "grad_norm": 0.16013909939251375,
+      "learning_rate": 5.077553752371708e-06,
+      "loss": 0.1678,
+      "step": 3020
+    },
+    {
+      "epoch": 2.4514563106796117,
+      "grad_norm": 0.16673936398632458,
+      "learning_rate": 4.936254100118656e-06,
+      "loss": 0.1709,
+      "step": 3030
+    },
+    {
+      "epoch": 2.459546925566343,
+      "grad_norm": 0.1605515108220787,
+      "learning_rate": 4.796732711484342e-06,
+      "loss": 0.1658,
+      "step": 3040
+    },
+    {
+      "epoch": 2.4676375404530746,
+      "grad_norm": 0.1585644224135113,
+      "learning_rate": 4.659001952345538e-06,
+      "loss": 0.1682,
+      "step": 3050
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.15466709640184034,
+      "learning_rate": 4.523074029874291e-06,
+      "loss": 0.1728,
+      "step": 3060
+    },
+    {
+      "epoch": 2.483818770226537,
+      "grad_norm": 0.15343018591614654,
+      "learning_rate": 4.388960991455998e-06,
+      "loss": 0.1692,
+      "step": 3070
+    },
+    {
+      "epoch": 2.4919093851132685,
+      "grad_norm": 0.14923408281267303,
+      "learning_rate": 4.256674723621621e-06,
+      "loss": 0.1706,
+      "step": 3080
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.16015551238015277,
+      "learning_rate": 4.126226950994211e-06,
+      "loss": 0.1708,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5080906148867315,
+      "grad_norm": 0.15550577660081535,
+      "learning_rate": 3.997629235249692e-06,
+      "loss": 0.1737,
+      "step": 3100
+    },
+    {
+      "epoch": 2.516181229773463,
+      "grad_norm": 0.16405843460813677,
+      "learning_rate": 3.870892974092197e-06,
+      "loss": 0.1677,
+      "step": 3110
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.15592864442189205,
+      "learning_rate": 3.7460294002438444e-06,
+      "loss": 0.1676,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5323624595469254,
+      "grad_norm": 0.15685374774957278,
+      "learning_rate": 3.6230495804491864e-06,
+      "loss": 0.174,
+      "step": 3130
+    },
+    {
+      "epoch": 2.540453074433657,
+      "grad_norm": 0.15046469900932222,
+      "learning_rate": 3.5019644144943576e-06,
+      "loss": 0.1723,
+      "step": 3140
+    },
+    {
+      "epoch": 2.5485436893203883,
+      "grad_norm": 0.15187030014871522,
+      "learning_rate": 3.382784634241015e-06,
+      "loss": 0.1679,
+      "step": 3150
+    },
+    {
+      "epoch": 2.55663430420712,
+      "grad_norm": 0.15370723807151057,
+      "learning_rate": 3.2655208026751816e-06,
+      "loss": 0.1707,
+      "step": 3160
+    },
+    {
+      "epoch": 2.5647249190938513,
+      "grad_norm": 0.15358348991682996,
+      "learning_rate": 3.150183312971014e-06,
+      "loss": 0.164,
+      "step": 3170
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.15495569068856682,
+      "learning_rate": 3.036782387569659e-06,
+      "loss": 0.1676,
+      "step": 3180
+    },
+    {
+      "epoch": 2.5809061488673137,
+      "grad_norm": 0.15152798447783355,
+      "learning_rate": 2.9253280772732595e-06,
+      "loss": 0.1668,
+      "step": 3190
+    },
+    {
+      "epoch": 2.588996763754045,
+      "grad_norm": 0.14795764100233588,
+      "learning_rate": 2.8158302603540965e-06,
+      "loss": 0.1677,
+      "step": 3200
+    },
+    {
+      "epoch": 2.5970873786407767,
+      "grad_norm": 0.1462435066429316,
+      "learning_rate": 2.708298641679105e-06,
+      "loss": 0.1634,
+      "step": 3210
+    },
+    {
+      "epoch": 2.605177993527508,
+      "grad_norm": 0.15564615414583508,
+      "learning_rate": 2.6027427518497153e-06,
+      "loss": 0.1675,
+      "step": 3220
+    },
+    {
+      "epoch": 2.6132686084142396,
+      "grad_norm": 0.15073879419059655,
+      "learning_rate": 2.49917194635714e-06,
+      "loss": 0.1652,
+      "step": 3230
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.1507824582262081,
+      "learning_rate": 2.397595404753225e-06,
+      "loss": 0.1708,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6294498381877025,
+      "grad_norm": 0.14447199827031953,
+      "learning_rate": 2.2980221298367995e-06,
+      "loss": 0.1718,
+      "step": 3250
+    },
+    {
+      "epoch": 2.6375404530744335,
+      "grad_norm": 0.15310920163624914,
+      "learning_rate": 2.2004609468558175e-06,
+      "loss": 0.1687,
+      "step": 3260
+    },
+    {
+      "epoch": 2.645631067961165,
+      "grad_norm": 0.1552961190816156,
+      "learning_rate": 2.1049205027251216e-06,
+      "loss": 0.168,
+      "step": 3270
+    },
+    {
+      "epoch": 2.6537216828478964,
+      "grad_norm": 0.1541563267360428,
+      "learning_rate": 2.0114092652600806e-06,
+      "loss": 0.1682,
+      "step": 3280
+    },
+    {
+      "epoch": 2.661812297734628,
+      "grad_norm": 0.1531794971521466,
+      "learning_rate": 1.919935522426081e-06,
+      "loss": 0.1688,
+      "step": 3290
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.1502769010065667,
+      "learning_rate": 1.8305073816039492e-06,
+      "loss": 0.17,
+      "step": 3300
+    },
+    {
+      "epoch": 2.6779935275080904,
+      "grad_norm": 0.14759059904949057,
+      "learning_rate": 1.7431327688714139e-06,
+      "loss": 0.1666,
+      "step": 3310
+    },
+    {
+      "epoch": 2.686084142394822,
+      "grad_norm": 0.14884028557553985,
+      "learning_rate": 1.6578194283005804e-06,
+      "loss": 0.1684,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6941747572815533,
+      "grad_norm": 0.15011522863824736,
+      "learning_rate": 1.5745749212715794e-06,
+      "loss": 0.1679,
+      "step": 3330
+    },
+    {
+      "epoch": 2.7022653721682848,
+      "grad_norm": 0.15248852381104064,
+      "learning_rate": 1.4934066258024182e-06,
+      "loss": 0.1741,
+      "step": 3340
+    },
+    {
+      "epoch": 2.7103559870550162,
+      "grad_norm": 0.14444830557078114,
+      "learning_rate": 1.4143217358950217e-06,
+      "loss": 0.168,
+      "step": 3350
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.16097353621052152,
+      "learning_rate": 1.3373272608976668e-06,
+      "loss": 0.1722,
+      "step": 3360
+    },
+    {
+      "epoch": 2.726537216828479,
+      "grad_norm": 0.15190902224222305,
+      "learning_rate": 1.2624300248836928e-06,
+      "loss": 0.1696,
+      "step": 3370
+    },
+    {
+      "epoch": 2.73462783171521,
+      "grad_norm": 0.15777837628961622,
+      "learning_rate": 1.1896366660467173e-06,
+      "loss": 0.168,
+      "step": 3380
+    },
+    {
+      "epoch": 2.7427184466019416,
+      "grad_norm": 0.14537830984259653,
+      "learning_rate": 1.1189536361122799e-06,
+      "loss": 0.1665,
+      "step": 3390
+    },
+    {
+      "epoch": 2.750809061488673,
+      "grad_norm": 0.14842465745819863,
+      "learning_rate": 1.0503871997660036e-06,
+      "loss": 0.1623,
+      "step": 3400
+    },
+    {
+      "epoch": 2.7588996763754046,
+      "grad_norm": 0.16097421255088173,
+      "learning_rate": 9.83943434098372e-07,
+      "loss": 0.166,
+      "step": 3410
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.1486212209344348,
+      "learning_rate": 9.196282280661023e-07,
+      "loss": 0.1688,
+      "step": 3420
+    },
+    {
+      "epoch": 2.775080906148867,
+      "grad_norm": 0.14198772068734017,
+      "learning_rate": 8.574472819702029e-07,
+      "loss": 0.1652,
+      "step": 3430
+    },
+    {
+      "epoch": 2.783171521035599,
+      "grad_norm": 0.1483955979101031,
+      "learning_rate": 7.974061069507571e-07,
+      "loss": 0.1679,
+      "step": 3440
+    },
+    {
+      "epoch": 2.79126213592233,
+      "grad_norm": 0.1510691708695613,
+      "learning_rate": 7.395100244984604e-07,
+      "loss": 0.1676,
+      "step": 3450
+    },
+    {
+      "epoch": 2.7993527508090614,
+      "grad_norm": 0.14675861790646366,
+      "learning_rate": 6.837641659829807e-07,
+      "loss": 0.1662,
+      "step": 3460
+    },
+    {
+      "epoch": 2.807443365695793,
+      "grad_norm": 0.1444417747058221,
+      "learning_rate": 6.301734721981533e-07,
+      "loss": 0.1669,
+      "step": 3470
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.14445462788403754,
+      "learning_rate": 5.787426929240808e-07,
+      "loss": 0.166,
+      "step": 3480
+    },
+    {
+      "epoch": 2.823624595469256,
+      "grad_norm": 0.1614165649662563,
+      "learning_rate": 5.294763865061558e-07,
+      "loss": 0.1683,
+      "step": 3490
+    },
+    {
+      "epoch": 2.831715210355987,
+      "grad_norm": 0.15711301471816222,
+      "learning_rate": 4.823789194510514e-07,
+      "loss": 0.1706,
+      "step": 3500
+    },
+    {
+      "epoch": 2.8398058252427183,
+      "grad_norm": 0.1453400612300062,
+      "learning_rate": 4.3745446603971064e-07,
+      "loss": 0.1688,
+      "step": 3510
+    },
+    {
+      "epoch": 2.8478964401294498,
+      "grad_norm": 0.15333438971393676,
+      "learning_rate": 3.947070079573872e-07,
+      "loss": 0.1638,
+      "step": 3520
+    },
+    {
+      "epoch": 2.855987055016181,
+      "grad_norm": 0.1472420355366777,
+      "learning_rate": 3.541403339407279e-07,
+      "loss": 0.171,
+      "step": 3530
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.1486881057392244,
+      "learning_rate": 3.1575803944199624e-07,
+      "loss": 0.1698,
+      "step": 3540
+    },
+    {
+      "epoch": 2.872168284789644,
+      "grad_norm": 0.140283471672274,
+      "learning_rate": 2.7956352631038906e-07,
+      "loss": 0.1659,
+      "step": 3550
+    },
+    {
+      "epoch": 2.8802588996763756,
+      "grad_norm": 0.14768247651236527,
+      "learning_rate": 2.4556000249054133e-07,
+      "loss": 0.1705,
+      "step": 3560
+    },
+    {
+      "epoch": 2.8883495145631066,
+      "grad_norm": 0.14227761591758845,
+      "learning_rate": 2.1375048173818412e-07,
+      "loss": 0.1676,
+      "step": 3570
+    },
+    {
+      "epoch": 2.896440129449838,
+      "grad_norm": 0.1397094378867443,
+      "learning_rate": 1.8413778335305e-07,
+      "loss": 0.1639,
+      "step": 3580
+    },
+    {
+      "epoch": 2.9045307443365695,
+      "grad_norm": 0.14544808567516948,
+      "learning_rate": 1.567245319290006e-07,
+      "loss": 0.169,
+      "step": 3590
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.14376315892226174,
+      "learning_rate": 1.315131571213879e-07,
+      "loss": 0.163,
+      "step": 3600
+    },
+    {
+      "epoch": 2.9207119741100325,
+      "grad_norm": 0.14749344179261822,
+      "learning_rate": 1.0850589343172624e-07,
+      "loss": 0.1669,
+      "step": 3610
+    },
+    {
+      "epoch": 2.9288025889967635,
+      "grad_norm": 0.1452354644803129,
+      "learning_rate": 8.770478000964532e-08,
+      "loss": 0.1684,
+      "step": 3620
+    },
+    {
+      "epoch": 2.9368932038834954,
+      "grad_norm": 0.14097833660885037,
+      "learning_rate": 6.911166047215145e-08,
+      "loss": 0.1694,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9449838187702264,
+      "grad_norm": 0.14528957043064877,
+      "learning_rate": 5.272818274023872e-08,
+      "loss": 0.1675,
+      "step": 3640
+    },
+    {
+      "epoch": 2.953074433656958,
+      "grad_norm": 0.14549261409943784,
+      "learning_rate": 3.855579889282257e-08,
+      "loss": 0.1679,
+      "step": 3650
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.14557934166615819,
+      "learning_rate": 2.6595765038045507e-08,
+      "loss": 0.1677,
+      "step": 3660
+    },
+    {
+      "epoch": 2.969255663430421,
+      "grad_norm": 0.1445108267077229,
+      "learning_rate": 1.6849141201946693e-08,
+      "loss": 0.165,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9773462783171523,
+      "grad_norm": 0.1425028501060893,
+      "learning_rate": 9.316791234506572e-09,
+      "loss": 0.1713,
+      "step": 3680
+    },
+    {
+      "epoch": 2.9854368932038833,
+      "grad_norm": 0.13458537716318286,
+      "learning_rate": 3.999382733096968e-09,
+      "loss": 0.1665,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9935275080906147,
+      "grad_norm": 0.13982192859442388,
+      "learning_rate": 8.973869832895609e-10,
+      "loss": 0.1686,
+      "step": 3700
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 3708,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.26626862104576e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-3708/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efde72064ee18fa35cca6fd639189f25db3e52f3bbc800f20c7c6eb49dea86ca
+size 7736

config.json

@@ -0,0 +1,68 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 12288,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 32768,
+  "max_window_layers": 36,
+  "model_type": "qwen3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.55.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048,
+  "transformers_version": "4.55.0"
+}

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53096ac7bb0e3a0ec6cecc4d902d62f520e74ac30e63b423214283a814c0776f
+size 4902257696

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e02590c760b06bef8b03c37ab5156352f2a1c7f5862d1aed747ea0a1a5bc6f4d
+size 4915960368

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca5be3fa1663f2990e1bdc3cc029546d69e03538f02e9faa9a049b93261cb3a5
+size 4983068496

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99a72c2cad443584179996d1d7fa5d512319232128449148213f31e469b660e7
+size 1580230264

model.safetensors.index.json

@@ -0,0 +1,407 @@
+{
+  "metadata": {
+    "total_parameters": 308224,
+    "total_size": 16381470720
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00004-of-00004.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.22.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.k_norm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.q_norm.weight": "model-00004-of-00004.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
+    "model.layers.9.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
+    "model.norm.weight": "model-00004-of-00004.safetensors"
+  }
+}

runs/Nov17_00-00-46_SH-IDCA1404-10-140-54-34/events.out.tfevents.1763309004.SH-IDCA1404-10-140-54-34.13603.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fe600e78f77c7b5af556baa9a751712d2c98a617716b0708d81135a469a2000
+size 84807

special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null,
+  "chat_template": "{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message + '\n' }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ 'Human: ' + content + '\nAssistant:' }}{% elif message['role'] == 'assistant' %}{{ content + '<|endoftext|>' + '\n' }}{% endif %}{% endfor %}"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 7.26626862104576e+18,
+    "train_loss": 0.24862830043482034,
+    "train_runtime": 257107.0797,
+    "train_samples_per_second": 0.461,
+    "train_steps_per_second": 0.014
+}

trainer_log.jsonl

@@ -0,0 +1,371 @@
+{"current_steps": 10, "total_steps": 3708, "loss": 2.5582, "lr": 1.2129380053908356e-06, "epoch": 0.008090614886731391, "percentage": 0.27, "elapsed_time": "0:11:37", "remaining_time": "2 days, 23:36:07"}
+{"current_steps": 20, "total_steps": 3708, "loss": 2.2025, "lr": 2.560646900269542e-06, "epoch": 0.016181229773462782, "percentage": 0.54, "elapsed_time": "0:23:09", "remaining_time": "2 days, 23:09:46"}
+{"current_steps": 30, "total_steps": 3708, "loss": 1.668, "lr": 3.908355795148248e-06, "epoch": 0.024271844660194174, "percentage": 0.81, "elapsed_time": "0:34:41", "remaining_time": "2 days, 22:53:49"}
+{"current_steps": 40, "total_steps": 3708, "loss": 1.1266, "lr": 5.256064690026954e-06, "epoch": 0.032362459546925564, "percentage": 1.08, "elapsed_time": "0:46:14", "remaining_time": "2 days, 22:40:20"}
+{"current_steps": 50, "total_steps": 3708, "loss": 0.7799, "lr": 6.60377358490566e-06, "epoch": 0.040453074433656956, "percentage": 1.35, "elapsed_time": "0:57:46", "remaining_time": "2 days, 22:27:08"}
+{"current_steps": 60, "total_steps": 3708, "loss": 0.5907, "lr": 7.951482479784367e-06, "epoch": 0.04854368932038835, "percentage": 1.62, "elapsed_time": "1:09:19", "remaining_time": "2 days, 22:14:46"}
+{"current_steps": 70, "total_steps": 3708, "loss": 0.4977, "lr": 9.299191374663074e-06, "epoch": 0.05663430420711974, "percentage": 1.89, "elapsed_time": "1:20:52", "remaining_time": "2 days, 22:02:46"}
+{"current_steps": 80, "total_steps": 3708, "loss": 0.4728, "lr": 1.0646900269541779e-05, "epoch": 0.06472491909385113, "percentage": 2.16, "elapsed_time": "1:32:24", "remaining_time": "2 days, 21:50:58"}
+{"current_steps": 90, "total_steps": 3708, "loss": 0.446, "lr": 1.1994609164420486e-05, "epoch": 0.07281553398058252, "percentage": 2.43, "elapsed_time": "1:43:57", "remaining_time": "2 days, 21:39:12"}
+{"current_steps": 100, "total_steps": 3708, "loss": 0.4164, "lr": 1.3342318059299191e-05, "epoch": 0.08090614886731391, "percentage": 2.7, "elapsed_time": "1:55:30", "remaining_time": "2 days, 21:27:39"}
+{"current_steps": 110, "total_steps": 3708, "loss": 0.412, "lr": 1.4690026954177898e-05, "epoch": 0.0889967637540453, "percentage": 2.97, "elapsed_time": "2:07:03", "remaining_time": "2 days, 21:15:52"}
+{"current_steps": 120, "total_steps": 3708, "loss": 0.3975, "lr": 1.6037735849056604e-05, "epoch": 0.0970873786407767, "percentage": 3.24, "elapsed_time": "2:18:35", "remaining_time": "2 days, 21:04:08"}
+{"current_steps": 130, "total_steps": 3708, "loss": 0.3857, "lr": 1.738544474393531e-05, "epoch": 0.10517799352750809, "percentage": 3.51, "elapsed_time": "2:30:08", "remaining_time": "2 days, 20:52:31"}
+{"current_steps": 140, "total_steps": 3708, "loss": 0.3837, "lr": 1.8733153638814018e-05, "epoch": 0.11326860841423948, "percentage": 3.78, "elapsed_time": "2:41:41", "remaining_time": "2 days, 20:40:54"}
+{"current_steps": 150, "total_steps": 3708, "loss": 0.3721, "lr": 2.0080862533692725e-05, "epoch": 0.12135922330097088, "percentage": 4.05, "elapsed_time": "2:53:14", "remaining_time": "2 days, 20:29:13"}
+{"current_steps": 160, "total_steps": 3708, "loss": 0.3734, "lr": 2.1428571428571428e-05, "epoch": 0.12944983818770225, "percentage": 4.31, "elapsed_time": "3:04:47", "remaining_time": "2 days, 20:17:34"}
+{"current_steps": 170, "total_steps": 3708, "loss": 0.365, "lr": 2.2776280323450135e-05, "epoch": 0.13754045307443366, "percentage": 4.58, "elapsed_time": "3:16:19", "remaining_time": "2 days, 20:05:54"}
+{"current_steps": 180, "total_steps": 3708, "loss": 0.3631, "lr": 2.4123989218328842e-05, "epoch": 0.14563106796116504, "percentage": 4.85, "elapsed_time": "3:27:52", "remaining_time": "2 days, 19:54:20"}
+{"current_steps": 190, "total_steps": 3708, "loss": 0.3509, "lr": 2.547169811320755e-05, "epoch": 0.15372168284789645, "percentage": 5.12, "elapsed_time": "3:39:25", "remaining_time": "2 days, 19:42:47"}
+{"current_steps": 200, "total_steps": 3708, "loss": 0.3474, "lr": 2.6819407008086256e-05, "epoch": 0.16181229773462782, "percentage": 5.39, "elapsed_time": "3:50:57", "remaining_time": "2 days, 19:31:06"}
+{"current_steps": 210, "total_steps": 3708, "loss": 0.3439, "lr": 2.8167115902964963e-05, "epoch": 0.16990291262135923, "percentage": 5.66, "elapsed_time": "4:02:30", "remaining_time": "2 days, 19:19:29"}
+{"current_steps": 220, "total_steps": 3708, "loss": 0.3477, "lr": 2.9514824797843667e-05, "epoch": 0.1779935275080906, "percentage": 5.93, "elapsed_time": "4:14:03", "remaining_time": "2 days, 19:07:54"}
+{"current_steps": 230, "total_steps": 3708, "loss": 0.3424, "lr": 3.086253369272237e-05, "epoch": 0.18608414239482202, "percentage": 6.2, "elapsed_time": "4:25:35", "remaining_time": "2 days, 18:56:18"}
+{"current_steps": 240, "total_steps": 3708, "loss": 0.3432, "lr": 3.221024258760108e-05, "epoch": 0.1941747572815534, "percentage": 6.47, "elapsed_time": "4:37:08", "remaining_time": "2 days, 18:44:45"}
+{"current_steps": 250, "total_steps": 3708, "loss": 0.3378, "lr": 3.355795148247979e-05, "epoch": 0.2022653721682848, "percentage": 6.74, "elapsed_time": "4:48:41", "remaining_time": "2 days, 18:33:12"}
+{"current_steps": 260, "total_steps": 3708, "loss": 0.34, "lr": 3.490566037735849e-05, "epoch": 0.21035598705501618, "percentage": 7.01, "elapsed_time": "5:00:14", "remaining_time": "2 days, 18:21:34"}
+{"current_steps": 270, "total_steps": 3708, "loss": 0.339, "lr": 3.62533692722372e-05, "epoch": 0.21844660194174756, "percentage": 7.28, "elapsed_time": "5:11:46", "remaining_time": "2 days, 18:09:59"}
+{"current_steps": 280, "total_steps": 3708, "loss": 0.3284, "lr": 3.76010781671159e-05, "epoch": 0.22653721682847897, "percentage": 7.55, "elapsed_time": "5:23:19", "remaining_time": "2 days, 17:58:24"}
+{"current_steps": 290, "total_steps": 3708, "loss": 0.3291, "lr": 3.894878706199461e-05, "epoch": 0.23462783171521034, "percentage": 7.82, "elapsed_time": "5:34:52", "remaining_time": "2 days, 17:46:50"}
+{"current_steps": 300, "total_steps": 3708, "loss": 0.3206, "lr": 4.0296495956873316e-05, "epoch": 0.24271844660194175, "percentage": 8.09, "elapsed_time": "5:46:24", "remaining_time": "2 days, 17:35:13"}
+{"current_steps": 310, "total_steps": 3708, "loss": 0.3276, "lr": 4.164420485175202e-05, "epoch": 0.25080906148867316, "percentage": 8.36, "elapsed_time": "5:57:57", "remaining_time": "2 days, 17:23:39"}
+{"current_steps": 320, "total_steps": 3708, "loss": 0.3258, "lr": 4.299191374663073e-05, "epoch": 0.2588996763754045, "percentage": 8.63, "elapsed_time": "6:09:30", "remaining_time": "2 days, 17:12:08"}
+{"current_steps": 330, "total_steps": 3708, "loss": 0.3227, "lr": 4.433962264150944e-05, "epoch": 0.2669902912621359, "percentage": 8.9, "elapsed_time": "6:21:02", "remaining_time": "2 days, 17:00:31"}
+{"current_steps": 340, "total_steps": 3708, "loss": 0.3196, "lr": 4.5687331536388144e-05, "epoch": 0.2750809061488673, "percentage": 9.17, "elapsed_time": "6:32:35", "remaining_time": "2 days, 16:48:58"}
+{"current_steps": 350, "total_steps": 3708, "loss": 0.3191, "lr": 4.703504043126685e-05, "epoch": 0.28317152103559873, "percentage": 9.44, "elapsed_time": "6:44:08", "remaining_time": "2 days, 16:37:26"}
+{"current_steps": 360, "total_steps": 3708, "loss": 0.3255, "lr": 4.838274932614555e-05, "epoch": 0.2912621359223301, "percentage": 9.71, "elapsed_time": "6:55:41", "remaining_time": "2 days, 16:25:55"}
+{"current_steps": 370, "total_steps": 3708, "loss": 0.3119, "lr": 4.973045822102426e-05, "epoch": 0.2993527508090615, "percentage": 9.98, "elapsed_time": "7:07:14", "remaining_time": "2 days, 16:14:21"}
+{"current_steps": 380, "total_steps": 3708, "loss": 0.3191, "lr": 4.9999290952604396e-05, "epoch": 0.3074433656957929, "percentage": 10.25, "elapsed_time": "7:18:47", "remaining_time": "2 days, 16:02:51"}
+{"current_steps": 390, "total_steps": 3708, "loss": 0.3098, "lr": 4.9996410516491115e-05, "epoch": 0.3155339805825243, "percentage": 10.52, "elapsed_time": "7:30:20", "remaining_time": "2 days, 15:51:17"}
+{"current_steps": 400, "total_steps": 3708, "loss": 0.3096, "lr": 4.9991314631296585e-05, "epoch": 0.32362459546925565, "percentage": 10.79, "elapsed_time": "7:41:52", "remaining_time": "2 days, 15:39:44"}
+{"current_steps": 410, "total_steps": 3708, "loss": 0.3095, "lr": 4.9984003748672604e-05, "epoch": 0.33171521035598706, "percentage": 11.06, "elapsed_time": "7:53:25", "remaining_time": "2 days, 15:28:10"}
+{"current_steps": 420, "total_steps": 3708, "loss": 0.3076, "lr": 4.997447851658774e-05, "epoch": 0.33980582524271846, "percentage": 11.33, "elapsed_time": "8:04:57", "remaining_time": "2 days, 15:16:33"}
+{"current_steps": 430, "total_steps": 3708, "loss": 0.3067, "lr": 4.9962739779269887e-05, "epoch": 0.3478964401294498, "percentage": 11.6, "elapsed_time": "8:16:30", "remaining_time": "2 days, 15:05:01"}
+{"current_steps": 440, "total_steps": 3708, "loss": 0.2989, "lr": 4.9948788577131414e-05, "epoch": 0.3559870550161812, "percentage": 11.87, "elapsed_time": "8:28:03", "remaining_time": "2 days, 14:53:29"}
+{"current_steps": 450, "total_steps": 3708, "loss": 0.3018, "lr": 4.993262614667696e-05, "epoch": 0.3640776699029126, "percentage": 12.14, "elapsed_time": "8:39:36", "remaining_time": "2 days, 14:41:56"}
+{"current_steps": 460, "total_steps": 3708, "loss": 0.3048, "lr": 4.9914253920393884e-05, "epoch": 0.37216828478964403, "percentage": 12.41, "elapsed_time": "8:51:09", "remaining_time": "2 days, 14:30:23"}
+{"current_steps": 470, "total_steps": 3708, "loss": 0.3025, "lr": 4.9893673526625265e-05, "epoch": 0.3802588996763754, "percentage": 12.68, "elapsed_time": "9:02:41", "remaining_time": "2 days, 14:18:50"}
+{"current_steps": 480, "total_steps": 3708, "loss": 0.2997, "lr": 4.987088678942555e-05, "epoch": 0.3883495145631068, "percentage": 12.94, "elapsed_time": "9:14:14", "remaining_time": "2 days, 14:07:19"}
+{"current_steps": 490, "total_steps": 3708, "loss": 0.2981, "lr": 4.984589572839897e-05, "epoch": 0.3964401294498382, "percentage": 13.21, "elapsed_time": "9:25:47", "remaining_time": "2 days, 13:55:47"}
+{"current_steps": 500, "total_steps": 3708, "loss": 0.2994, "lr": 4.9818702558520485e-05, "epoch": 0.4045307443365696, "percentage": 13.48, "elapsed_time": "9:37:20", "remaining_time": "2 days, 13:44:14"}
+{"current_steps": 510, "total_steps": 3708, "loss": 0.3027, "lr": 4.978930968993946e-05, "epoch": 0.41262135922330095, "percentage": 13.75, "elapsed_time": "9:48:53", "remaining_time": "2 days, 13:32:41"}
+{"current_steps": 520, "total_steps": 3708, "loss": 0.2995, "lr": 4.9757719727766085e-05, "epoch": 0.42071197411003236, "percentage": 14.02, "elapsed_time": "10:00:26", "remaining_time": "2 days, 13:21:10"}
+{"current_steps": 530, "total_steps": 3708, "loss": 0.2968, "lr": 4.972393547184046e-05, "epoch": 0.42880258899676377, "percentage": 14.29, "elapsed_time": "10:11:59", "remaining_time": "2 days, 13:09:36"}
+{"current_steps": 540, "total_steps": 3708, "loss": 0.2951, "lr": 4.968795991648446e-05, "epoch": 0.4368932038834951, "percentage": 14.56, "elapsed_time": "10:23:31", "remaining_time": "2 days, 12:58:01"}
+{"current_steps": 550, "total_steps": 3708, "loss": 0.2842, "lr": 4.9649796250236344e-05, "epoch": 0.4449838187702265, "percentage": 14.83, "elapsed_time": "10:35:04", "remaining_time": "2 days, 12:46:28"}
+{"current_steps": 560, "total_steps": 3708, "loss": 0.2862, "lr": 4.960944785556814e-05, "epoch": 0.45307443365695793, "percentage": 15.1, "elapsed_time": "10:46:37", "remaining_time": "2 days, 12:34:59"}
+{"current_steps": 570, "total_steps": 3708, "loss": 0.2925, "lr": 4.956691830858585e-05, "epoch": 0.46116504854368934, "percentage": 15.37, "elapsed_time": "10:58:10", "remaining_time": "2 days, 12:23:26"}
+{"current_steps": 580, "total_steps": 3708, "loss": 0.2884, "lr": 4.952221137871252e-05, "epoch": 0.4692556634304207, "percentage": 15.64, "elapsed_time": "11:09:43", "remaining_time": "2 days, 12:11:52"}
+{"current_steps": 590, "total_steps": 3708, "loss": 0.2937, "lr": 4.947533102835413e-05, "epoch": 0.4773462783171521, "percentage": 15.91, "elapsed_time": "11:21:16", "remaining_time": "2 days, 12:00:20"}
+{"current_steps": 600, "total_steps": 3708, "loss": 0.2888, "lr": 4.942628141254843e-05, "epoch": 0.4854368932038835, "percentage": 16.18, "elapsed_time": "11:32:49", "remaining_time": "2 days, 11:48:49"}
+{"current_steps": 610, "total_steps": 3708, "loss": 0.2839, "lr": 4.937506687859666e-05, "epoch": 0.4935275080906149, "percentage": 16.45, "elapsed_time": "11:44:22", "remaining_time": "2 days, 11:37:16"}
+{"current_steps": 620, "total_steps": 3708, "loss": 0.2926, "lr": 4.932169196567824e-05, "epoch": 0.5016181229773463, "percentage": 16.72, "elapsed_time": "11:55:54", "remaining_time": "2 days, 11:25:43"}
+{"current_steps": 630, "total_steps": 3708, "loss": 0.2914, "lr": 4.9266161404448454e-05, "epoch": 0.5097087378640777, "percentage": 16.99, "elapsed_time": "12:07:27", "remaining_time": "2 days, 11:14:10"}
+{"current_steps": 640, "total_steps": 3708, "loss": 0.2844, "lr": 4.920848011661919e-05, "epoch": 0.517799352750809, "percentage": 17.26, "elapsed_time": "12:19:00", "remaining_time": "2 days, 11:02:37"}
+{"current_steps": 650, "total_steps": 3708, "loss": 0.2838, "lr": 4.914865321452274e-05, "epoch": 0.5258899676375405, "percentage": 17.53, "elapsed_time": "12:30:33", "remaining_time": "2 days, 10:51:05"}
+{"current_steps": 660, "total_steps": 3708, "loss": 0.2853, "lr": 4.908668600065862e-05, "epoch": 0.5339805825242718, "percentage": 17.8, "elapsed_time": "12:42:06", "remaining_time": "2 days, 10:39:31"}
+{"current_steps": 670, "total_steps": 3708, "loss": 0.2844, "lr": 4.90225839672237e-05, "epoch": 0.5420711974110033, "percentage": 18.07, "elapsed_time": "12:53:38", "remaining_time": "2 days, 10:27:57"}
+{"current_steps": 680, "total_steps": 3708, "loss": 0.2912, "lr": 4.8956352795625325e-05, "epoch": 0.5501618122977346, "percentage": 18.34, "elapsed_time": "13:05:11", "remaining_time": "2 days, 10:16:25"}
+{"current_steps": 690, "total_steps": 3708, "loss": 0.2826, "lr": 4.8887998355977886e-05, "epoch": 0.558252427184466, "percentage": 18.61, "elapsed_time": "13:16:44", "remaining_time": "2 days, 10:04:52"}
+{"current_steps": 700, "total_steps": 3708, "loss": 0.2799, "lr": 4.881752670658244e-05, "epoch": 0.5663430420711975, "percentage": 18.88, "elapsed_time": "13:28:17", "remaining_time": "2 days, 9:53:20"}
+{"current_steps": 710, "total_steps": 3708, "loss": 0.2713, "lr": 4.87449440933898e-05, "epoch": 0.5744336569579288, "percentage": 19.15, "elapsed_time": "13:39:50", "remaining_time": "2 days, 9:41:48"}
+{"current_steps": 720, "total_steps": 3708, "loss": 0.2813, "lr": 4.867025694944698e-05, "epoch": 0.5825242718446602, "percentage": 19.42, "elapsed_time": "13:51:23", "remaining_time": "2 days, 9:30:15"}
+{"current_steps": 730, "total_steps": 3708, "loss": 0.28, "lr": 4.859347189432699e-05, "epoch": 0.5906148867313916, "percentage": 19.69, "elapsed_time": "14:02:56", "remaining_time": "2 days, 9:18:43"}
+{"current_steps": 740, "total_steps": 3708, "loss": 0.2709, "lr": 4.8514595733542144e-05, "epoch": 0.598705501618123, "percentage": 19.96, "elapsed_time": "14:14:29", "remaining_time": "2 days, 9:07:10"}
+{"current_steps": 750, "total_steps": 3708, "loss": 0.2814, "lr": 4.8433635457940915e-05, "epoch": 0.6067961165048543, "percentage": 20.23, "elapsed_time": "14:26:01", "remaining_time": "2 days, 8:55:36"}
+{"current_steps": 760, "total_steps": 3708, "loss": 0.2776, "lr": 4.8350598243088283e-05, "epoch": 0.6148867313915858, "percentage": 20.5, "elapsed_time": "14:37:34", "remaining_time": "2 days, 8:44:04"}
+{"current_steps": 770, "total_steps": 3708, "loss": 0.2767, "lr": 4.8265491448629804e-05, "epoch": 0.6229773462783171, "percentage": 20.77, "elapsed_time": "14:49:07", "remaining_time": "2 days, 8:32:30"}
+{"current_steps": 780, "total_steps": 3708, "loss": 0.2772, "lr": 4.817832261763928e-05, "epoch": 0.6310679611650486, "percentage": 21.04, "elapsed_time": "15:00:40", "remaining_time": "2 days, 8:20:59"}
+{"current_steps": 790, "total_steps": 3708, "loss": 0.276, "lr": 4.8089099475950257e-05, "epoch": 0.63915857605178, "percentage": 21.31, "elapsed_time": "15:12:13", "remaining_time": "2 days, 8:09:27"}
+{"current_steps": 800, "total_steps": 3708, "loss": 0.2765, "lr": 4.7997829931471225e-05, "epoch": 0.6472491909385113, "percentage": 21.57, "elapsed_time": "15:23:46", "remaining_time": "2 days, 7:57:55"}
+{"current_steps": 810, "total_steps": 3708, "loss": 0.2778, "lr": 4.7904522073484786e-05, "epoch": 0.6553398058252428, "percentage": 21.84, "elapsed_time": "15:35:19", "remaining_time": "2 days, 7:46:22"}
+{"current_steps": 820, "total_steps": 3708, "loss": 0.2775, "lr": 4.780918417193065e-05, "epoch": 0.6634304207119741, "percentage": 22.11, "elapsed_time": "15:46:51", "remaining_time": "2 days, 7:34:48"}
+{"current_steps": 830, "total_steps": 3708, "loss": 0.2733, "lr": 4.7711824676672726e-05, "epoch": 0.6715210355987055, "percentage": 22.38, "elapsed_time": "15:58:24", "remaining_time": "2 days, 7:23:15"}
+{"current_steps": 840, "total_steps": 3708, "loss": 0.2746, "lr": 4.76124522167501e-05, "epoch": 0.6796116504854369, "percentage": 22.65, "elapsed_time": "16:09:57", "remaining_time": "2 days, 7:11:42"}
+{"current_steps": 850, "total_steps": 3708, "loss": 0.2745, "lr": 4.751107559961238e-05, "epoch": 0.6877022653721683, "percentage": 22.92, "elapsed_time": "16:21:30", "remaining_time": "2 days, 7:00:09"}
+{"current_steps": 860, "total_steps": 3708, "loss": 0.275, "lr": 4.740770381033894e-05, "epoch": 0.6957928802588996, "percentage": 23.19, "elapsed_time": "16:33:02", "remaining_time": "2 days, 6:48:35"}
+{"current_steps": 870, "total_steps": 3708, "loss": 0.2716, "lr": 4.730234601084268e-05, "epoch": 0.7038834951456311, "percentage": 23.46, "elapsed_time": "16:44:35", "remaining_time": "2 days, 6:37:03"}
+{"current_steps": 880, "total_steps": 3708, "loss": 0.2673, "lr": 4.719501153905793e-05, "epoch": 0.7119741100323624, "percentage": 23.73, "elapsed_time": "16:56:08", "remaining_time": "2 days, 6:25:31"}
+{"current_steps": 890, "total_steps": 3708, "loss": 0.2731, "lr": 4.7085709908112866e-05, "epoch": 0.7200647249190939, "percentage": 24.0, "elapsed_time": "17:07:41", "remaining_time": "2 days, 6:13:58"}
+{"current_steps": 900, "total_steps": 3708, "loss": 0.2639, "lr": 4.6974450805486305e-05, "epoch": 0.7281553398058253, "percentage": 24.27, "elapsed_time": "17:19:14", "remaining_time": "2 days, 6:02:26"}
+{"current_steps": 910, "total_steps": 3708, "loss": 0.2715, "lr": 4.686124409214917e-05, "epoch": 0.7362459546925566, "percentage": 24.54, "elapsed_time": "17:30:47", "remaining_time": "2 days, 5:50:53"}
+{"current_steps": 920, "total_steps": 3708, "loss": 0.2712, "lr": 4.674609980169042e-05, "epoch": 0.7443365695792881, "percentage": 24.81, "elapsed_time": "17:42:20", "remaining_time": "2 days, 5:39:20"}
+{"current_steps": 930, "total_steps": 3708, "loss": 0.272, "lr": 4.662902813942784e-05, "epoch": 0.7524271844660194, "percentage": 25.08, "elapsed_time": "17:53:52", "remaining_time": "2 days, 5:27:47"}
+{"current_steps": 940, "total_steps": 3708, "loss": 0.2689, "lr": 4.651003948150349e-05, "epoch": 0.7605177993527508, "percentage": 25.35, "elapsed_time": "18:05:25", "remaining_time": "2 days, 5:16:14"}
+{"current_steps": 950, "total_steps": 3708, "loss": 0.2693, "lr": 4.638914437396408e-05, "epoch": 0.7686084142394822, "percentage": 25.62, "elapsed_time": "18:16:58", "remaining_time": "2 days, 5:04:41"}
+{"current_steps": 960, "total_steps": 3708, "loss": 0.2659, "lr": 4.626635353182626e-05, "epoch": 0.7766990291262136, "percentage": 25.89, "elapsed_time": "18:28:31", "remaining_time": "2 days, 4:53:08"}
+{"current_steps": 970, "total_steps": 3708, "loss": 0.2753, "lr": 4.614167783812694e-05, "epoch": 0.7847896440129449, "percentage": 26.16, "elapsed_time": "18:40:03", "remaining_time": "2 days, 4:41:35"}
+{"current_steps": 980, "total_steps": 3708, "loss": 0.272, "lr": 4.601512834295874e-05, "epoch": 0.7928802588996764, "percentage": 26.43, "elapsed_time": "18:51:36", "remaining_time": "2 days, 4:30:02"}
+{"current_steps": 990, "total_steps": 3708, "loss": 0.2631, "lr": 4.588671626249057e-05, "epoch": 0.8009708737864077, "percentage": 26.7, "elapsed_time": "19:03:09", "remaining_time": "2 days, 4:18:29"}
+{"current_steps": 1000, "total_steps": 3708, "loss": 0.2609, "lr": 4.5756452977973585e-05, "epoch": 0.8090614886731392, "percentage": 26.97, "elapsed_time": "19:14:42", "remaining_time": "2 days, 4:06:56"}
+{"current_steps": 1010, "total_steps": 3708, "loss": 0.2651, "lr": 4.56243500347324e-05, "epoch": 0.8171521035598706, "percentage": 27.24, "elapsed_time": "19:26:14", "remaining_time": "2 days, 3:55:23"}
+{"current_steps": 1020, "total_steps": 3708, "loss": 0.2687, "lr": 4.549041914114188e-05, "epoch": 0.8252427184466019, "percentage": 27.51, "elapsed_time": "19:37:47", "remaining_time": "2 days, 3:43:50"}
+{"current_steps": 1030, "total_steps": 3708, "loss": 0.2666, "lr": 4.535467216758936e-05, "epoch": 0.8333333333333334, "percentage": 27.78, "elapsed_time": "19:49:20", "remaining_time": "2 days, 3:32:17"}
+{"current_steps": 1040, "total_steps": 3708, "loss": 0.2631, "lr": 4.5217121145422616e-05, "epoch": 0.8414239482200647, "percentage": 28.05, "elapsed_time": "20:00:53", "remaining_time": "2 days, 3:20:44"}
+{"current_steps": 1050, "total_steps": 3708, "loss": 0.2632, "lr": 4.5077778265883477e-05, "epoch": 0.8495145631067961, "percentage": 28.32, "elapsed_time": "20:12:26", "remaining_time": "2 days, 3:09:12"}
+{"current_steps": 1060, "total_steps": 3708, "loss": 0.2634, "lr": 4.4936655879027336e-05, "epoch": 0.8576051779935275, "percentage": 28.59, "elapsed_time": "20:23:59", "remaining_time": "2 days, 2:57:39"}
+{"current_steps": 1070, "total_steps": 3708, "loss": 0.266, "lr": 4.479376649262855e-05, "epoch": 0.8656957928802589, "percentage": 28.86, "elapsed_time": "20:35:32", "remaining_time": "2 days, 2:46:07"}
+{"current_steps": 1080, "total_steps": 3708, "loss": 0.2658, "lr": 4.464912277107185e-05, "epoch": 0.8737864077669902, "percentage": 29.13, "elapsed_time": "20:47:04", "remaining_time": "2 days, 2:34:34"}
+{"current_steps": 1090, "total_steps": 3708, "loss": 0.2653, "lr": 4.450273753422992e-05, "epoch": 0.8818770226537217, "percentage": 29.4, "elapsed_time": "20:58:38", "remaining_time": "2 days, 2:23:01"}
+{"current_steps": 1100, "total_steps": 3708, "loss": 0.2651, "lr": 4.435462375632711e-05, "epoch": 0.889967637540453, "percentage": 29.67, "elapsed_time": "21:10:10", "remaining_time": "2 days, 2:11:29"}
+{"current_steps": 1110, "total_steps": 3708, "loss": 0.2621, "lr": 4.420479456478957e-05, "epoch": 0.8980582524271845, "percentage": 29.94, "elapsed_time": "21:21:44", "remaining_time": "2 days, 1:59:57"}
+{"current_steps": 1120, "total_steps": 3708, "loss": 0.2635, "lr": 4.405326323908178e-05, "epoch": 0.9061488673139159, "percentage": 30.2, "elapsed_time": "21:33:17", "remaining_time": "2 days, 1:48:24"}
+{"current_steps": 1130, "total_steps": 3708, "loss": 0.2639, "lr": 4.390004320952947e-05, "epoch": 0.9142394822006472, "percentage": 30.47, "elapsed_time": "21:44:50", "remaining_time": "2 days, 1:36:52"}
+{"current_steps": 1140, "total_steps": 3708, "loss": 0.2667, "lr": 4.374514805612942e-05, "epoch": 0.9223300970873787, "percentage": 30.74, "elapsed_time": "21:56:23", "remaining_time": "2 days, 1:25:20"}
+{"current_steps": 1150, "total_steps": 3708, "loss": 0.2592, "lr": 4.358859150734576e-05, "epoch": 0.93042071197411, "percentage": 31.01, "elapsed_time": "22:07:55", "remaining_time": "2 days, 1:13:47"}
+{"current_steps": 1160, "total_steps": 3708, "loss": 0.2602, "lr": 4.343038743889324e-05, "epoch": 0.9385113268608414, "percentage": 31.28, "elapsed_time": "22:19:29", "remaining_time": "2 days, 1:02:14"}
+{"current_steps": 1170, "total_steps": 3708, "loss": 0.26, "lr": 4.3270549872507415e-05, "epoch": 0.9466019417475728, "percentage": 31.55, "elapsed_time": "22:31:01", "remaining_time": "2 days, 0:50:42"}
+{"current_steps": 1180, "total_steps": 3708, "loss": 0.2591, "lr": 4.3109092974701895e-05, "epoch": 0.9546925566343042, "percentage": 31.82, "elapsed_time": "22:42:34", "remaining_time": "2 days, 0:39:08"}
+{"current_steps": 1190, "total_steps": 3708, "loss": 0.26, "lr": 4.2946031055512733e-05, "epoch": 0.9627831715210357, "percentage": 32.09, "elapsed_time": "22:54:07", "remaining_time": "2 days, 0:27:36"}
+{"current_steps": 1200, "total_steps": 3708, "loss": 0.2633, "lr": 4.2781378567230145e-05, "epoch": 0.970873786407767, "percentage": 32.36, "elapsed_time": "23:05:40", "remaining_time": "2 days, 0:16:03"}
+{"current_steps": 1210, "total_steps": 3708, "loss": 0.2598, "lr": 4.2615150103117576e-05, "epoch": 0.9789644012944984, "percentage": 32.63, "elapsed_time": "23:17:13", "remaining_time": "2 days, 0:04:30"}
+{"current_steps": 1220, "total_steps": 3708, "loss": 0.2587, "lr": 4.24473603961183e-05, "epoch": 0.9870550161812298, "percentage": 32.9, "elapsed_time": "23:28:46", "remaining_time": "1 day, 23:52:58"}
+{"current_steps": 1230, "total_steps": 3708, "loss": 0.2589, "lr": 4.227802431754961e-05, "epoch": 0.9951456310679612, "percentage": 33.17, "elapsed_time": "23:40:19", "remaining_time": "1 day, 23:41:26"}
+{"current_steps": 1240, "total_steps": 3708, "loss": 0.2468, "lr": 4.210715687578481e-05, "epoch": 1.0032362459546926, "percentage": 33.44, "elapsed_time": "23:52:32", "remaining_time": "1 day, 23:31:12"}
+{"current_steps": 1250, "total_steps": 3708, "loss": 0.2219, "lr": 4.193477321492293e-05, "epoch": 1.0113268608414239, "percentage": 33.71, "elapsed_time": "1 day, 0:04:05", "remaining_time": "1 day, 23:19:40"}
+{"current_steps": 1260, "total_steps": 3708, "loss": 0.2251, "lr": 4.176088861344657e-05, "epoch": 1.0194174757281553, "percentage": 33.98, "elapsed_time": "1 day, 0:15:38", "remaining_time": "1 day, 23:08:06"}
+{"current_steps": 1270, "total_steps": 3708, "loss": 0.2178, "lr": 4.158551848286773e-05, "epoch": 1.0275080906148868, "percentage": 34.25, "elapsed_time": "1 day, 0:27:11", "remaining_time": "1 day, 22:56:33"}
+{"current_steps": 1280, "total_steps": 3708, "loss": 0.2289, "lr": 4.140867836636189e-05, "epoch": 1.035598705501618, "percentage": 34.52, "elapsed_time": "1 day, 0:38:44", "remaining_time": "1 day, 22:44:59"}
+{"current_steps": 1290, "total_steps": 3708, "loss": 0.2258, "lr": 4.1230383937390374e-05, "epoch": 1.0436893203883495, "percentage": 34.79, "elapsed_time": "1 day, 0:50:17", "remaining_time": "1 day, 22:33:26"}
+{"current_steps": 1300, "total_steps": 3708, "loss": 0.2216, "lr": 4.1050650998311215e-05, "epoch": 1.051779935275081, "percentage": 35.06, "elapsed_time": "1 day, 1:01:50", "remaining_time": "1 day, 22:21:52"}
+{"current_steps": 1310, "total_steps": 3708, "loss": 0.2225, "lr": 4.086949547897862e-05, "epoch": 1.0598705501618122, "percentage": 35.33, "elapsed_time": "1 day, 1:13:23", "remaining_time": "1 day, 22:10:19"}
+{"current_steps": 1320, "total_steps": 3708, "loss": 0.2227, "lr": 4.068693343533103e-05, "epoch": 1.0679611650485437, "percentage": 35.6, "elapsed_time": "1 day, 1:24:56", "remaining_time": "1 day, 21:58:45"}
+{"current_steps": 1330, "total_steps": 3708, "loss": 0.2268, "lr": 4.050298104796812e-05, "epoch": 1.0760517799352751, "percentage": 35.87, "elapsed_time": "1 day, 1:36:29", "remaining_time": "1 day, 21:47:12"}
+{"current_steps": 1340, "total_steps": 3708, "loss": 0.2306, "lr": 4.0317654620716704e-05, "epoch": 1.0841423948220066, "percentage": 36.14, "elapsed_time": "1 day, 1:48:02", "remaining_time": "1 day, 21:35:38"}
+{"current_steps": 1350, "total_steps": 3708, "loss": 0.2254, "lr": 4.013097057918566e-05, "epoch": 1.0922330097087378, "percentage": 36.41, "elapsed_time": "1 day, 1:59:35", "remaining_time": "1 day, 21:24:05"}
+{"current_steps": 1360, "total_steps": 3708, "loss": 0.2279, "lr": 3.9942945469310175e-05, "epoch": 1.1003236245954693, "percentage": 36.68, "elapsed_time": "1 day, 2:11:08", "remaining_time": "1 day, 21:12:32"}
+{"current_steps": 1370, "total_steps": 3708, "loss": 0.2256, "lr": 3.9753595955885264e-05, "epoch": 1.1084142394822007, "percentage": 36.95, "elapsed_time": "1 day, 2:22:42", "remaining_time": "1 day, 21:00:59"}
+{"current_steps": 1380, "total_steps": 3708, "loss": 0.2235, "lr": 3.9562938821088705e-05, "epoch": 1.116504854368932, "percentage": 37.22, "elapsed_time": "1 day, 2:34:15", "remaining_time": "1 day, 20:49:26"}
+{"current_steps": 1390, "total_steps": 3708, "loss": 0.2237, "lr": 3.9370990962993695e-05, "epoch": 1.1245954692556634, "percentage": 37.49, "elapsed_time": "1 day, 2:45:48", "remaining_time": "1 day, 20:37:53"}
+{"current_steps": 1400, "total_steps": 3708, "loss": 0.2215, "lr": 3.9177769394071086e-05, "epoch": 1.132686084142395, "percentage": 37.76, "elapsed_time": "1 day, 2:57:21", "remaining_time": "1 day, 20:26:19"}
+{"current_steps": 1410, "total_steps": 3708, "loss": 0.2254, "lr": 3.898329123968163e-05, "epoch": 1.1407766990291262, "percentage": 38.03, "elapsed_time": "1 day, 3:08:54", "remaining_time": "1 day, 20:14:46"}
+{"current_steps": 1420, "total_steps": 3708, "loss": 0.2184, "lr": 3.87875737365581e-05, "epoch": 1.1488673139158576, "percentage": 38.3, "elapsed_time": "1 day, 3:20:27", "remaining_time": "1 day, 20:03:12"}
+{"current_steps": 1430, "total_steps": 3708, "loss": 0.2176, "lr": 3.8590634231277626e-05, "epoch": 1.156957928802589, "percentage": 38.57, "elapsed_time": "1 day, 3:32:00", "remaining_time": "1 day, 19:51:40"}
+{"current_steps": 1440, "total_steps": 3708, "loss": 0.2207, "lr": 3.8392490178724184e-05, "epoch": 1.1650485436893203, "percentage": 38.83, "elapsed_time": "1 day, 3:43:33", "remaining_time": "1 day, 19:40:06"}
+{"current_steps": 1450, "total_steps": 3708, "loss": 0.223, "lr": 3.8193159140541645e-05, "epoch": 1.1731391585760518, "percentage": 39.1, "elapsed_time": "1 day, 3:55:06", "remaining_time": "1 day, 19:28:33"}
+{"current_steps": 1460, "total_steps": 3708, "loss": 0.2264, "lr": 3.7992658783577215e-05, "epoch": 1.1812297734627832, "percentage": 39.37, "elapsed_time": "1 day, 4:06:39", "remaining_time": "1 day, 19:16:59"}
+{"current_steps": 1470, "total_steps": 3708, "loss": 0.2216, "lr": 3.779100687831563e-05, "epoch": 1.1893203883495145, "percentage": 39.64, "elapsed_time": "1 day, 4:18:12", "remaining_time": "1 day, 19:05:26"}
+{"current_steps": 1480, "total_steps": 3708, "loss": 0.2201, "lr": 3.758822129730415e-05, "epoch": 1.197411003236246, "percentage": 39.91, "elapsed_time": "1 day, 4:29:45", "remaining_time": "1 day, 18:53:53"}
+{"current_steps": 1490, "total_steps": 3708, "loss": 0.2239, "lr": 3.738432001356851e-05, "epoch": 1.2055016181229774, "percentage": 40.18, "elapsed_time": "1 day, 4:41:19", "remaining_time": "1 day, 18:42:20"}
+{"current_steps": 1500, "total_steps": 3708, "loss": 0.2281, "lr": 3.7179321099019916e-05, "epoch": 1.2135922330097086, "percentage": 40.45, "elapsed_time": "1 day, 4:52:51", "remaining_time": "1 day, 18:30:46"}
+{"current_steps": 1510, "total_steps": 3708, "loss": 0.225, "lr": 3.6973242722853365e-05, "epoch": 1.22168284789644, "percentage": 40.72, "elapsed_time": "1 day, 5:04:24", "remaining_time": "1 day, 18:19:13"}
+{"current_steps": 1520, "total_steps": 3708, "loss": 0.2246, "lr": 3.6766103149937295e-05, "epoch": 1.2297734627831716, "percentage": 40.99, "elapsed_time": "1 day, 5:15:57", "remaining_time": "1 day, 18:07:39"}
+{"current_steps": 1530, "total_steps": 3708, "loss": 0.2249, "lr": 3.655792073919471e-05, "epoch": 1.237864077669903, "percentage": 41.26, "elapsed_time": "1 day, 5:27:30", "remaining_time": "1 day, 17:56:06"}
+{"current_steps": 1540, "total_steps": 3708, "loss": 0.2263, "lr": 3.634871394197607e-05, "epoch": 1.2459546925566343, "percentage": 41.53, "elapsed_time": "1 day, 5:39:03", "remaining_time": "1 day, 17:44:32"}
+{"current_steps": 1550, "total_steps": 3708, "loss": 0.2211, "lr": 3.6138501300423934e-05, "epoch": 1.2540453074433657, "percentage": 41.8, "elapsed_time": "1 day, 5:50:36", "remaining_time": "1 day, 17:32:59"}
+{"current_steps": 1560, "total_steps": 3708, "loss": 0.2243, "lr": 3.592730144582948e-05, "epoch": 1.262135922330097, "percentage": 42.07, "elapsed_time": "1 day, 6:02:09", "remaining_time": "1 day, 17:21:26"}
+{"current_steps": 1570, "total_steps": 3708, "loss": 0.2202, "lr": 3.571513309698131e-05, "epoch": 1.2702265372168284, "percentage": 42.34, "elapsed_time": "1 day, 6:13:42", "remaining_time": "1 day, 17:09:53"}
+{"current_steps": 1580, "total_steps": 3708, "loss": 0.2229, "lr": 3.5502015058506335e-05, "epoch": 1.27831715210356, "percentage": 42.61, "elapsed_time": "1 day, 6:25:16", "remaining_time": "1 day, 16:58:20"}
+{"current_steps": 1590, "total_steps": 3708, "loss": 0.219, "lr": 3.528796621920307e-05, "epoch": 1.2864077669902914, "percentage": 42.88, "elapsed_time": "1 day, 6:36:49", "remaining_time": "1 day, 16:46:47"}
+{"current_steps": 1600, "total_steps": 3708, "loss": 0.222, "lr": 3.50730055503676e-05, "epoch": 1.2944983818770226, "percentage": 43.15, "elapsed_time": "1 day, 6:48:22", "remaining_time": "1 day, 16:35:14"}
+{"current_steps": 1610, "total_steps": 3708, "loss": 0.2262, "lr": 3.485715210411204e-05, "epoch": 1.302588996763754, "percentage": 43.42, "elapsed_time": "1 day, 6:59:56", "remaining_time": "1 day, 16:23:42"}
+{"current_steps": 1620, "total_steps": 3708, "loss": 0.2226, "lr": 3.4640425011676034e-05, "epoch": 1.3106796116504853, "percentage": 43.69, "elapsed_time": "1 day, 7:11:30", "remaining_time": "1 day, 16:12:09"}
+{"current_steps": 1630, "total_steps": 3708, "loss": 0.2254, "lr": 3.442284348173106e-05, "epoch": 1.3187702265372168, "percentage": 43.96, "elapsed_time": "1 day, 7:23:03", "remaining_time": "1 day, 16:00:36"}
+{"current_steps": 1640, "total_steps": 3708, "loss": 0.2244, "lr": 3.420442679867796e-05, "epoch": 1.3268608414239482, "percentage": 44.23, "elapsed_time": "1 day, 7:34:36", "remaining_time": "1 day, 15:49:03"}
+{"current_steps": 1650, "total_steps": 3708, "loss": 0.2227, "lr": 3.398519432093782e-05, "epoch": 1.3349514563106797, "percentage": 44.5, "elapsed_time": "1 day, 7:46:10", "remaining_time": "1 day, 15:37:30"}
+{"current_steps": 1660, "total_steps": 3708, "loss": 0.2195, "lr": 3.376516547923614e-05, "epoch": 1.343042071197411, "percentage": 44.77, "elapsed_time": "1 day, 7:57:43", "remaining_time": "1 day, 15:25:57"}
+{"current_steps": 1670, "total_steps": 3708, "loss": 0.2203, "lr": 3.3544359774880714e-05, "epoch": 1.3511326860841424, "percentage": 45.04, "elapsed_time": "1 day, 8:09:16", "remaining_time": "1 day, 15:14:24"}
+{"current_steps": 1680, "total_steps": 3708, "loss": 0.2151, "lr": 3.3322796778033204e-05, "epoch": 1.3592233009708738, "percentage": 45.31, "elapsed_time": "1 day, 8:20:49", "remaining_time": "1 day, 15:02:51"}
+{"current_steps": 1690, "total_steps": 3708, "loss": 0.216, "lr": 3.3100496125974624e-05, "epoch": 1.367313915857605, "percentage": 45.58, "elapsed_time": "1 day, 8:32:22", "remaining_time": "1 day, 14:51:18"}
+{"current_steps": 1700, "total_steps": 3708, "loss": 0.2162, "lr": 3.2877477521364895e-05, "epoch": 1.3754045307443366, "percentage": 45.85, "elapsed_time": "1 day, 8:43:55", "remaining_time": "1 day, 14:39:44"}
+{"current_steps": 1710, "total_steps": 3708, "loss": 0.2187, "lr": 3.2653760730496555e-05, "epoch": 1.383495145631068, "percentage": 46.12, "elapsed_time": "1 day, 8:55:28", "remaining_time": "1 day, 14:28:11"}
+{"current_steps": 1720, "total_steps": 3708, "loss": 0.225, "lr": 3.242936558154285e-05, "epoch": 1.3915857605177995, "percentage": 46.39, "elapsed_time": "1 day, 9:07:02", "remaining_time": "1 day, 14:16:38"}
+{"current_steps": 1730, "total_steps": 3708, "loss": 0.2239, "lr": 3.2204311962800426e-05, "epoch": 1.3996763754045307, "percentage": 46.66, "elapsed_time": "1 day, 9:18:35", "remaining_time": "1 day, 14:05:05"}
+{"current_steps": 1740, "total_steps": 3708, "loss": 0.2201, "lr": 3.197861982092651e-05, "epoch": 1.4077669902912622, "percentage": 46.93, "elapsed_time": "1 day, 9:30:07", "remaining_time": "1 day, 13:53:31"}
+{"current_steps": 1750, "total_steps": 3708, "loss": 0.2234, "lr": 3.175230915917108e-05, "epoch": 1.4158576051779934, "percentage": 47.2, "elapsed_time": "1 day, 9:41:41", "remaining_time": "1 day, 13:41:58"}
+{"current_steps": 1760, "total_steps": 3708, "loss": 0.219, "lr": 3.152540003560398e-05, "epoch": 1.4239482200647249, "percentage": 47.46, "elapsed_time": "1 day, 9:53:14", "remaining_time": "1 day, 13:30:25"}
+{"current_steps": 1770, "total_steps": 3708, "loss": 0.2158, "lr": 3.129791256133712e-05, "epoch": 1.4320388349514563, "percentage": 47.73, "elapsed_time": "1 day, 10:04:47", "remaining_time": "1 day, 13:18:52"}
+{"current_steps": 1780, "total_steps": 3708, "loss": 0.2208, "lr": 3.106986689874204e-05, "epoch": 1.4401294498381878, "percentage": 48.0, "elapsed_time": "1 day, 10:16:20", "remaining_time": "1 day, 13:07:19"}
+{"current_steps": 1790, "total_steps": 3708, "loss": 0.2227, "lr": 3.0841283259662875e-05, "epoch": 1.448220064724919, "percentage": 48.27, "elapsed_time": "1 day, 10:27:54", "remaining_time": "1 day, 12:55:46"}
+{"current_steps": 1800, "total_steps": 3708, "loss": 0.222, "lr": 3.0612181903625014e-05, "epoch": 1.4563106796116505, "percentage": 48.54, "elapsed_time": "1 day, 10:39:27", "remaining_time": "1 day, 12:44:13"}
+{"current_steps": 1810, "total_steps": 3708, "loss": 0.2174, "lr": 3.0382583136039444e-05, "epoch": 1.4644012944983817, "percentage": 48.81, "elapsed_time": "1 day, 10:51:00", "remaining_time": "1 day, 12:32:40"}
+{"current_steps": 1820, "total_steps": 3708, "loss": 0.2167, "lr": 3.015250730640308e-05, "epoch": 1.4724919093851132, "percentage": 49.08, "elapsed_time": "1 day, 11:02:34", "remaining_time": "1 day, 12:21:08"}
+{"current_steps": 1830, "total_steps": 3708, "loss": 0.2141, "lr": 2.9921974806495178e-05, "epoch": 1.4805825242718447, "percentage": 49.35, "elapsed_time": "1 day, 11:14:07", "remaining_time": "1 day, 12:09:34"}
+{"current_steps": 1840, "total_steps": 3708, "loss": 0.2189, "lr": 2.969100606856998e-05, "epoch": 1.4886731391585761, "percentage": 49.62, "elapsed_time": "1 day, 11:25:40", "remaining_time": "1 day, 11:58:01"}
+{"current_steps": 1850, "total_steps": 3708, "loss": 0.2236, "lr": 2.9459621563545825e-05, "epoch": 1.4967637540453074, "percentage": 49.89, "elapsed_time": "1 day, 11:37:13", "remaining_time": "1 day, 11:46:28"}
+{"current_steps": 1860, "total_steps": 3708, "loss": 0.2182, "lr": 2.9227841799190775e-05, "epoch": 1.5048543689320388, "percentage": 50.16, "elapsed_time": "1 day, 11:48:47", "remaining_time": "1 day, 11:34:55"}
+{"current_steps": 1870, "total_steps": 3708, "loss": 0.2224, "lr": 2.8995687318304975e-05, "epoch": 1.51294498381877, "percentage": 50.43, "elapsed_time": "1 day, 12:00:20", "remaining_time": "1 day, 11:23:22"}
+{"current_steps": 1880, "total_steps": 3708, "loss": 0.219, "lr": 2.8763178696899995e-05, "epoch": 1.5210355987055015, "percentage": 50.7, "elapsed_time": "1 day, 12:11:53", "remaining_time": "1 day, 11:11:48"}
+{"current_steps": 1890, "total_steps": 3708, "loss": 0.2207, "lr": 2.853033654237507e-05, "epoch": 1.529126213592233, "percentage": 50.97, "elapsed_time": "1 day, 12:23:26", "remaining_time": "1 day, 11:00:15"}
+{"current_steps": 1900, "total_steps": 3708, "loss": 0.2191, "lr": 2.8297181491690756e-05, "epoch": 1.5372168284789645, "percentage": 51.24, "elapsed_time": "1 day, 12:34:59", "remaining_time": "1 day, 10:48:42"}
+{"current_steps": 1910, "total_steps": 3708, "loss": 0.2192, "lr": 2.8063734209539773e-05, "epoch": 1.545307443365696, "percentage": 51.51, "elapsed_time": "1 day, 12:46:32", "remaining_time": "1 day, 10:37:09"}
+{"current_steps": 1920, "total_steps": 3708, "loss": 0.2196, "lr": 2.783001538651554e-05, "epoch": 1.5533980582524272, "percentage": 51.78, "elapsed_time": "1 day, 12:58:05", "remaining_time": "1 day, 10:25:36"}
+{"current_steps": 1930, "total_steps": 3708, "loss": 0.2202, "lr": 2.7596045737278336e-05, "epoch": 1.5614886731391586, "percentage": 52.05, "elapsed_time": "1 day, 13:09:38", "remaining_time": "1 day, 10:14:02"}
+{"current_steps": 1940, "total_steps": 3708, "loss": 0.2224, "lr": 2.7361845998719315e-05, "epoch": 1.5695792880258899, "percentage": 52.32, "elapsed_time": "1 day, 13:21:11", "remaining_time": "1 day, 10:02:29"}
+{"current_steps": 1950, "total_steps": 3708, "loss": 0.2196, "lr": 2.7127436928122612e-05, "epoch": 1.5776699029126213, "percentage": 52.59, "elapsed_time": "1 day, 13:32:44", "remaining_time": "1 day, 9:50:56"}
+{"current_steps": 1960, "total_steps": 3708, "loss": 0.2134, "lr": 2.6892839301325623e-05, "epoch": 1.5857605177993528, "percentage": 52.86, "elapsed_time": "1 day, 13:44:18", "remaining_time": "1 day, 9:39:23"}
+{"current_steps": 1970, "total_steps": 3708, "loss": 0.2167, "lr": 2.6658073910877603e-05, "epoch": 1.5938511326860842, "percentage": 53.13, "elapsed_time": "1 day, 13:55:50", "remaining_time": "1 day, 9:27:49"}
+{"current_steps": 1980, "total_steps": 3708, "loss": 0.2161, "lr": 2.6423161564196803e-05, "epoch": 1.6019417475728155, "percentage": 53.4, "elapsed_time": "1 day, 14:07:23", "remaining_time": "1 day, 9:16:16"}
+{"current_steps": 1990, "total_steps": 3708, "loss": 0.2174, "lr": 2.6188123081726306e-05, "epoch": 1.610032362459547, "percentage": 53.67, "elapsed_time": "1 day, 14:18:57", "remaining_time": "1 day, 9:04:43"}
+{"current_steps": 2000, "total_steps": 3708, "loss": 0.2149, "lr": 2.5952979295088714e-05, "epoch": 1.6181229773462782, "percentage": 53.94, "elapsed_time": "1 day, 14:30:30", "remaining_time": "1 day, 8:53:10"}
+{"current_steps": 2010, "total_steps": 3708, "loss": 0.2148, "lr": 2.57177510452398e-05, "epoch": 1.6262135922330097, "percentage": 54.21, "elapsed_time": "1 day, 14:42:03", "remaining_time": "1 day, 8:41:37"}
+{"current_steps": 2020, "total_steps": 3708, "loss": 0.2191, "lr": 2.5482459180621377e-05, "epoch": 1.6343042071197411, "percentage": 54.48, "elapsed_time": "1 day, 14:53:36", "remaining_time": "1 day, 8:30:03"}
+{"current_steps": 2030, "total_steps": 3708, "loss": 0.2136, "lr": 2.524712455531347e-05, "epoch": 1.6423948220064726, "percentage": 54.75, "elapsed_time": "1 day, 15:05:09", "remaining_time": "1 day, 8:18:30"}
+{"current_steps": 2040, "total_steps": 3708, "loss": 0.2153, "lr": 2.501176802718599e-05, "epoch": 1.650485436893204, "percentage": 55.02, "elapsed_time": "1 day, 15:16:42", "remaining_time": "1 day, 8:06:57"}
+{"current_steps": 2050, "total_steps": 3708, "loss": 0.2149, "lr": 2.4776410456050165e-05, "epoch": 1.6585760517799353, "percentage": 55.29, "elapsed_time": "1 day, 15:28:15", "remaining_time": "1 day, 7:55:24"}
+{"current_steps": 2060, "total_steps": 3708, "loss": 0.2187, "lr": 2.4541072701809624e-05, "epoch": 1.6666666666666665, "percentage": 55.56, "elapsed_time": "1 day, 15:39:49", "remaining_time": "1 day, 7:43:51"}
+{"current_steps": 2070, "total_steps": 3708, "loss": 0.2156, "lr": 2.4305775622611627e-05, "epoch": 1.674757281553398, "percentage": 55.83, "elapsed_time": "1 day, 15:51:22", "remaining_time": "1 day, 7:32:18"}
+{"current_steps": 2080, "total_steps": 3708, "loss": 0.2186, "lr": 2.4070540072998372e-05, "epoch": 1.6828478964401294, "percentage": 56.09, "elapsed_time": "1 day, 16:02:55", "remaining_time": "1 day, 7:20:45"}
+{"current_steps": 2090, "total_steps": 3708, "loss": 0.2145, "lr": 2.3835386902058637e-05, "epoch": 1.690938511326861, "percentage": 56.36, "elapsed_time": "1 day, 16:14:28", "remaining_time": "1 day, 7:09:11"}
+{"current_steps": 2100, "total_steps": 3708, "loss": 0.213, "lr": 2.360033695157995e-05, "epoch": 1.6990291262135924, "percentage": 56.63, "elapsed_time": "1 day, 16:26:01", "remaining_time": "1 day, 6:57:38"}
+{"current_steps": 2110, "total_steps": 3708, "loss": 0.2154, "lr": 2.3365411054201315e-05, "epoch": 1.7071197411003236, "percentage": 56.9, "elapsed_time": "1 day, 16:37:34", "remaining_time": "1 day, 6:46:05"}
+{"current_steps": 2120, "total_steps": 3708, "loss": 0.2209, "lr": 2.3130630031566818e-05, "epoch": 1.715210355987055, "percentage": 57.17, "elapsed_time": "1 day, 16:49:07", "remaining_time": "1 day, 6:34:32"}
+{"current_steps": 2130, "total_steps": 3708, "loss": 0.2156, "lr": 2.2896014692480226e-05, "epoch": 1.7233009708737863, "percentage": 57.44, "elapsed_time": "1 day, 17:00:41", "remaining_time": "1 day, 6:22:59"}
+{"current_steps": 2140, "total_steps": 3708, "loss": 0.2156, "lr": 2.266158583106063e-05, "epoch": 1.7313915857605178, "percentage": 57.71, "elapsed_time": "1 day, 17:12:14", "remaining_time": "1 day, 6:11:25"}
+{"current_steps": 2150, "total_steps": 3708, "loss": 0.2152, "lr": 2.2427364224899502e-05, "epoch": 1.7394822006472492, "percentage": 57.98, "elapsed_time": "1 day, 17:23:46", "remaining_time": "1 day, 5:59:52"}
+{"current_steps": 2160, "total_steps": 3708, "loss": 0.2137, "lr": 2.2193370633219115e-05, "epoch": 1.7475728155339807, "percentage": 58.25, "elapsed_time": "1 day, 17:35:19", "remaining_time": "1 day, 5:48:19"}
+{"current_steps": 2170, "total_steps": 3708, "loss": 0.2113, "lr": 2.1959625795032664e-05, "epoch": 1.755663430420712, "percentage": 58.52, "elapsed_time": "1 day, 17:46:52", "remaining_time": "1 day, 5:36:45"}
+{"current_steps": 2180, "total_steps": 3708, "loss": 0.2146, "lr": 2.1726150427306182e-05, "epoch": 1.7637540453074434, "percentage": 58.79, "elapsed_time": "1 day, 17:58:25", "remaining_time": "1 day, 5:25:12"}
+{"current_steps": 2190, "total_steps": 3708, "loss": 0.2132, "lr": 2.1492965223122305e-05, "epoch": 1.7718446601941746, "percentage": 59.06, "elapsed_time": "1 day, 18:09:58", "remaining_time": "1 day, 5:13:39"}
+{"current_steps": 2200, "total_steps": 3708, "loss": 0.2124, "lr": 2.126009084984629e-05, "epoch": 1.779935275080906, "percentage": 59.33, "elapsed_time": "1 day, 18:21:31", "remaining_time": "1 day, 5:02:05"}
+{"current_steps": 2210, "total_steps": 3708, "loss": 0.2151, "lr": 2.102754794729426e-05, "epoch": 1.7880258899676376, "percentage": 59.6, "elapsed_time": "1 day, 18:33:04", "remaining_time": "1 day, 4:50:32"}
+{"current_steps": 2220, "total_steps": 3708, "loss": 0.2107, "lr": 2.079535712590382e-05, "epoch": 1.796116504854369, "percentage": 59.87, "elapsed_time": "1 day, 18:44:37", "remaining_time": "1 day, 4:38:59"}
+{"current_steps": 2230, "total_steps": 3708, "loss": 0.2091, "lr": 2.056353896490742e-05, "epoch": 1.8042071197411005, "percentage": 60.14, "elapsed_time": "1 day, 18:56:09", "remaining_time": "1 day, 4:27:25"}
+{"current_steps": 2240, "total_steps": 3708, "loss": 0.2114, "lr": 2.0332114010508334e-05, "epoch": 1.8122977346278317, "percentage": 60.41, "elapsed_time": "1 day, 19:07:42", "remaining_time": "1 day, 4:15:52"}
+{"current_steps": 2250, "total_steps": 3708, "loss": 0.213, "lr": 2.010110277405966e-05, "epoch": 1.820388349514563, "percentage": 60.68, "elapsed_time": "1 day, 19:19:15", "remaining_time": "1 day, 4:04:19"}
+{"current_steps": 2260, "total_steps": 3708, "loss": 0.2142, "lr": 1.9870525730246424e-05, "epoch": 1.8284789644012944, "percentage": 60.95, "elapsed_time": "1 day, 19:30:49", "remaining_time": "1 day, 3:52:46"}
+{"current_steps": 2270, "total_steps": 3708, "loss": 0.212, "lr": 1.9640403315270824e-05, "epoch": 1.8365695792880259, "percentage": 61.22, "elapsed_time": "1 day, 19:42:22", "remaining_time": "1 day, 3:41:13"}
+{"current_steps": 2280, "total_steps": 3708, "loss": 0.2107, "lr": 1.9410755925041006e-05, "epoch": 1.8446601941747574, "percentage": 61.49, "elapsed_time": "1 day, 19:53:55", "remaining_time": "1 day, 3:29:39"}
+{"current_steps": 2290, "total_steps": 3708, "loss": 0.2117, "lr": 1.918160391336335e-05, "epoch": 1.8527508090614888, "percentage": 61.76, "elapsed_time": "1 day, 20:05:27", "remaining_time": "1 day, 3:18:06"}
+{"current_steps": 2300, "total_steps": 3708, "loss": 0.2099, "lr": 1.8952967590138472e-05, "epoch": 1.86084142394822, "percentage": 62.03, "elapsed_time": "1 day, 20:17:00", "remaining_time": "1 day, 3:06:33"}
+{"current_steps": 2310, "total_steps": 3708, "loss": 0.2154, "lr": 1.8724867219561203e-05, "epoch": 1.8689320388349513, "percentage": 62.3, "elapsed_time": "1 day, 20:28:33", "remaining_time": "1 day, 2:54:59"}
+{"current_steps": 2320, "total_steps": 3708, "loss": 0.2145, "lr": 1.8497323018324476e-05, "epoch": 1.8770226537216828, "percentage": 62.57, "elapsed_time": "1 day, 20:40:06", "remaining_time": "1 day, 2:43:26"}
+{"current_steps": 2330, "total_steps": 3708, "loss": 0.2077, "lr": 1.8270355153827598e-05, "epoch": 1.8851132686084142, "percentage": 62.84, "elapsed_time": "1 day, 20:51:39", "remaining_time": "1 day, 2:31:53"}
+{"current_steps": 2340, "total_steps": 3708, "loss": 0.2102, "lr": 1.804398374238872e-05, "epoch": 1.8932038834951457, "percentage": 63.11, "elapsed_time": "1 day, 21:03:12", "remaining_time": "1 day, 2:20:20"}
+{"current_steps": 2350, "total_steps": 3708, "loss": 0.2098, "lr": 1.781822884746196e-05, "epoch": 1.9012944983818771, "percentage": 63.38, "elapsed_time": "1 day, 21:14:45", "remaining_time": "1 day, 2:08:47"}
+{"current_steps": 2360, "total_steps": 3708, "loss": 0.2114, "lr": 1.7593110477859153e-05, "epoch": 1.9093851132686084, "percentage": 63.65, "elapsed_time": "1 day, 21:26:19", "remaining_time": "1 day, 1:57:14"}
+{"current_steps": 2370, "total_steps": 3708, "loss": 0.2125, "lr": 1.736864858597645e-05, "epoch": 1.9174757281553398, "percentage": 63.92, "elapsed_time": "1 day, 21:37:52", "remaining_time": "1 day, 1:45:40"}
+{"current_steps": 2380, "total_steps": 3708, "loss": 0.2101, "lr": 1.7144863066025955e-05, "epoch": 1.925566343042071, "percentage": 64.19, "elapsed_time": "1 day, 21:49:25", "remaining_time": "1 day, 1:34:07"}
+{"current_steps": 2390, "total_steps": 3708, "loss": 0.2148, "lr": 1.692177375227242e-05, "epoch": 1.9336569579288025, "percentage": 64.46, "elapsed_time": "1 day, 22:00:58", "remaining_time": "1 day, 1:22:34"}
+{"current_steps": 2400, "total_steps": 3708, "loss": 0.2135, "lr": 1.669940041727538e-05, "epoch": 1.941747572815534, "percentage": 64.72, "elapsed_time": "1 day, 22:12:31", "remaining_time": "1 day, 1:11:01"}
+{"current_steps": 2410, "total_steps": 3708, "loss": 0.2107, "lr": 1.6477762770136707e-05, "epoch": 1.9498381877022655, "percentage": 64.99, "elapsed_time": "1 day, 22:24:04", "remaining_time": "1 day, 0:59:28"}
+{"current_steps": 2420, "total_steps": 3708, "loss": 0.2154, "lr": 1.625688045475371e-05, "epoch": 1.9579288025889967, "percentage": 65.26, "elapsed_time": "1 day, 22:35:38", "remaining_time": "1 day, 0:47:55"}
+{"current_steps": 2430, "total_steps": 3708, "loss": 0.2085, "lr": 1.603677304807815e-05, "epoch": 1.9660194174757282, "percentage": 65.53, "elapsed_time": "1 day, 22:47:11", "remaining_time": "1 day, 0:36:22"}
+{"current_steps": 2440, "total_steps": 3708, "loss": 0.2089, "lr": 1.5817460058381088e-05, "epoch": 1.9741100323624594, "percentage": 65.8, "elapsed_time": "1 day, 22:58:44", "remaining_time": "1 day, 0:24:49"}
+{"current_steps": 2450, "total_steps": 3708, "loss": 0.2088, "lr": 1.5598960923523842e-05, "epoch": 1.9822006472491909, "percentage": 66.07, "elapsed_time": "1 day, 23:10:17", "remaining_time": "1 day, 0:13:16"}
+{"current_steps": 2460, "total_steps": 3708, "loss": 0.2033, "lr": 1.5381295009235262e-05, "epoch": 1.9902912621359223, "percentage": 66.34, "elapsed_time": "1 day, 23:21:50", "remaining_time": "1 day, 0:01:42"}
+{"current_steps": 2470, "total_steps": 3708, "loss": 0.2073, "lr": 1.5164481607395238e-05, "epoch": 1.9983818770226538, "percentage": 66.61, "elapsed_time": "1 day, 23:33:23", "remaining_time": "23:50:09"}
+{"current_steps": 2480, "total_steps": 3708, "loss": 0.184, "lr": 1.4948539934324923e-05, "epoch": 2.0064724919093853, "percentage": 66.88, "elapsed_time": "1 day, 23:45:38", "remaining_time": "23:38:57"}
+{"current_steps": 2490, "total_steps": 3708, "loss": 0.1729, "lr": 1.4733489129083534e-05, "epoch": 2.0145631067961167, "percentage": 67.15, "elapsed_time": "1 day, 23:57:11", "remaining_time": "23:27:23"}
+{"current_steps": 2500, "total_steps": 3708, "loss": 0.1732, "lr": 1.4519348251772058e-05, "epoch": 2.0226537216828477, "percentage": 67.42, "elapsed_time": "2 days, 0:08:46", "remaining_time": "23:15:51"}
+{"current_steps": 2510, "total_steps": 3708, "loss": 0.1712, "lr": 1.4306136281843962e-05, "epoch": 2.030744336569579, "percentage": 67.69, "elapsed_time": "2 days, 0:20:22", "remaining_time": "23:04:19"}
+{"current_steps": 2520, "total_steps": 3708, "loss": 0.1703, "lr": 1.4093872116422979e-05, "epoch": 2.0388349514563107, "percentage": 67.96, "elapsed_time": "2 days, 0:31:57", "remaining_time": "22:52:46"}
+{"current_steps": 2530, "total_steps": 3708, "loss": 0.1706, "lr": 1.3882574568628315e-05, "epoch": 2.046925566343042, "percentage": 68.23, "elapsed_time": "2 days, 0:43:31", "remaining_time": "22:41:13"}
+{"current_steps": 2540, "total_steps": 3708, "loss": 0.1695, "lr": 1.3672262365907163e-05, "epoch": 2.0550161812297736, "percentage": 68.5, "elapsed_time": "2 days, 0:55:04", "remaining_time": "22:29:40"}
+{"current_steps": 2550, "total_steps": 3708, "loss": 0.1701, "lr": 1.3462954148374899e-05, "epoch": 2.063106796116505, "percentage": 68.77, "elapsed_time": "2 days, 1:06:37", "remaining_time": "22:18:06"}
+{"current_steps": 2560, "total_steps": 3708, "loss": 0.1722, "lr": 1.3254668467163029e-05, "epoch": 2.071197411003236, "percentage": 69.04, "elapsed_time": "2 days, 1:18:10", "remaining_time": "22:06:33"}
+{"current_steps": 2570, "total_steps": 3708, "loss": 0.1696, "lr": 1.3047423782774937e-05, "epoch": 2.0792880258899675, "percentage": 69.31, "elapsed_time": "2 days, 1:29:43", "remaining_time": "21:55:00"}
+{"current_steps": 2580, "total_steps": 3708, "loss": 0.1712, "lr": 1.2841238463449743e-05, "epoch": 2.087378640776699, "percentage": 69.58, "elapsed_time": "2 days, 1:41:16", "remaining_time": "21:43:26"}
+{"current_steps": 2590, "total_steps": 3708, "loss": 0.1711, "lr": 1.2636130783534319e-05, "epoch": 2.0954692556634305, "percentage": 69.85, "elapsed_time": "2 days, 1:52:50", "remaining_time": "21:31:53"}
+{"current_steps": 2600, "total_steps": 3708, "loss": 0.1661, "lr": 1.2432118921863604e-05, "epoch": 2.103559870550162, "percentage": 70.12, "elapsed_time": "2 days, 2:04:23", "remaining_time": "21:20:19"}
+{"current_steps": 2610, "total_steps": 3708, "loss": 0.1721, "lr": 1.2229220960149431e-05, "epoch": 2.1116504854368934, "percentage": 70.39, "elapsed_time": "2 days, 2:15:55", "remaining_time": "21:08:46"}
+{"current_steps": 2620, "total_steps": 3708, "loss": 0.1681, "lr": 1.2027454881377889e-05, "epoch": 2.1197411003236244, "percentage": 70.66, "elapsed_time": "2 days, 2:27:28", "remaining_time": "20:57:12"}
+{"current_steps": 2630, "total_steps": 3708, "loss": 0.1693, "lr": 1.1826838568215526e-05, "epoch": 2.127831715210356, "percentage": 70.93, "elapsed_time": "2 days, 2:39:01", "remaining_time": "20:45:39"}
+{"current_steps": 2640, "total_steps": 3708, "loss": 0.1674, "lr": 1.1627389801424351e-05, "epoch": 2.1359223300970873, "percentage": 71.2, "elapsed_time": "2 days, 2:50:35", "remaining_time": "20:34:06"}
+{"current_steps": 2650, "total_steps": 3708, "loss": 0.1743, "lr": 1.1429126258285946e-05, "epoch": 2.144012944983819, "percentage": 71.47, "elapsed_time": "2 days, 3:02:08", "remaining_time": "20:22:32"}
+{"current_steps": 2660, "total_steps": 3708, "loss": 0.1749, "lr": 1.1232065511034696e-05, "epoch": 2.1521035598705502, "percentage": 71.74, "elapsed_time": "2 days, 3:13:41", "remaining_time": "20:10:59"}
+{"current_steps": 2670, "total_steps": 3708, "loss": 0.1758, "lr": 1.1036225025300357e-05, "epoch": 2.1601941747572817, "percentage": 72.01, "elapsed_time": "2 days, 3:25:14", "remaining_time": "19:59:25"}
+{"current_steps": 2680, "total_steps": 3708, "loss": 0.1762, "lr": 1.0841622158560085e-05, "epoch": 2.168284789644013, "percentage": 72.28, "elapsed_time": "2 days, 3:36:47", "remaining_time": "19:47:52"}
+{"current_steps": 2690, "total_steps": 3708, "loss": 0.1747, "lr": 1.0648274158599994e-05, "epoch": 2.176375404530744, "percentage": 72.55, "elapsed_time": "2 days, 3:48:20", "remaining_time": "19:36:18"}
+{"current_steps": 2700, "total_steps": 3708, "loss": 0.171, "lr": 1.0456198161986489e-05, "epoch": 2.1844660194174756, "percentage": 72.82, "elapsed_time": "2 days, 3:59:53", "remaining_time": "19:24:45"}
+{"current_steps": 2710, "total_steps": 3708, "loss": 0.1712, "lr": 1.0265411192547462e-05, "epoch": 2.192556634304207, "percentage": 73.09, "elapsed_time": "2 days, 4:11:26", "remaining_time": "19:13:12"}
+{"current_steps": 2720, "total_steps": 3708, "loss": 0.1713, "lr": 1.0075930159863416e-05, "epoch": 2.2006472491909386, "percentage": 73.35, "elapsed_time": "2 days, 4:23:00", "remaining_time": "19:01:38"}
+{"current_steps": 2730, "total_steps": 3708, "loss": 0.1716, "lr": 9.887771857768796e-06, "epoch": 2.20873786407767, "percentage": 73.62, "elapsed_time": "2 days, 4:34:32", "remaining_time": "18:50:05"}
+{"current_steps": 2740, "total_steps": 3708, "loss": 0.1724, "lr": 9.700952962863513e-06, "epoch": 2.2168284789644015, "percentage": 73.89, "elapsed_time": "2 days, 4:46:06", "remaining_time": "18:38:32"}
+{"current_steps": 2750, "total_steps": 3708, "loss": 0.1747, "lr": 9.515490033034893e-06, "epoch": 2.2249190938511325, "percentage": 74.16, "elapsed_time": "2 days, 4:57:38", "remaining_time": "18:26:58"}
+{"current_steps": 2760, "total_steps": 3708, "loss": 0.17, "lr": 9.331399505990168e-06, "epoch": 2.233009708737864, "percentage": 74.43, "elapsed_time": "2 days, 5:09:11", "remaining_time": "18:15:25"}
+{"current_steps": 2770, "total_steps": 3708, "loss": 0.1744, "lr": 9.148697697799533e-06, "epoch": 2.2411003236245954, "percentage": 74.7, "elapsed_time": "2 days, 5:20:45", "remaining_time": "18:03:51"}
+{"current_steps": 2780, "total_steps": 3708, "loss": 0.1683, "lr": 8.967400801450105e-06, "epoch": 2.249190938511327, "percentage": 74.97, "elapsed_time": "2 days, 5:32:17", "remaining_time": "17:52:18"}
+{"current_steps": 2790, "total_steps": 3708, "loss": 0.1684, "lr": 8.787524885410678e-06, "epoch": 2.2572815533980584, "percentage": 75.24, "elapsed_time": "2 days, 5:43:51", "remaining_time": "17:40:45"}
+{"current_steps": 2800, "total_steps": 3708, "loss": 0.1695, "lr": 8.60908589220758e-06, "epoch": 2.26537216828479, "percentage": 75.51, "elapsed_time": "2 days, 5:55:24", "remaining_time": "17:29:11"}
+{"current_steps": 2810, "total_steps": 3708, "loss": 0.1686, "lr": 8.432099637011693e-06, "epoch": 2.273462783171521, "percentage": 75.78, "elapsed_time": "2 days, 6:06:56", "remaining_time": "17:17:38"}
+{"current_steps": 2820, "total_steps": 3708, "loss": 0.1727, "lr": 8.256581806236704e-06, "epoch": 2.2815533980582523, "percentage": 76.05, "elapsed_time": "2 days, 6:18:29", "remaining_time": "17:06:04"}
+{"current_steps": 2830, "total_steps": 3708, "loss": 0.1702, "lr": 8.082547956148873e-06, "epoch": 2.2896440129449838, "percentage": 76.32, "elapsed_time": "2 days, 6:30:02", "remaining_time": "16:54:31"}
+{"current_steps": 2840, "total_steps": 3708, "loss": 0.1688, "lr": 7.91001351148819e-06, "epoch": 2.2977346278317152, "percentage": 76.59, "elapsed_time": "2 days, 6:41:35", "remaining_time": "16:42:57"}
+{"current_steps": 2850, "total_steps": 3708, "loss": 0.1687, "lr": 7.738993764101324e-06, "epoch": 2.3058252427184467, "percentage": 76.86, "elapsed_time": "2 days, 6:53:08", "remaining_time": "16:31:24"}
+{"current_steps": 2860, "total_steps": 3708, "loss": 0.1709, "lr": 7.569503871586292e-06, "epoch": 2.313915857605178, "percentage": 77.13, "elapsed_time": "2 days, 7:04:41", "remaining_time": "16:19:51"}
+{"current_steps": 2870, "total_steps": 3708, "loss": 0.168, "lr": 7.401558855949004e-06, "epoch": 2.3220064724919096, "percentage": 77.4, "elapsed_time": "2 days, 7:16:14", "remaining_time": "16:08:17"}
+{"current_steps": 2880, "total_steps": 3708, "loss": 0.1697, "lr": 7.235173602271875e-06, "epoch": 2.3300970873786406, "percentage": 77.67, "elapsed_time": "2 days, 7:27:48", "remaining_time": "15:56:44"}
+{"current_steps": 2890, "total_steps": 3708, "loss": 0.1712, "lr": 7.070362857394538e-06, "epoch": 2.338187702265372, "percentage": 77.94, "elapsed_time": "2 days, 7:39:21", "remaining_time": "15:45:11"}
+{"current_steps": 2900, "total_steps": 3708, "loss": 0.165, "lr": 6.907141228606831e-06, "epoch": 2.3462783171521036, "percentage": 78.21, "elapsed_time": "2 days, 7:50:54", "remaining_time": "15:33:37"}
+{"current_steps": 2910, "total_steps": 3708, "loss": 0.1717, "lr": 6.745523182354147e-06, "epoch": 2.354368932038835, "percentage": 78.48, "elapsed_time": "2 days, 8:02:27", "remaining_time": "15:22:04"}
+{"current_steps": 2920, "total_steps": 3708, "loss": 0.1668, "lr": 6.585523042955233e-06, "epoch": 2.3624595469255665, "percentage": 78.75, "elapsed_time": "2 days, 8:14:00", "remaining_time": "15:10:31"}
+{"current_steps": 2930, "total_steps": 3708, "loss": 0.1721, "lr": 6.427154991332665e-06, "epoch": 2.3705501618122975, "percentage": 79.02, "elapsed_time": "2 days, 8:25:33", "remaining_time": "14:58:57"}
+{"current_steps": 2940, "total_steps": 3708, "loss": 0.1709, "lr": 6.2704330637559315e-06, "epoch": 2.378640776699029, "percentage": 79.29, "elapsed_time": "2 days, 8:37:06", "remaining_time": "14:47:24"}
+{"current_steps": 2950, "total_steps": 3708, "loss": 0.1702, "lr": 6.115371150597413e-06, "epoch": 2.3867313915857604, "percentage": 79.56, "elapsed_time": "2 days, 8:48:39", "remaining_time": "14:35:51"}
+{"current_steps": 2960, "total_steps": 3708, "loss": 0.1729, "lr": 5.961982995101301e-06, "epoch": 2.394822006472492, "percentage": 79.83, "elapsed_time": "2 days, 9:00:12", "remaining_time": "14:24:17"}
+{"current_steps": 2970, "total_steps": 3708, "loss": 0.1722, "lr": 5.810282192165442e-06, "epoch": 2.4029126213592233, "percentage": 80.1, "elapsed_time": "2 days, 9:11:45", "remaining_time": "14:12:44"}
+{"current_steps": 2980, "total_steps": 3708, "loss": 0.1689, "lr": 5.660282187136507e-06, "epoch": 2.411003236245955, "percentage": 80.37, "elapsed_time": "2 days, 9:23:18", "remaining_time": "14:01:11"}
+{"current_steps": 2990, "total_steps": 3708, "loss": 0.1702, "lr": 5.511996274618253e-06, "epoch": 2.4190938511326863, "percentage": 80.64, "elapsed_time": "2 days, 9:34:51", "remaining_time": "13:49:37"}
+{"current_steps": 3000, "total_steps": 3708, "loss": 0.1702, "lr": 5.365437597293238e-06, "epoch": 2.4271844660194173, "percentage": 80.91, "elapsed_time": "2 days, 9:46:25", "remaining_time": "13:38:04"}
+{"current_steps": 3010, "total_steps": 3708, "loss": 0.1693, "lr": 5.220619144757996e-06, "epoch": 2.4352750809061487, "percentage": 81.18, "elapsed_time": "2 days, 9:57:58", "remaining_time": "13:26:31"}
+{"current_steps": 3020, "total_steps": 3708, "loss": 0.1678, "lr": 5.077553752371708e-06, "epoch": 2.44336569579288, "percentage": 81.45, "elapsed_time": "2 days, 10:09:31", "remaining_time": "13:14:57"}
+{"current_steps": 3030, "total_steps": 3708, "loss": 0.1709, "lr": 4.936254100118656e-06, "epoch": 2.4514563106796117, "percentage": 81.72, "elapsed_time": "2 days, 10:21:04", "remaining_time": "13:03:24"}
+{"current_steps": 3040, "total_steps": 3708, "loss": 0.1658, "lr": 4.796732711484342e-06, "epoch": 2.459546925566343, "percentage": 81.98, "elapsed_time": "2 days, 10:32:37", "remaining_time": "12:51:51"}
+{"current_steps": 3050, "total_steps": 3708, "loss": 0.1682, "lr": 4.659001952345538e-06, "epoch": 2.4676375404530746, "percentage": 82.25, "elapsed_time": "2 days, 10:44:11", "remaining_time": "12:40:18"}
+{"current_steps": 3060, "total_steps": 3708, "loss": 0.1728, "lr": 4.523074029874291e-06, "epoch": 2.475728155339806, "percentage": 82.52, "elapsed_time": "2 days, 10:55:44", "remaining_time": "12:28:44"}
+{"current_steps": 3070, "total_steps": 3708, "loss": 0.1692, "lr": 4.388960991455998e-06, "epoch": 2.483818770226537, "percentage": 82.79, "elapsed_time": "2 days, 11:07:17", "remaining_time": "12:17:11"}
+{"current_steps": 3080, "total_steps": 3708, "loss": 0.1706, "lr": 4.256674723621621e-06, "epoch": 2.4919093851132685, "percentage": 83.06, "elapsed_time": "2 days, 11:18:51", "remaining_time": "12:05:38"}
+{"current_steps": 3090, "total_steps": 3708, "loss": 0.1708, "lr": 4.126226950994211e-06, "epoch": 2.5, "percentage": 83.33, "elapsed_time": "2 days, 11:30:24", "remaining_time": "11:54:04"}
+{"current_steps": 3100, "total_steps": 3708, "loss": 0.1737, "lr": 3.997629235249692e-06, "epoch": 2.5080906148867315, "percentage": 83.6, "elapsed_time": "2 days, 11:41:58", "remaining_time": "11:42:31"}
+{"current_steps": 3110, "total_steps": 3708, "loss": 0.1677, "lr": 3.870892974092197e-06, "epoch": 2.516181229773463, "percentage": 83.87, "elapsed_time": "2 days, 11:53:31", "remaining_time": "11:30:58"}
+{"current_steps": 3120, "total_steps": 3708, "loss": 0.1676, "lr": 3.7460294002438444e-06, "epoch": 2.524271844660194, "percentage": 84.14, "elapsed_time": "2 days, 12:05:05", "remaining_time": "11:19:25"}
+{"current_steps": 3130, "total_steps": 3708, "loss": 0.174, "lr": 3.6230495804491864e-06, "epoch": 2.5323624595469254, "percentage": 84.41, "elapsed_time": "2 days, 12:16:38", "remaining_time": "11:07:51"}
+{"current_steps": 3140, "total_steps": 3708, "loss": 0.1723, "lr": 3.5019644144943576e-06, "epoch": 2.540453074433657, "percentage": 84.68, "elapsed_time": "2 days, 12:28:11", "remaining_time": "10:56:18"}
+{"current_steps": 3150, "total_steps": 3708, "loss": 0.1679, "lr": 3.382784634241015e-06, "epoch": 2.5485436893203883, "percentage": 84.95, "elapsed_time": "2 days, 12:39:44", "remaining_time": "10:44:45"}
+{"current_steps": 3160, "total_steps": 3708, "loss": 0.1707, "lr": 3.2655208026751816e-06, "epoch": 2.55663430420712, "percentage": 85.22, "elapsed_time": "2 days, 12:51:17", "remaining_time": "10:33:12"}
+{"current_steps": 3170, "total_steps": 3708, "loss": 0.164, "lr": 3.150183312971014e-06, "epoch": 2.5647249190938513, "percentage": 85.49, "elapsed_time": "2 days, 13:02:50", "remaining_time": "10:21:38"}
+{"current_steps": 3180, "total_steps": 3708, "loss": 0.1676, "lr": 3.036782387569659e-06, "epoch": 2.5728155339805827, "percentage": 85.76, "elapsed_time": "2 days, 13:14:24", "remaining_time": "10:10:05"}
+{"current_steps": 3190, "total_steps": 3708, "loss": 0.1668, "lr": 2.9253280772732595e-06, "epoch": 2.5809061488673137, "percentage": 86.03, "elapsed_time": "2 days, 13:25:57", "remaining_time": "9:58:32"}
+{"current_steps": 3200, "total_steps": 3708, "loss": 0.1677, "lr": 2.8158302603540965e-06, "epoch": 2.588996763754045, "percentage": 86.3, "elapsed_time": "2 days, 13:37:30", "remaining_time": "9:46:58"}
+{"current_steps": 3210, "total_steps": 3708, "loss": 0.1634, "lr": 2.708298641679105e-06, "epoch": 2.5970873786407767, "percentage": 86.57, "elapsed_time": "2 days, 13:49:04", "remaining_time": "9:35:25"}
+{"current_steps": 3220, "total_steps": 3708, "loss": 0.1675, "lr": 2.6027427518497153e-06, "epoch": 2.605177993527508, "percentage": 86.84, "elapsed_time": "2 days, 14:00:37", "remaining_time": "9:23:52"}
+{"current_steps": 3230, "total_steps": 3708, "loss": 0.1652, "lr": 2.49917194635714e-06, "epoch": 2.6132686084142396, "percentage": 87.11, "elapsed_time": "2 days, 14:12:11", "remaining_time": "9:12:19"}
+{"current_steps": 3240, "total_steps": 3708, "loss": 0.1708, "lr": 2.397595404753225e-06, "epoch": 2.6213592233009706, "percentage": 87.38, "elapsed_time": "2 days, 14:23:44", "remaining_time": "9:00:45"}
+{"current_steps": 3250, "total_steps": 3708, "loss": 0.1718, "lr": 2.2980221298367995e-06, "epoch": 2.6294498381877025, "percentage": 87.65, "elapsed_time": "2 days, 14:35:17", "remaining_time": "8:49:12"}
+{"current_steps": 3260, "total_steps": 3708, "loss": 0.1687, "lr": 2.2004609468558175e-06, "epoch": 2.6375404530744335, "percentage": 87.92, "elapsed_time": "2 days, 14:46:50", "remaining_time": "8:37:39"}
+{"current_steps": 3270, "total_steps": 3708, "loss": 0.168, "lr": 2.1049205027251216e-06, "epoch": 2.645631067961165, "percentage": 88.19, "elapsed_time": "2 days, 14:58:23", "remaining_time": "8:26:05"}
+{"current_steps": 3280, "total_steps": 3708, "loss": 0.1682, "lr": 2.0114092652600806e-06, "epoch": 2.6537216828478964, "percentage": 88.46, "elapsed_time": "2 days, 15:09:56", "remaining_time": "8:14:32"}
+{"current_steps": 3290, "total_steps": 3708, "loss": 0.1688, "lr": 1.919935522426081e-06, "epoch": 2.661812297734628, "percentage": 88.73, "elapsed_time": "2 days, 15:21:29", "remaining_time": "8:02:59"}
+{"current_steps": 3300, "total_steps": 3708, "loss": 0.17, "lr": 1.8305073816039492e-06, "epoch": 2.6699029126213594, "percentage": 89.0, "elapsed_time": "2 days, 15:33:02", "remaining_time": "7:51:25"}
+{"current_steps": 3310, "total_steps": 3708, "loss": 0.1666, "lr": 1.7431327688714139e-06, "epoch": 2.6779935275080904, "percentage": 89.27, "elapsed_time": "2 days, 15:44:36", "remaining_time": "7:39:52"}
+{"current_steps": 3320, "total_steps": 3708, "loss": 0.1684, "lr": 1.6578194283005804e-06, "epoch": 2.686084142394822, "percentage": 89.54, "elapsed_time": "2 days, 15:56:09", "remaining_time": "7:28:19"}
+{"current_steps": 3330, "total_steps": 3708, "loss": 0.1679, "lr": 1.5745749212715794e-06, "epoch": 2.6941747572815533, "percentage": 89.81, "elapsed_time": "2 days, 16:07:42", "remaining_time": "7:16:46"}
+{"current_steps": 3340, "total_steps": 3708, "loss": 0.1741, "lr": 1.4934066258024182e-06, "epoch": 2.7022653721682848, "percentage": 90.08, "elapsed_time": "2 days, 16:19:15", "remaining_time": "7:05:12"}
+{"current_steps": 3350, "total_steps": 3708, "loss": 0.168, "lr": 1.4143217358950217e-06, "epoch": 2.7103559870550162, "percentage": 90.35, "elapsed_time": "2 days, 16:30:49", "remaining_time": "6:53:39"}
+{"current_steps": 3360, "total_steps": 3708, "loss": 0.1722, "lr": 1.3373272608976668e-06, "epoch": 2.7184466019417477, "percentage": 90.61, "elapsed_time": "2 days, 16:42:23", "remaining_time": "6:42:06"}
+{"current_steps": 3370, "total_steps": 3708, "loss": 0.1696, "lr": 1.2624300248836928e-06, "epoch": 2.726537216828479, "percentage": 90.88, "elapsed_time": "2 days, 16:53:56", "remaining_time": "6:30:32"}
+{"current_steps": 3380, "total_steps": 3708, "loss": 0.168, "lr": 1.1896366660467173e-06, "epoch": 2.73462783171521, "percentage": 91.15, "elapsed_time": "2 days, 17:05:29", "remaining_time": "6:18:59"}
+{"current_steps": 3390, "total_steps": 3708, "loss": 0.1665, "lr": 1.1189536361122799e-06, "epoch": 2.7427184466019416, "percentage": 91.42, "elapsed_time": "2 days, 17:17:02", "remaining_time": "6:07:26"}
+{"current_steps": 3400, "total_steps": 3708, "loss": 0.1623, "lr": 1.0503871997660036e-06, "epoch": 2.750809061488673, "percentage": 91.69, "elapsed_time": "2 days, 17:28:36", "remaining_time": "5:55:53"}
+{"current_steps": 3410, "total_steps": 3708, "loss": 0.166, "lr": 9.83943434098372e-07, "epoch": 2.7588996763754046, "percentage": 91.96, "elapsed_time": "2 days, 17:40:09", "remaining_time": "5:44:19"}
+{"current_steps": 3420, "total_steps": 3708, "loss": 0.1688, "lr": 9.196282280661023e-07, "epoch": 2.766990291262136, "percentage": 92.23, "elapsed_time": "2 days, 17:51:42", "remaining_time": "5:32:46"}
+{"current_steps": 3430, "total_steps": 3708, "loss": 0.1652, "lr": 8.574472819702029e-07, "epoch": 2.775080906148867, "percentage": 92.5, "elapsed_time": "2 days, 18:03:15", "remaining_time": "5:21:13"}
+{"current_steps": 3440, "total_steps": 3708, "loss": 0.1679, "lr": 7.974061069507571e-07, "epoch": 2.783171521035599, "percentage": 92.77, "elapsed_time": "2 days, 18:14:49", "remaining_time": "5:09:39"}
+{"current_steps": 3450, "total_steps": 3708, "loss": 0.1676, "lr": 7.395100244984604e-07, "epoch": 2.79126213592233, "percentage": 93.04, "elapsed_time": "2 days, 18:26:22", "remaining_time": "4:58:06"}
+{"current_steps": 3460, "total_steps": 3708, "loss": 0.1662, "lr": 6.837641659829807e-07, "epoch": 2.7993527508090614, "percentage": 93.31, "elapsed_time": "2 days, 18:37:55", "remaining_time": "4:46:33"}
+{"current_steps": 3470, "total_steps": 3708, "loss": 0.1669, "lr": 6.301734721981533e-07, "epoch": 2.807443365695793, "percentage": 93.58, "elapsed_time": "2 days, 18:49:29", "remaining_time": "4:35:00"}
+{"current_steps": 3480, "total_steps": 3708, "loss": 0.166, "lr": 5.787426929240808e-07, "epoch": 2.8155339805825244, "percentage": 93.85, "elapsed_time": "2 days, 19:01:02", "remaining_time": "4:23:26"}
+{"current_steps": 3490, "total_steps": 3708, "loss": 0.1683, "lr": 5.294763865061558e-07, "epoch": 2.823624595469256, "percentage": 94.12, "elapsed_time": "2 days, 19:12:36", "remaining_time": "4:11:53"}
+{"current_steps": 3500, "total_steps": 3708, "loss": 0.1706, "lr": 4.823789194510514e-07, "epoch": 2.831715210355987, "percentage": 94.39, "elapsed_time": "2 days, 19:24:09", "remaining_time": "4:00:20"}
+{"current_steps": 3510, "total_steps": 3708, "loss": 0.1688, "lr": 4.3745446603971064e-07, "epoch": 2.8398058252427183, "percentage": 94.66, "elapsed_time": "2 days, 19:35:42", "remaining_time": "3:48:47"}
+{"current_steps": 3520, "total_steps": 3708, "loss": 0.1638, "lr": 3.947070079573872e-07, "epoch": 2.8478964401294498, "percentage": 94.93, "elapsed_time": "2 days, 19:47:15", "remaining_time": "3:37:13"}
+{"current_steps": 3530, "total_steps": 3708, "loss": 0.171, "lr": 3.541403339407279e-07, "epoch": 2.855987055016181, "percentage": 95.2, "elapsed_time": "2 days, 19:58:48", "remaining_time": "3:25:40"}
+{"current_steps": 3540, "total_steps": 3708, "loss": 0.1698, "lr": 3.1575803944199624e-07, "epoch": 2.8640776699029127, "percentage": 95.47, "elapsed_time": "2 days, 20:10:21", "remaining_time": "3:14:07"}
+{"current_steps": 3550, "total_steps": 3708, "loss": 0.1659, "lr": 2.7956352631038906e-07, "epoch": 2.872168284789644, "percentage": 95.74, "elapsed_time": "2 days, 20:21:55", "remaining_time": "3:02:33"}
+{"current_steps": 3560, "total_steps": 3708, "loss": 0.1705, "lr": 2.4556000249054133e-07, "epoch": 2.8802588996763756, "percentage": 96.01, "elapsed_time": "2 days, 20:33:28", "remaining_time": "2:51:00"}
+{"current_steps": 3570, "total_steps": 3708, "loss": 0.1676, "lr": 2.1375048173818412e-07, "epoch": 2.8883495145631066, "percentage": 96.28, "elapsed_time": "2 days, 20:45:02", "remaining_time": "2:39:27"}
+{"current_steps": 3580, "total_steps": 3708, "loss": 0.1639, "lr": 1.8413778335305e-07, "epoch": 2.896440129449838, "percentage": 96.55, "elapsed_time": "2 days, 20:56:35", "remaining_time": "2:27:54"}
+{"current_steps": 3590, "total_steps": 3708, "loss": 0.169, "lr": 1.567245319290006e-07, "epoch": 2.9045307443365695, "percentage": 96.82, "elapsed_time": "2 days, 21:08:08", "remaining_time": "2:16:20"}
+{"current_steps": 3600, "total_steps": 3708, "loss": 0.163, "lr": 1.315131571213879e-07, "epoch": 2.912621359223301, "percentage": 97.09, "elapsed_time": "2 days, 21:19:42", "remaining_time": "2:04:47"}
+{"current_steps": 3610, "total_steps": 3708, "loss": 0.1669, "lr": 1.0850589343172624e-07, "epoch": 2.9207119741100325, "percentage": 97.36, "elapsed_time": "2 days, 21:31:16", "remaining_time": "1:53:14"}
+{"current_steps": 3620, "total_steps": 3708, "loss": 0.1684, "lr": 8.770478000964532e-08, "epoch": 2.9288025889967635, "percentage": 97.63, "elapsed_time": "2 days, 21:42:49", "remaining_time": "1:41:40"}
+{"current_steps": 3630, "total_steps": 3708, "loss": 0.1694, "lr": 6.911166047215145e-08, "epoch": 2.9368932038834954, "percentage": 97.9, "elapsed_time": "2 days, 21:54:22", "remaining_time": "1:30:07"}
+{"current_steps": 3640, "total_steps": 3708, "loss": 0.1675, "lr": 5.272818274023872e-08, "epoch": 2.9449838187702264, "percentage": 98.17, "elapsed_time": "2 days, 22:05:55", "remaining_time": "1:18:34"}
+{"current_steps": 3650, "total_steps": 3708, "loss": 0.1679, "lr": 3.855579889282257e-08, "epoch": 2.953074433656958, "percentage": 98.44, "elapsed_time": "2 days, 22:17:28", "remaining_time": "1:07:01"}
+{"current_steps": 3660, "total_steps": 3708, "loss": 0.1677, "lr": 2.6595765038045507e-08, "epoch": 2.9611650485436893, "percentage": 98.71, "elapsed_time": "2 days, 22:29:01", "remaining_time": "0:55:27"}
+{"current_steps": 3670, "total_steps": 3708, "loss": 0.165, "lr": 1.6849141201946693e-08, "epoch": 2.969255663430421, "percentage": 98.98, "elapsed_time": "2 days, 22:40:34", "remaining_time": "0:43:54"}
+{"current_steps": 3680, "total_steps": 3708, "loss": 0.1713, "lr": 9.316791234506572e-09, "epoch": 2.9773462783171523, "percentage": 99.24, "elapsed_time": "2 days, 22:52:08", "remaining_time": "0:32:21"}
+{"current_steps": 3690, "total_steps": 3708, "loss": 0.1665, "lr": 3.999382733096968e-09, "epoch": 2.9854368932038833, "percentage": 99.51, "elapsed_time": "2 days, 23:03:41", "remaining_time": "0:20:47"}
+{"current_steps": 3700, "total_steps": 3708, "loss": 0.1686, "lr": 8.973869832895609e-10, "epoch": 2.9935275080906147, "percentage": 99.78, "elapsed_time": "2 days, 23:15:14", "remaining_time": "0:09:14"}
+{"current_steps": 3708, "total_steps": 3708, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "2 days, 23:25:07", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,2633 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3708,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008090614886731391,
+      "grad_norm": 22.52252141115457,
+      "learning_rate": 1.2129380053908356e-06,
+      "loss": 2.5582,
+      "step": 10
+    },
+    {
+      "epoch": 0.016181229773462782,
+      "grad_norm": 11.441648898342923,
+      "learning_rate": 2.560646900269542e-06,
+      "loss": 2.2025,
+      "step": 20
+    },
+    {
+      "epoch": 0.024271844660194174,
+      "grad_norm": 5.370882200215843,
+      "learning_rate": 3.908355795148248e-06,
+      "loss": 1.668,
+      "step": 30
+    },
+    {
+      "epoch": 0.032362459546925564,
+      "grad_norm": 2.0115313434456827,
+      "learning_rate": 5.256064690026954e-06,
+      "loss": 1.1266,
+      "step": 40
+    },
+    {
+      "epoch": 0.040453074433656956,
+      "grad_norm": 1.2036774860036918,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.7799,
+      "step": 50
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 0.8106871294631399,
+      "learning_rate": 7.951482479784367e-06,
+      "loss": 0.5907,
+      "step": 60
+    },
+    {
+      "epoch": 0.05663430420711974,
+      "grad_norm": 0.6572037767143685,
+      "learning_rate": 9.299191374663074e-06,
+      "loss": 0.4977,
+      "step": 70
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 0.9743925361281945,
+      "learning_rate": 1.0646900269541779e-05,
+      "loss": 0.4728,
+      "step": 80
+    },
+    {
+      "epoch": 0.07281553398058252,
+      "grad_norm": 0.5302813737105591,
+      "learning_rate": 1.1994609164420486e-05,
+      "loss": 0.446,
+      "step": 90
+    },
+    {
+      "epoch": 0.08090614886731391,
+      "grad_norm": 0.5504892707748702,
+      "learning_rate": 1.3342318059299191e-05,
+      "loss": 0.4164,
+      "step": 100
+    },
+    {
+      "epoch": 0.0889967637540453,
+      "grad_norm": 0.46842577857609835,
+      "learning_rate": 1.4690026954177898e-05,
+      "loss": 0.412,
+      "step": 110
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 0.5228315101871648,
+      "learning_rate": 1.6037735849056604e-05,
+      "loss": 0.3975,
+      "step": 120
+    },
+    {
+      "epoch": 0.10517799352750809,
+      "grad_norm": 0.5623706583567197,
+      "learning_rate": 1.738544474393531e-05,
+      "loss": 0.3857,
+      "step": 130
+    },
+    {
+      "epoch": 0.11326860841423948,
+      "grad_norm": 0.3506071062876399,
+      "learning_rate": 1.8733153638814018e-05,
+      "loss": 0.3837,
+      "step": 140
+    },
+    {
+      "epoch": 0.12135922330097088,
+      "grad_norm": 0.491171903870991,
+      "learning_rate": 2.0080862533692725e-05,
+      "loss": 0.3721,
+      "step": 150
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 0.5505284472221431,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.3734,
+      "step": 160
+    },
+    {
+      "epoch": 0.13754045307443366,
+      "grad_norm": 0.8767975628519997,
+      "learning_rate": 2.2776280323450135e-05,
+      "loss": 0.365,
+      "step": 170
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 0.5343891120951508,
+      "learning_rate": 2.4123989218328842e-05,
+      "loss": 0.3631,
+      "step": 180
+    },
+    {
+      "epoch": 0.15372168284789645,
+      "grad_norm": 0.47266629407427385,
+      "learning_rate": 2.547169811320755e-05,
+      "loss": 0.3509,
+      "step": 190
+    },
+    {
+      "epoch": 0.16181229773462782,
+      "grad_norm": 0.49039579917599835,
+      "learning_rate": 2.6819407008086256e-05,
+      "loss": 0.3474,
+      "step": 200
+    },
+    {
+      "epoch": 0.16990291262135923,
+      "grad_norm": 0.5095673082261557,
+      "learning_rate": 2.8167115902964963e-05,
+      "loss": 0.3439,
+      "step": 210
+    },
+    {
+      "epoch": 0.1779935275080906,
+      "grad_norm": 0.7559655324445135,
+      "learning_rate": 2.9514824797843667e-05,
+      "loss": 0.3477,
+      "step": 220
+    },
+    {
+      "epoch": 0.18608414239482202,
+      "grad_norm": 0.717837164870727,
+      "learning_rate": 3.086253369272237e-05,
+      "loss": 0.3424,
+      "step": 230
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 0.44154340671189124,
+      "learning_rate": 3.221024258760108e-05,
+      "loss": 0.3432,
+      "step": 240
+    },
+    {
+      "epoch": 0.2022653721682848,
+      "grad_norm": 0.645302977385022,
+      "learning_rate": 3.355795148247979e-05,
+      "loss": 0.3378,
+      "step": 250
+    },
+    {
+      "epoch": 0.21035598705501618,
+      "grad_norm": 0.45392402289172107,
+      "learning_rate": 3.490566037735849e-05,
+      "loss": 0.34,
+      "step": 260
+    },
+    {
+      "epoch": 0.21844660194174756,
+      "grad_norm": 0.7064296610826926,
+      "learning_rate": 3.62533692722372e-05,
+      "loss": 0.339,
+      "step": 270
+    },
+    {
+      "epoch": 0.22653721682847897,
+      "grad_norm": 0.5611292380408083,
+      "learning_rate": 3.76010781671159e-05,
+      "loss": 0.3284,
+      "step": 280
+    },
+    {
+      "epoch": 0.23462783171521034,
+      "grad_norm": 0.5448155574311832,
+      "learning_rate": 3.894878706199461e-05,
+      "loss": 0.3291,
+      "step": 290
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 0.43795871099809547,
+      "learning_rate": 4.0296495956873316e-05,
+      "loss": 0.3206,
+      "step": 300
+    },
+    {
+      "epoch": 0.25080906148867316,
+      "grad_norm": 0.5584253534316864,
+      "learning_rate": 4.164420485175202e-05,
+      "loss": 0.3276,
+      "step": 310
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 0.5617857932874714,
+      "learning_rate": 4.299191374663073e-05,
+      "loss": 0.3258,
+      "step": 320
+    },
+    {
+      "epoch": 0.2669902912621359,
+      "grad_norm": 0.5441488728977509,
+      "learning_rate": 4.433962264150944e-05,
+      "loss": 0.3227,
+      "step": 330
+    },
+    {
+      "epoch": 0.2750809061488673,
+      "grad_norm": 0.8386999082473845,
+      "learning_rate": 4.5687331536388144e-05,
+      "loss": 0.3196,
+      "step": 340
+    },
+    {
+      "epoch": 0.28317152103559873,
+      "grad_norm": 0.43608786652562637,
+      "learning_rate": 4.703504043126685e-05,
+      "loss": 0.3191,
+      "step": 350
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.4937683097487802,
+      "learning_rate": 4.838274932614555e-05,
+      "loss": 0.3255,
+      "step": 360
+    },
+    {
+      "epoch": 0.2993527508090615,
+      "grad_norm": 0.5061197818365496,
+      "learning_rate": 4.973045822102426e-05,
+      "loss": 0.3119,
+      "step": 370
+    },
+    {
+      "epoch": 0.3074433656957929,
+      "grad_norm": 0.4744335618999143,
+      "learning_rate": 4.9999290952604396e-05,
+      "loss": 0.3191,
+      "step": 380
+    },
+    {
+      "epoch": 0.3155339805825243,
+      "grad_norm": 0.5037259460580787,
+      "learning_rate": 4.9996410516491115e-05,
+      "loss": 0.3098,
+      "step": 390
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 0.5710713036434062,
+      "learning_rate": 4.9991314631296585e-05,
+      "loss": 0.3096,
+      "step": 400
+    },
+    {
+      "epoch": 0.33171521035598706,
+      "grad_norm": 0.33451540591505297,
+      "learning_rate": 4.9984003748672604e-05,
+      "loss": 0.3095,
+      "step": 410
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 0.4340252206831366,
+      "learning_rate": 4.997447851658774e-05,
+      "loss": 0.3076,
+      "step": 420
+    },
+    {
+      "epoch": 0.3478964401294498,
+      "grad_norm": 0.3838413262491291,
+      "learning_rate": 4.9962739779269887e-05,
+      "loss": 0.3067,
+      "step": 430
+    },
+    {
+      "epoch": 0.3559870550161812,
+      "grad_norm": 0.42248192127450584,
+      "learning_rate": 4.9948788577131414e-05,
+      "loss": 0.2989,
+      "step": 440
+    },
+    {
+      "epoch": 0.3640776699029126,
+      "grad_norm": 0.4026141748836561,
+      "learning_rate": 4.993262614667696e-05,
+      "loss": 0.3018,
+      "step": 450
+    },
+    {
+      "epoch": 0.37216828478964403,
+      "grad_norm": 0.4295230581487718,
+      "learning_rate": 4.9914253920393884e-05,
+      "loss": 0.3048,
+      "step": 460
+    },
+    {
+      "epoch": 0.3802588996763754,
+      "grad_norm": 0.3384646839950234,
+      "learning_rate": 4.9893673526625265e-05,
+      "loss": 0.3025,
+      "step": 470
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.41348113959511545,
+      "learning_rate": 4.987088678942555e-05,
+      "loss": 0.2997,
+      "step": 480
+    },
+    {
+      "epoch": 0.3964401294498382,
+      "grad_norm": 0.3673430436712548,
+      "learning_rate": 4.984589572839897e-05,
+      "loss": 0.2981,
+      "step": 490
+    },
+    {
+      "epoch": 0.4045307443365696,
+      "grad_norm": 0.3637541700718482,
+      "learning_rate": 4.9818702558520485e-05,
+      "loss": 0.2994,
+      "step": 500
+    },
+    {
+      "epoch": 0.41262135922330095,
+      "grad_norm": 0.3667744606459147,
+      "learning_rate": 4.978930968993946e-05,
+      "loss": 0.3027,
+      "step": 510
+    },
+    {
+      "epoch": 0.42071197411003236,
+      "grad_norm": 0.3783856834139639,
+      "learning_rate": 4.9757719727766085e-05,
+      "loss": 0.2995,
+      "step": 520
+    },
+    {
+      "epoch": 0.42880258899676377,
+      "grad_norm": 0.39358597558975966,
+      "learning_rate": 4.972393547184046e-05,
+      "loss": 0.2968,
+      "step": 530
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.3961355130202433,
+      "learning_rate": 4.968795991648446e-05,
+      "loss": 0.2951,
+      "step": 540
+    },
+    {
+      "epoch": 0.4449838187702265,
+      "grad_norm": 0.37157203976493997,
+      "learning_rate": 4.9649796250236344e-05,
+      "loss": 0.2842,
+      "step": 550
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 0.2840234936379272,
+      "learning_rate": 4.960944785556814e-05,
+      "loss": 0.2862,
+      "step": 560
+    },
+    {
+      "epoch": 0.46116504854368934,
+      "grad_norm": 0.320746304504261,
+      "learning_rate": 4.956691830858585e-05,
+      "loss": 0.2925,
+      "step": 570
+    },
+    {
+      "epoch": 0.4692556634304207,
+      "grad_norm": 0.3627941299734378,
+      "learning_rate": 4.952221137871252e-05,
+      "loss": 0.2884,
+      "step": 580
+    },
+    {
+      "epoch": 0.4773462783171521,
+      "grad_norm": 0.44860484414052815,
+      "learning_rate": 4.947533102835413e-05,
+      "loss": 0.2937,
+      "step": 590
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.3404390735893878,
+      "learning_rate": 4.942628141254843e-05,
+      "loss": 0.2888,
+      "step": 600
+    },
+    {
+      "epoch": 0.4935275080906149,
+      "grad_norm": 0.3346590749564513,
+      "learning_rate": 4.937506687859666e-05,
+      "loss": 0.2839,
+      "step": 610
+    },
+    {
+      "epoch": 0.5016181229773463,
+      "grad_norm": 0.3540605558733981,
+      "learning_rate": 4.932169196567824e-05,
+      "loss": 0.2926,
+      "step": 620
+    },
+    {
+      "epoch": 0.5097087378640777,
+      "grad_norm": 0.3225115085834403,
+      "learning_rate": 4.9266161404448454e-05,
+      "loss": 0.2914,
+      "step": 630
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 0.333757773957392,
+      "learning_rate": 4.920848011661919e-05,
+      "loss": 0.2844,
+      "step": 640
+    },
+    {
+      "epoch": 0.5258899676375405,
+      "grad_norm": 0.3691189283201263,
+      "learning_rate": 4.914865321452274e-05,
+      "loss": 0.2838,
+      "step": 650
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 0.41635131408315823,
+      "learning_rate": 4.908668600065862e-05,
+      "loss": 0.2853,
+      "step": 660
+    },
+    {
+      "epoch": 0.5420711974110033,
+      "grad_norm": 0.3161547689784877,
+      "learning_rate": 4.90225839672237e-05,
+      "loss": 0.2844,
+      "step": 670
+    },
+    {
+      "epoch": 0.5501618122977346,
+      "grad_norm": 0.32264889950016085,
+      "learning_rate": 4.8956352795625325e-05,
+      "loss": 0.2912,
+      "step": 680
+    },
+    {
+      "epoch": 0.558252427184466,
+      "grad_norm": 0.3354534030777558,
+      "learning_rate": 4.8887998355977886e-05,
+      "loss": 0.2826,
+      "step": 690
+    },
+    {
+      "epoch": 0.5663430420711975,
+      "grad_norm": 0.3409096712904377,
+      "learning_rate": 4.881752670658244e-05,
+      "loss": 0.2799,
+      "step": 700
+    },
+    {
+      "epoch": 0.5744336569579288,
+      "grad_norm": 0.3012793393888447,
+      "learning_rate": 4.87449440933898e-05,
+      "loss": 0.2713,
+      "step": 710
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.3127062042429517,
+      "learning_rate": 4.867025694944698e-05,
+      "loss": 0.2813,
+      "step": 720
+    },
+    {
+      "epoch": 0.5906148867313916,
+      "grad_norm": 0.3269471051340985,
+      "learning_rate": 4.859347189432699e-05,
+      "loss": 0.28,
+      "step": 730
+    },
+    {
+      "epoch": 0.598705501618123,
+      "grad_norm": 0.32543025148595867,
+      "learning_rate": 4.8514595733542144e-05,
+      "loss": 0.2709,
+      "step": 740
+    },
+    {
+      "epoch": 0.6067961165048543,
+      "grad_norm": 0.3147551999550132,
+      "learning_rate": 4.8433635457940915e-05,
+      "loss": 0.2814,
+      "step": 750
+    },
+    {
+      "epoch": 0.6148867313915858,
+      "grad_norm": 0.26606594884988966,
+      "learning_rate": 4.8350598243088283e-05,
+      "loss": 0.2776,
+      "step": 760
+    },
+    {
+      "epoch": 0.6229773462783171,
+      "grad_norm": 0.3007757906909445,
+      "learning_rate": 4.8265491448629804e-05,
+      "loss": 0.2767,
+      "step": 770
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.3795450058305756,
+      "learning_rate": 4.817832261763928e-05,
+      "loss": 0.2772,
+      "step": 780
+    },
+    {
+      "epoch": 0.63915857605178,
+      "grad_norm": 0.3106723870622729,
+      "learning_rate": 4.8089099475950257e-05,
+      "loss": 0.276,
+      "step": 790
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 0.34373131058866446,
+      "learning_rate": 4.7997829931471225e-05,
+      "loss": 0.2765,
+      "step": 800
+    },
+    {
+      "epoch": 0.6553398058252428,
+      "grad_norm": 0.2846997869981113,
+      "learning_rate": 4.7904522073484786e-05,
+      "loss": 0.2778,
+      "step": 810
+    },
+    {
+      "epoch": 0.6634304207119741,
+      "grad_norm": 0.31767113534168995,
+      "learning_rate": 4.780918417193065e-05,
+      "loss": 0.2775,
+      "step": 820
+    },
+    {
+      "epoch": 0.6715210355987055,
+      "grad_norm": 0.31169855988969003,
+      "learning_rate": 4.7711824676672726e-05,
+      "loss": 0.2733,
+      "step": 830
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.3175387528791534,
+      "learning_rate": 4.76124522167501e-05,
+      "loss": 0.2746,
+      "step": 840
+    },
+    {
+      "epoch": 0.6877022653721683,
+      "grad_norm": 0.2975263968714564,
+      "learning_rate": 4.751107559961238e-05,
+      "loss": 0.2745,
+      "step": 850
+    },
+    {
+      "epoch": 0.6957928802588996,
+      "grad_norm": 0.29446496624170493,
+      "learning_rate": 4.740770381033894e-05,
+      "loss": 0.275,
+      "step": 860
+    },
+    {
+      "epoch": 0.7038834951456311,
+      "grad_norm": 0.2973639432877628,
+      "learning_rate": 4.730234601084268e-05,
+      "loss": 0.2716,
+      "step": 870
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 0.2497924694108393,
+      "learning_rate": 4.719501153905793e-05,
+      "loss": 0.2673,
+      "step": 880
+    },
+    {
+      "epoch": 0.7200647249190939,
+      "grad_norm": 0.2948558193473281,
+      "learning_rate": 4.7085709908112866e-05,
+      "loss": 0.2731,
+      "step": 890
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.2459558041018532,
+      "learning_rate": 4.6974450805486305e-05,
+      "loss": 0.2639,
+      "step": 900
+    },
+    {
+      "epoch": 0.7362459546925566,
+      "grad_norm": 0.2559758327802977,
+      "learning_rate": 4.686124409214917e-05,
+      "loss": 0.2715,
+      "step": 910
+    },
+    {
+      "epoch": 0.7443365695792881,
+      "grad_norm": 0.2811398346458149,
+      "learning_rate": 4.674609980169042e-05,
+      "loss": 0.2712,
+      "step": 920
+    },
+    {
+      "epoch": 0.7524271844660194,
+      "grad_norm": 0.2686221572337129,
+      "learning_rate": 4.662902813942784e-05,
+      "loss": 0.272,
+      "step": 930
+    },
+    {
+      "epoch": 0.7605177993527508,
+      "grad_norm": 0.2755997369035635,
+      "learning_rate": 4.651003948150349e-05,
+      "loss": 0.2689,
+      "step": 940
+    },
+    {
+      "epoch": 0.7686084142394822,
+      "grad_norm": 0.2672412475643565,
+      "learning_rate": 4.638914437396408e-05,
+      "loss": 0.2693,
+      "step": 950
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.23259525487399887,
+      "learning_rate": 4.626635353182626e-05,
+      "loss": 0.2659,
+      "step": 960
+    },
+    {
+      "epoch": 0.7847896440129449,
+      "grad_norm": 0.26440852988259955,
+      "learning_rate": 4.614167783812694e-05,
+      "loss": 0.2753,
+      "step": 970
+    },
+    {
+      "epoch": 0.7928802588996764,
+      "grad_norm": 0.2786648778182051,
+      "learning_rate": 4.601512834295874e-05,
+      "loss": 0.272,
+      "step": 980
+    },
+    {
+      "epoch": 0.8009708737864077,
+      "grad_norm": 0.2475631464266887,
+      "learning_rate": 4.588671626249057e-05,
+      "loss": 0.2631,
+      "step": 990
+    },
+    {
+      "epoch": 0.8090614886731392,
+      "grad_norm": 0.23380841398688618,
+      "learning_rate": 4.5756452977973585e-05,
+      "loss": 0.2609,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8171521035598706,
+      "grad_norm": 0.27565297705248776,
+      "learning_rate": 4.56243500347324e-05,
+      "loss": 0.2651,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.29059051987658346,
+      "learning_rate": 4.549041914114188e-05,
+      "loss": 0.2687,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.29594618957533503,
+      "learning_rate": 4.535467216758936e-05,
+      "loss": 0.2666,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 0.3931170619188329,
+      "learning_rate": 4.5217121145422616e-05,
+      "loss": 0.2631,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8495145631067961,
+      "grad_norm": 0.2857742902635426,
+      "learning_rate": 4.5077778265883477e-05,
+      "loss": 0.2632,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8576051779935275,
+      "grad_norm": 0.24489808715354155,
+      "learning_rate": 4.4936655879027336e-05,
+      "loss": 0.2634,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8656957928802589,
+      "grad_norm": 0.26376201661023657,
+      "learning_rate": 4.479376649262855e-05,
+      "loss": 0.266,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.25183442403581696,
+      "learning_rate": 4.464912277107185e-05,
+      "loss": 0.2658,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8818770226537217,
+      "grad_norm": 0.2613869002025676,
+      "learning_rate": 4.450273753422992e-05,
+      "loss": 0.2653,
+      "step": 1090
+    },
+    {
+      "epoch": 0.889967637540453,
+      "grad_norm": 0.29304343706553704,
+      "learning_rate": 4.435462375632711e-05,
+      "loss": 0.2651,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8980582524271845,
+      "grad_norm": 0.263122059677539,
+      "learning_rate": 4.420479456478957e-05,
+      "loss": 0.2621,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 0.24882087207880094,
+      "learning_rate": 4.405326323908178e-05,
+      "loss": 0.2635,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9142394822006472,
+      "grad_norm": 0.2744368878036377,
+      "learning_rate": 4.390004320952947e-05,
+      "loss": 0.2639,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.25617953359336987,
+      "learning_rate": 4.374514805612942e-05,
+      "loss": 0.2667,
+      "step": 1140
+    },
+    {
+      "epoch": 0.93042071197411,
+      "grad_norm": 0.26969276785611856,
+      "learning_rate": 4.358859150734576e-05,
+      "loss": 0.2592,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9385113268608414,
+      "grad_norm": 0.2703864111603607,
+      "learning_rate": 4.343038743889324e-05,
+      "loss": 0.2602,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9466019417475728,
+      "grad_norm": 0.3115866006041078,
+      "learning_rate": 4.3270549872507415e-05,
+      "loss": 0.26,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9546925566343042,
+      "grad_norm": 0.2611547498092126,
+      "learning_rate": 4.3109092974701895e-05,
+      "loss": 0.2591,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9627831715210357,
+      "grad_norm": 0.26931477864354814,
+      "learning_rate": 4.2946031055512733e-05,
+      "loss": 0.26,
+      "step": 1190
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.2578523252517678,
+      "learning_rate": 4.2781378567230145e-05,
+      "loss": 0.2633,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9789644012944984,
+      "grad_norm": 0.27451378253592124,
+      "learning_rate": 4.2615150103117576e-05,
+      "loss": 0.2598,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9870550161812298,
+      "grad_norm": 0.2609329028667746,
+      "learning_rate": 4.24473603961183e-05,
+      "loss": 0.2587,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9951456310679612,
+      "grad_norm": 0.2280958023284859,
+      "learning_rate": 4.227802431754961e-05,
+      "loss": 0.2589,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0032362459546926,
+      "grad_norm": 0.30070050651075947,
+      "learning_rate": 4.210715687578481e-05,
+      "loss": 0.2468,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0113268608414239,
+      "grad_norm": 0.26616355843357287,
+      "learning_rate": 4.193477321492293e-05,
+      "loss": 0.2219,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.2713492665233822,
+      "learning_rate": 4.176088861344657e-05,
+      "loss": 0.2251,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0275080906148868,
+      "grad_norm": 0.21687794735432822,
+      "learning_rate": 4.158551848286773e-05,
+      "loss": 0.2178,
+      "step": 1270
+    },
+    {
+      "epoch": 1.035598705501618,
+      "grad_norm": 0.2539615253588423,
+      "learning_rate": 4.140867836636189e-05,
+      "loss": 0.2289,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0436893203883495,
+      "grad_norm": 0.24657642734847438,
+      "learning_rate": 4.1230383937390374e-05,
+      "loss": 0.2258,
+      "step": 1290
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 0.2609125461360237,
+      "learning_rate": 4.1050650998311215e-05,
+      "loss": 0.2216,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0598705501618122,
+      "grad_norm": 0.3630380994135679,
+      "learning_rate": 4.086949547897862e-05,
+      "loss": 0.2225,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.2504138070534478,
+      "learning_rate": 4.068693343533103e-05,
+      "loss": 0.2227,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0760517799352751,
+      "grad_norm": 0.3126812459563885,
+      "learning_rate": 4.050298104796812e-05,
+      "loss": 0.2268,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0841423948220066,
+      "grad_norm": 0.2257523743221347,
+      "learning_rate": 4.0317654620716704e-05,
+      "loss": 0.2306,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0922330097087378,
+      "grad_norm": 0.21960504136006673,
+      "learning_rate": 4.013097057918566e-05,
+      "loss": 0.2254,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1003236245954693,
+      "grad_norm": 0.250009270408735,
+      "learning_rate": 3.9942945469310175e-05,
+      "loss": 0.2279,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1084142394822007,
+      "grad_norm": 0.23480544241701537,
+      "learning_rate": 3.9753595955885264e-05,
+      "loss": 0.2256,
+      "step": 1370
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.23720387815460467,
+      "learning_rate": 3.9562938821088705e-05,
+      "loss": 0.2235,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1245954692556634,
+      "grad_norm": 0.23437981665878366,
+      "learning_rate": 3.9370990962993695e-05,
+      "loss": 0.2237,
+      "step": 1390
+    },
+    {
+      "epoch": 1.132686084142395,
+      "grad_norm": 0.2286079680729574,
+      "learning_rate": 3.9177769394071086e-05,
+      "loss": 0.2215,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1407766990291262,
+      "grad_norm": 0.22687828650289346,
+      "learning_rate": 3.898329123968163e-05,
+      "loss": 0.2254,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1488673139158576,
+      "grad_norm": 0.2224235012119624,
+      "learning_rate": 3.87875737365581e-05,
+      "loss": 0.2184,
+      "step": 1420
+    },
+    {
+      "epoch": 1.156957928802589,
+      "grad_norm": 0.20195576636612733,
+      "learning_rate": 3.8590634231277626e-05,
+      "loss": 0.2176,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.20327918718818466,
+      "learning_rate": 3.8392490178724184e-05,
+      "loss": 0.2207,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1731391585760518,
+      "grad_norm": 0.21744881881616449,
+      "learning_rate": 3.8193159140541645e-05,
+      "loss": 0.223,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 0.30658821143823883,
+      "learning_rate": 3.7992658783577215e-05,
+      "loss": 0.2264,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1893203883495145,
+      "grad_norm": 0.25174532799295857,
+      "learning_rate": 3.779100687831563e-05,
+      "loss": 0.2216,
+      "step": 1470
+    },
+    {
+      "epoch": 1.197411003236246,
+      "grad_norm": 0.2561981869189091,
+      "learning_rate": 3.758822129730415e-05,
+      "loss": 0.2201,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2055016181229774,
+      "grad_norm": 0.22464952614110853,
+      "learning_rate": 3.738432001356851e-05,
+      "loss": 0.2239,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.2100829197882277,
+      "learning_rate": 3.7179321099019916e-05,
+      "loss": 0.2281,
+      "step": 1500
+    },
+    {
+      "epoch": 1.22168284789644,
+      "grad_norm": 0.2469541669228544,
+      "learning_rate": 3.6973242722853365e-05,
+      "loss": 0.225,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2297734627831716,
+      "grad_norm": 0.23312096116267617,
+      "learning_rate": 3.6766103149937295e-05,
+      "loss": 0.2246,
+      "step": 1520
+    },
+    {
+      "epoch": 1.237864077669903,
+      "grad_norm": 0.20677868772581298,
+      "learning_rate": 3.655792073919471e-05,
+      "loss": 0.2249,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 0.2563183131089248,
+      "learning_rate": 3.634871394197607e-05,
+      "loss": 0.2263,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2540453074433657,
+      "grad_norm": 0.22128405881374966,
+      "learning_rate": 3.6138501300423934e-05,
+      "loss": 0.2211,
+      "step": 1550
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.21006117612744632,
+      "learning_rate": 3.592730144582948e-05,
+      "loss": 0.2243,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2702265372168284,
+      "grad_norm": 0.1975344463355082,
+      "learning_rate": 3.571513309698131e-05,
+      "loss": 0.2202,
+      "step": 1570
+    },
+    {
+      "epoch": 1.27831715210356,
+      "grad_norm": 0.20997495813930359,
+      "learning_rate": 3.5502015058506335e-05,
+      "loss": 0.2229,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2864077669902914,
+      "grad_norm": 0.2332593631600526,
+      "learning_rate": 3.528796621920307e-05,
+      "loss": 0.219,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2944983818770226,
+      "grad_norm": 0.22815002833576079,
+      "learning_rate": 3.50730055503676e-05,
+      "loss": 0.222,
+      "step": 1600
+    },
+    {
+      "epoch": 1.302588996763754,
+      "grad_norm": 0.20437252443820447,
+      "learning_rate": 3.485715210411204e-05,
+      "loss": 0.2262,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.2395647563898874,
+      "learning_rate": 3.4640425011676034e-05,
+      "loss": 0.2226,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3187702265372168,
+      "grad_norm": 0.22072548502378347,
+      "learning_rate": 3.442284348173106e-05,
+      "loss": 0.2254,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3268608414239482,
+      "grad_norm": 0.21606415951025024,
+      "learning_rate": 3.420442679867796e-05,
+      "loss": 0.2244,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3349514563106797,
+      "grad_norm": 0.27863651532858746,
+      "learning_rate": 3.398519432093782e-05,
+      "loss": 0.2227,
+      "step": 1650
+    },
+    {
+      "epoch": 1.343042071197411,
+      "grad_norm": 0.2202632134109216,
+      "learning_rate": 3.376516547923614e-05,
+      "loss": 0.2195,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3511326860841424,
+      "grad_norm": 0.21524751059743225,
+      "learning_rate": 3.3544359774880714e-05,
+      "loss": 0.2203,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.20473236269563994,
+      "learning_rate": 3.3322796778033204e-05,
+      "loss": 0.2151,
+      "step": 1680
+    },
+    {
+      "epoch": 1.367313915857605,
+      "grad_norm": 0.21469153751522554,
+      "learning_rate": 3.3100496125974624e-05,
+      "loss": 0.216,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 0.19954679234440933,
+      "learning_rate": 3.2877477521364895e-05,
+      "loss": 0.2162,
+      "step": 1700
+    },
+    {
+      "epoch": 1.383495145631068,
+      "grad_norm": 0.18861748166722786,
+      "learning_rate": 3.2653760730496555e-05,
+      "loss": 0.2187,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3915857605177995,
+      "grad_norm": 0.20341053182705085,
+      "learning_rate": 3.242936558154285e-05,
+      "loss": 0.225,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3996763754045307,
+      "grad_norm": 0.19372792735024014,
+      "learning_rate": 3.2204311962800426e-05,
+      "loss": 0.2239,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.27663774087598847,
+      "learning_rate": 3.197861982092651e-05,
+      "loss": 0.2201,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4158576051779934,
+      "grad_norm": 0.20101300631018146,
+      "learning_rate": 3.175230915917108e-05,
+      "loss": 0.2234,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4239482200647249,
+      "grad_norm": 0.20722987740988344,
+      "learning_rate": 3.152540003560398e-05,
+      "loss": 0.219,
+      "step": 1760
+    },
+    {
+      "epoch": 1.4320388349514563,
+      "grad_norm": 0.23987446750061972,
+      "learning_rate": 3.129791256133712e-05,
+      "loss": 0.2158,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 0.2072184752121984,
+      "learning_rate": 3.106986689874204e-05,
+      "loss": 0.2208,
+      "step": 1780
+    },
+    {
+      "epoch": 1.448220064724919,
+      "grad_norm": 0.19280709770923388,
+      "learning_rate": 3.0841283259662875e-05,
+      "loss": 0.2227,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.20536031119127787,
+      "learning_rate": 3.0612181903625014e-05,
+      "loss": 0.222,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4644012944983817,
+      "grad_norm": 0.22626011728625453,
+      "learning_rate": 3.0382583136039444e-05,
+      "loss": 0.2174,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4724919093851132,
+      "grad_norm": 0.2105569387760891,
+      "learning_rate": 3.015250730640308e-05,
+      "loss": 0.2167,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4805825242718447,
+      "grad_norm": 0.20632917153940236,
+      "learning_rate": 2.9921974806495178e-05,
+      "loss": 0.2141,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4886731391585761,
+      "grad_norm": 0.2021951348802143,
+      "learning_rate": 2.969100606856998e-05,
+      "loss": 0.2189,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4967637540453074,
+      "grad_norm": 0.21580708092456813,
+      "learning_rate": 2.9459621563545825e-05,
+      "loss": 0.2236,
+      "step": 1850
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.19252241914699963,
+      "learning_rate": 2.9227841799190775e-05,
+      "loss": 0.2182,
+      "step": 1860
+    },
+    {
+      "epoch": 1.51294498381877,
+      "grad_norm": 0.28670687006718837,
+      "learning_rate": 2.8995687318304975e-05,
+      "loss": 0.2224,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5210355987055015,
+      "grad_norm": 0.19972242558140968,
+      "learning_rate": 2.8763178696899995e-05,
+      "loss": 0.219,
+      "step": 1880
+    },
+    {
+      "epoch": 1.529126213592233,
+      "grad_norm": 0.21361706883180026,
+      "learning_rate": 2.853033654237507e-05,
+      "loss": 0.2207,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5372168284789645,
+      "grad_norm": 0.22498931655736576,
+      "learning_rate": 2.8297181491690756e-05,
+      "loss": 0.2191,
+      "step": 1900
+    },
+    {
+      "epoch": 1.545307443365696,
+      "grad_norm": 0.23189215817184283,
+      "learning_rate": 2.8063734209539773e-05,
+      "loss": 0.2192,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 0.22214336417666555,
+      "learning_rate": 2.783001538651554e-05,
+      "loss": 0.2196,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5614886731391586,
+      "grad_norm": 0.22032252634525124,
+      "learning_rate": 2.7596045737278336e-05,
+      "loss": 0.2202,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5695792880258899,
+      "grad_norm": 0.19861461201811947,
+      "learning_rate": 2.7361845998719315e-05,
+      "loss": 0.2224,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5776699029126213,
+      "grad_norm": 0.20017500987570222,
+      "learning_rate": 2.7127436928122612e-05,
+      "loss": 0.2196,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5857605177993528,
+      "grad_norm": 0.2065059828986781,
+      "learning_rate": 2.6892839301325623e-05,
+      "loss": 0.2134,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5938511326860842,
+      "grad_norm": 0.20590200374163903,
+      "learning_rate": 2.6658073910877603e-05,
+      "loss": 0.2167,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.1855510838919491,
+      "learning_rate": 2.6423161564196803e-05,
+      "loss": 0.2161,
+      "step": 1980
+    },
+    {
+      "epoch": 1.610032362459547,
+      "grad_norm": 0.23296838329760938,
+      "learning_rate": 2.6188123081726306e-05,
+      "loss": 0.2174,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6181229773462782,
+      "grad_norm": 0.21045668674547632,
+      "learning_rate": 2.5952979295088714e-05,
+      "loss": 0.2149,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6262135922330097,
+      "grad_norm": 0.19447307206087575,
+      "learning_rate": 2.57177510452398e-05,
+      "loss": 0.2148,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6343042071197411,
+      "grad_norm": 0.23090130774168546,
+      "learning_rate": 2.5482459180621377e-05,
+      "loss": 0.2191,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6423948220064726,
+      "grad_norm": 0.18913378090947097,
+      "learning_rate": 2.524712455531347e-05,
+      "loss": 0.2136,
+      "step": 2030
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.20240602934546922,
+      "learning_rate": 2.501176802718599e-05,
+      "loss": 0.2153,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6585760517799353,
+      "grad_norm": 0.18200383138065016,
+      "learning_rate": 2.4776410456050165e-05,
+      "loss": 0.2149,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.1919414220183083,
+      "learning_rate": 2.4541072701809624e-05,
+      "loss": 0.2187,
+      "step": 2060
+    },
+    {
+      "epoch": 1.674757281553398,
+      "grad_norm": 0.194424204771118,
+      "learning_rate": 2.4305775622611627e-05,
+      "loss": 0.2156,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6828478964401294,
+      "grad_norm": 0.1986457800050313,
+      "learning_rate": 2.4070540072998372e-05,
+      "loss": 0.2186,
+      "step": 2080
+    },
+    {
+      "epoch": 1.690938511326861,
+      "grad_norm": 0.20702148931842815,
+      "learning_rate": 2.3835386902058637e-05,
+      "loss": 0.2145,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.1933456613050236,
+      "learning_rate": 2.360033695157995e-05,
+      "loss": 0.213,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7071197411003236,
+      "grad_norm": 0.1902051798469252,
+      "learning_rate": 2.3365411054201315e-05,
+      "loss": 0.2154,
+      "step": 2110
+    },
+    {
+      "epoch": 1.715210355987055,
+      "grad_norm": 0.2128582736922773,
+      "learning_rate": 2.3130630031566818e-05,
+      "loss": 0.2209,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7233009708737863,
+      "grad_norm": 0.16887432451943601,
+      "learning_rate": 2.2896014692480226e-05,
+      "loss": 0.2156,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7313915857605178,
+      "grad_norm": 0.20654290160014185,
+      "learning_rate": 2.266158583106063e-05,
+      "loss": 0.2156,
+      "step": 2140
+    },
+    {
+      "epoch": 1.7394822006472492,
+      "grad_norm": 0.1889181780602885,
+      "learning_rate": 2.2427364224899502e-05,
+      "loss": 0.2152,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.20357991974988612,
+      "learning_rate": 2.2193370633219115e-05,
+      "loss": 0.2137,
+      "step": 2160
+    },
+    {
+      "epoch": 1.755663430420712,
+      "grad_norm": 0.18958212912498537,
+      "learning_rate": 2.1959625795032664e-05,
+      "loss": 0.2113,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7637540453074434,
+      "grad_norm": 0.18671823093768689,
+      "learning_rate": 2.1726150427306182e-05,
+      "loss": 0.2146,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7718446601941746,
+      "grad_norm": 0.18915530672104663,
+      "learning_rate": 2.1492965223122305e-05,
+      "loss": 0.2132,
+      "step": 2190
+    },
+    {
+      "epoch": 1.779935275080906,
+      "grad_norm": 0.23999090664815534,
+      "learning_rate": 2.126009084984629e-05,
+      "loss": 0.2124,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7880258899676376,
+      "grad_norm": 0.1958651936225041,
+      "learning_rate": 2.102754794729426e-05,
+      "loss": 0.2151,
+      "step": 2210
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.1977750621585522,
+      "learning_rate": 2.079535712590382e-05,
+      "loss": 0.2107,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8042071197411005,
+      "grad_norm": 0.17805905963116433,
+      "learning_rate": 2.056353896490742e-05,
+      "loss": 0.2091,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8122977346278317,
+      "grad_norm": 0.17973603360437174,
+      "learning_rate": 2.0332114010508334e-05,
+      "loss": 0.2114,
+      "step": 2240
+    },
+    {
+      "epoch": 1.820388349514563,
+      "grad_norm": 0.18242429802007595,
+      "learning_rate": 2.010110277405966e-05,
+      "loss": 0.213,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8284789644012944,
+      "grad_norm": 0.18356014531964865,
+      "learning_rate": 1.9870525730246424e-05,
+      "loss": 0.2142,
+      "step": 2260
+    },
+    {
+      "epoch": 1.8365695792880259,
+      "grad_norm": 0.18174264807445242,
+      "learning_rate": 1.9640403315270824e-05,
+      "loss": 0.212,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.19024588215963859,
+      "learning_rate": 1.9410755925041006e-05,
+      "loss": 0.2107,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8527508090614888,
+      "grad_norm": 0.19119554332148833,
+      "learning_rate": 1.918160391336335e-05,
+      "loss": 0.2117,
+      "step": 2290
+    },
+    {
+      "epoch": 1.86084142394822,
+      "grad_norm": 0.19462064918607552,
+      "learning_rate": 1.8952967590138472e-05,
+      "loss": 0.2099,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8689320388349513,
+      "grad_norm": 0.1800377494816668,
+      "learning_rate": 1.8724867219561203e-05,
+      "loss": 0.2154,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8770226537216828,
+      "grad_norm": 0.20601322758987628,
+      "learning_rate": 1.8497323018324476e-05,
+      "loss": 0.2145,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8851132686084142,
+      "grad_norm": 0.1930709584493409,
+      "learning_rate": 1.8270355153827598e-05,
+      "loss": 0.2077,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.1911370351967725,
+      "learning_rate": 1.804398374238872e-05,
+      "loss": 0.2102,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9012944983818771,
+      "grad_norm": 0.18726153065650458,
+      "learning_rate": 1.781822884746196e-05,
+      "loss": 0.2098,
+      "step": 2350
+    },
+    {
+      "epoch": 1.9093851132686084,
+      "grad_norm": 0.1786359702194776,
+      "learning_rate": 1.7593110477859153e-05,
+      "loss": 0.2114,
+      "step": 2360
+    },
+    {
+      "epoch": 1.9174757281553398,
+      "grad_norm": 0.17464418924684114,
+      "learning_rate": 1.736864858597645e-05,
+      "loss": 0.2125,
+      "step": 2370
+    },
+    {
+      "epoch": 1.925566343042071,
+      "grad_norm": 0.19553554568438702,
+      "learning_rate": 1.7144863066025955e-05,
+      "loss": 0.2101,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9336569579288025,
+      "grad_norm": 0.1919410672013163,
+      "learning_rate": 1.692177375227242e-05,
+      "loss": 0.2148,
+      "step": 2390
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.19222701771364425,
+      "learning_rate": 1.669940041727538e-05,
+      "loss": 0.2135,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9498381877022655,
+      "grad_norm": 0.19090816514102554,
+      "learning_rate": 1.6477762770136707e-05,
+      "loss": 0.2107,
+      "step": 2410
+    },
+    {
+      "epoch": 1.9579288025889967,
+      "grad_norm": 0.196303834522943,
+      "learning_rate": 1.625688045475371e-05,
+      "loss": 0.2154,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9660194174757282,
+      "grad_norm": 0.2021430652987766,
+      "learning_rate": 1.603677304807815e-05,
+      "loss": 0.2085,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9741100323624594,
+      "grad_norm": 0.19099797924736667,
+      "learning_rate": 1.5817460058381088e-05,
+      "loss": 0.2089,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9822006472491909,
+      "grad_norm": 0.19913570866606684,
+      "learning_rate": 1.5598960923523842e-05,
+      "loss": 0.2088,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.18039032373193892,
+      "learning_rate": 1.5381295009235262e-05,
+      "loss": 0.2033,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9983818770226538,
+      "grad_norm": 0.17656611003487743,
+      "learning_rate": 1.5164481607395238e-05,
+      "loss": 0.2073,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0064724919093853,
+      "grad_norm": 0.2203078879292808,
+      "learning_rate": 1.4948539934324923e-05,
+      "loss": 0.184,
+      "step": 2480
+    },
+    {
+      "epoch": 2.0145631067961167,
+      "grad_norm": 0.19377235611735016,
+      "learning_rate": 1.4733489129083534e-05,
+      "loss": 0.1729,
+      "step": 2490
+    },
+    {
+      "epoch": 2.0226537216828477,
+      "grad_norm": 0.19353245562742485,
+      "learning_rate": 1.4519348251772058e-05,
+      "loss": 0.1732,
+      "step": 2500
+    },
+    {
+      "epoch": 2.030744336569579,
+      "grad_norm": 0.19647629833009353,
+      "learning_rate": 1.4306136281843962e-05,
+      "loss": 0.1712,
+      "step": 2510
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.16744835404717143,
+      "learning_rate": 1.4093872116422979e-05,
+      "loss": 0.1703,
+      "step": 2520
+    },
+    {
+      "epoch": 2.046925566343042,
+      "grad_norm": 0.17703272841109144,
+      "learning_rate": 1.3882574568628315e-05,
+      "loss": 0.1706,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0550161812297736,
+      "grad_norm": 0.2024362245821369,
+      "learning_rate": 1.3672262365907163e-05,
+      "loss": 0.1695,
+      "step": 2540
+    },
+    {
+      "epoch": 2.063106796116505,
+      "grad_norm": 0.1767034203926602,
+      "learning_rate": 1.3462954148374899e-05,
+      "loss": 0.1701,
+      "step": 2550
+    },
+    {
+      "epoch": 2.071197411003236,
+      "grad_norm": 0.18197977986545294,
+      "learning_rate": 1.3254668467163029e-05,
+      "loss": 0.1722,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0792880258899675,
+      "grad_norm": 0.17760389056330828,
+      "learning_rate": 1.3047423782774937e-05,
+      "loss": 0.1696,
+      "step": 2570
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.19203143690929964,
+      "learning_rate": 1.2841238463449743e-05,
+      "loss": 0.1712,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0954692556634305,
+      "grad_norm": 0.17183295745816207,
+      "learning_rate": 1.2636130783534319e-05,
+      "loss": 0.1711,
+      "step": 2590
+    },
+    {
+      "epoch": 2.103559870550162,
+      "grad_norm": 0.171911071755841,
+      "learning_rate": 1.2432118921863604e-05,
+      "loss": 0.1661,
+      "step": 2600
+    },
+    {
+      "epoch": 2.1116504854368934,
+      "grad_norm": 0.16535945547449518,
+      "learning_rate": 1.2229220960149431e-05,
+      "loss": 0.1721,
+      "step": 2610
+    },
+    {
+      "epoch": 2.1197411003236244,
+      "grad_norm": 0.15768148348111047,
+      "learning_rate": 1.2027454881377889e-05,
+      "loss": 0.1681,
+      "step": 2620
+    },
+    {
+      "epoch": 2.127831715210356,
+      "grad_norm": 0.16958702285946278,
+      "learning_rate": 1.1826838568215526e-05,
+      "loss": 0.1693,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.16417738734331427,
+      "learning_rate": 1.1627389801424351e-05,
+      "loss": 0.1674,
+      "step": 2640
+    },
+    {
+      "epoch": 2.144012944983819,
+      "grad_norm": 0.1875141618399819,
+      "learning_rate": 1.1429126258285946e-05,
+      "loss": 0.1743,
+      "step": 2650
+    },
+    {
+      "epoch": 2.1521035598705502,
+      "grad_norm": 0.1840991587252681,
+      "learning_rate": 1.1232065511034696e-05,
+      "loss": 0.1749,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1601941747572817,
+      "grad_norm": 0.18075731721273597,
+      "learning_rate": 1.1036225025300357e-05,
+      "loss": 0.1758,
+      "step": 2670
+    },
+    {
+      "epoch": 2.168284789644013,
+      "grad_norm": 0.1751267297742484,
+      "learning_rate": 1.0841622158560085e-05,
+      "loss": 0.1762,
+      "step": 2680
+    },
+    {
+      "epoch": 2.176375404530744,
+      "grad_norm": 0.17441997867016168,
+      "learning_rate": 1.0648274158599994e-05,
+      "loss": 0.1747,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.16061544075925469,
+      "learning_rate": 1.0456198161986489e-05,
+      "loss": 0.171,
+      "step": 2700
+    },
+    {
+      "epoch": 2.192556634304207,
+      "grad_norm": 0.17229966007961503,
+      "learning_rate": 1.0265411192547462e-05,
+      "loss": 0.1712,
+      "step": 2710
+    },
+    {
+      "epoch": 2.2006472491909386,
+      "grad_norm": 0.17576497117173268,
+      "learning_rate": 1.0075930159863416e-05,
+      "loss": 0.1713,
+      "step": 2720
+    },
+    {
+      "epoch": 2.20873786407767,
+      "grad_norm": 0.17654170464263277,
+      "learning_rate": 9.887771857768796e-06,
+      "loss": 0.1716,
+      "step": 2730
+    },
+    {
+      "epoch": 2.2168284789644015,
+      "grad_norm": 0.15949424974193477,
+      "learning_rate": 9.700952962863513e-06,
+      "loss": 0.1724,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2249190938511325,
+      "grad_norm": 0.1797493219887846,
+      "learning_rate": 9.515490033034893e-06,
+      "loss": 0.1747,
+      "step": 2750
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.1640136227884059,
+      "learning_rate": 9.331399505990168e-06,
+      "loss": 0.17,
+      "step": 2760
+    },
+    {
+      "epoch": 2.2411003236245954,
+      "grad_norm": 0.1761964552588544,
+      "learning_rate": 9.148697697799533e-06,
+      "loss": 0.1744,
+      "step": 2770
+    },
+    {
+      "epoch": 2.249190938511327,
+      "grad_norm": 0.16581441786521817,
+      "learning_rate": 8.967400801450105e-06,
+      "loss": 0.1683,
+      "step": 2780
+    },
+    {
+      "epoch": 2.2572815533980584,
+      "grad_norm": 0.15986397423733795,
+      "learning_rate": 8.787524885410678e-06,
+      "loss": 0.1684,
+      "step": 2790
+    },
+    {
+      "epoch": 2.26537216828479,
+      "grad_norm": 0.16854533205632355,
+      "learning_rate": 8.60908589220758e-06,
+      "loss": 0.1695,
+      "step": 2800
+    },
+    {
+      "epoch": 2.273462783171521,
+      "grad_norm": 0.16192651639080963,
+      "learning_rate": 8.432099637011693e-06,
+      "loss": 0.1686,
+      "step": 2810
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.15876123318459937,
+      "learning_rate": 8.256581806236704e-06,
+      "loss": 0.1727,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2896440129449838,
+      "grad_norm": 0.15953467663865786,
+      "learning_rate": 8.082547956148873e-06,
+      "loss": 0.1702,
+      "step": 2830
+    },
+    {
+      "epoch": 2.2977346278317152,
+      "grad_norm": 0.15986140875482005,
+      "learning_rate": 7.91001351148819e-06,
+      "loss": 0.1688,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3058252427184467,
+      "grad_norm": 0.16937360962913361,
+      "learning_rate": 7.738993764101324e-06,
+      "loss": 0.1687,
+      "step": 2850
+    },
+    {
+      "epoch": 2.313915857605178,
+      "grad_norm": 0.1624566036577771,
+      "learning_rate": 7.569503871586292e-06,
+      "loss": 0.1709,
+      "step": 2860
+    },
+    {
+      "epoch": 2.3220064724919096,
+      "grad_norm": 0.17174966186956542,
+      "learning_rate": 7.401558855949004e-06,
+      "loss": 0.168,
+      "step": 2870
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.1619257670204373,
+      "learning_rate": 7.235173602271875e-06,
+      "loss": 0.1697,
+      "step": 2880
+    },
+    {
+      "epoch": 2.338187702265372,
+      "grad_norm": 0.1659355702344565,
+      "learning_rate": 7.070362857394538e-06,
+      "loss": 0.1712,
+      "step": 2890
+    },
+    {
+      "epoch": 2.3462783171521036,
+      "grad_norm": 0.16020534539972092,
+      "learning_rate": 6.907141228606831e-06,
+      "loss": 0.165,
+      "step": 2900
+    },
+    {
+      "epoch": 2.354368932038835,
+      "grad_norm": 0.16872762980492875,
+      "learning_rate": 6.745523182354147e-06,
+      "loss": 0.1717,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3624595469255665,
+      "grad_norm": 0.15648717008884194,
+      "learning_rate": 6.585523042955233e-06,
+      "loss": 0.1668,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3705501618122975,
+      "grad_norm": 0.16263064803747318,
+      "learning_rate": 6.427154991332665e-06,
+      "loss": 0.1721,
+      "step": 2930
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.15947499101323748,
+      "learning_rate": 6.2704330637559315e-06,
+      "loss": 0.1709,
+      "step": 2940
+    },
+    {
+      "epoch": 2.3867313915857604,
+      "grad_norm": 0.16083163188925498,
+      "learning_rate": 6.115371150597413e-06,
+      "loss": 0.1702,
+      "step": 2950
+    },
+    {
+      "epoch": 2.394822006472492,
+      "grad_norm": 0.16542590461200707,
+      "learning_rate": 5.961982995101301e-06,
+      "loss": 0.1729,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4029126213592233,
+      "grad_norm": 0.16420989704643807,
+      "learning_rate": 5.810282192165442e-06,
+      "loss": 0.1722,
+      "step": 2970
+    },
+    {
+      "epoch": 2.411003236245955,
+      "grad_norm": 0.1551104913079497,
+      "learning_rate": 5.660282187136507e-06,
+      "loss": 0.1689,
+      "step": 2980
+    },
+    {
+      "epoch": 2.4190938511326863,
+      "grad_norm": 0.15940095271603197,
+      "learning_rate": 5.511996274618253e-06,
+      "loss": 0.1702,
+      "step": 2990
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.15033675616819664,
+      "learning_rate": 5.365437597293238e-06,
+      "loss": 0.1702,
+      "step": 3000
+    },
+    {
+      "epoch": 2.4352750809061487,
+      "grad_norm": 0.15579865123668477,
+      "learning_rate": 5.220619144757996e-06,
+      "loss": 0.1693,
+      "step": 3010
+    },
+    {
+      "epoch": 2.44336569579288,
+      "grad_norm": 0.16013909939251375,
+      "learning_rate": 5.077553752371708e-06,
+      "loss": 0.1678,
+      "step": 3020
+    },
+    {
+      "epoch": 2.4514563106796117,
+      "grad_norm": 0.16673936398632458,
+      "learning_rate": 4.936254100118656e-06,
+      "loss": 0.1709,
+      "step": 3030
+    },
+    {
+      "epoch": 2.459546925566343,
+      "grad_norm": 0.1605515108220787,
+      "learning_rate": 4.796732711484342e-06,
+      "loss": 0.1658,
+      "step": 3040
+    },
+    {
+      "epoch": 2.4676375404530746,
+      "grad_norm": 0.1585644224135113,
+      "learning_rate": 4.659001952345538e-06,
+      "loss": 0.1682,
+      "step": 3050
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.15466709640184034,
+      "learning_rate": 4.523074029874291e-06,
+      "loss": 0.1728,
+      "step": 3060
+    },
+    {
+      "epoch": 2.483818770226537,
+      "grad_norm": 0.15343018591614654,
+      "learning_rate": 4.388960991455998e-06,
+      "loss": 0.1692,
+      "step": 3070
+    },
+    {
+      "epoch": 2.4919093851132685,
+      "grad_norm": 0.14923408281267303,
+      "learning_rate": 4.256674723621621e-06,
+      "loss": 0.1706,
+      "step": 3080
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.16015551238015277,
+      "learning_rate": 4.126226950994211e-06,
+      "loss": 0.1708,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5080906148867315,
+      "grad_norm": 0.15550577660081535,
+      "learning_rate": 3.997629235249692e-06,
+      "loss": 0.1737,
+      "step": 3100
+    },
+    {
+      "epoch": 2.516181229773463,
+      "grad_norm": 0.16405843460813677,
+      "learning_rate": 3.870892974092197e-06,
+      "loss": 0.1677,
+      "step": 3110
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.15592864442189205,
+      "learning_rate": 3.7460294002438444e-06,
+      "loss": 0.1676,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5323624595469254,
+      "grad_norm": 0.15685374774957278,
+      "learning_rate": 3.6230495804491864e-06,
+      "loss": 0.174,
+      "step": 3130
+    },
+    {
+      "epoch": 2.540453074433657,
+      "grad_norm": 0.15046469900932222,
+      "learning_rate": 3.5019644144943576e-06,
+      "loss": 0.1723,
+      "step": 3140
+    },
+    {
+      "epoch": 2.5485436893203883,
+      "grad_norm": 0.15187030014871522,
+      "learning_rate": 3.382784634241015e-06,
+      "loss": 0.1679,
+      "step": 3150
+    },
+    {
+      "epoch": 2.55663430420712,
+      "grad_norm": 0.15370723807151057,
+      "learning_rate": 3.2655208026751816e-06,
+      "loss": 0.1707,
+      "step": 3160
+    },
+    {
+      "epoch": 2.5647249190938513,
+      "grad_norm": 0.15358348991682996,
+      "learning_rate": 3.150183312971014e-06,
+      "loss": 0.164,
+      "step": 3170
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.15495569068856682,
+      "learning_rate": 3.036782387569659e-06,
+      "loss": 0.1676,
+      "step": 3180
+    },
+    {
+      "epoch": 2.5809061488673137,
+      "grad_norm": 0.15152798447783355,
+      "learning_rate": 2.9253280772732595e-06,
+      "loss": 0.1668,
+      "step": 3190
+    },
+    {
+      "epoch": 2.588996763754045,
+      "grad_norm": 0.14795764100233588,
+      "learning_rate": 2.8158302603540965e-06,
+      "loss": 0.1677,
+      "step": 3200
+    },
+    {
+      "epoch": 2.5970873786407767,
+      "grad_norm": 0.1462435066429316,
+      "learning_rate": 2.708298641679105e-06,
+      "loss": 0.1634,
+      "step": 3210
+    },
+    {
+      "epoch": 2.605177993527508,
+      "grad_norm": 0.15564615414583508,
+      "learning_rate": 2.6027427518497153e-06,
+      "loss": 0.1675,
+      "step": 3220
+    },
+    {
+      "epoch": 2.6132686084142396,
+      "grad_norm": 0.15073879419059655,
+      "learning_rate": 2.49917194635714e-06,
+      "loss": 0.1652,
+      "step": 3230
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.1507824582262081,
+      "learning_rate": 2.397595404753225e-06,
+      "loss": 0.1708,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6294498381877025,
+      "grad_norm": 0.14447199827031953,
+      "learning_rate": 2.2980221298367995e-06,
+      "loss": 0.1718,
+      "step": 3250
+    },
+    {
+      "epoch": 2.6375404530744335,
+      "grad_norm": 0.15310920163624914,
+      "learning_rate": 2.2004609468558175e-06,
+      "loss": 0.1687,
+      "step": 3260
+    },
+    {
+      "epoch": 2.645631067961165,
+      "grad_norm": 0.1552961190816156,
+      "learning_rate": 2.1049205027251216e-06,
+      "loss": 0.168,
+      "step": 3270
+    },
+    {
+      "epoch": 2.6537216828478964,
+      "grad_norm": 0.1541563267360428,
+      "learning_rate": 2.0114092652600806e-06,
+      "loss": 0.1682,
+      "step": 3280
+    },
+    {
+      "epoch": 2.661812297734628,
+      "grad_norm": 0.1531794971521466,
+      "learning_rate": 1.919935522426081e-06,
+      "loss": 0.1688,
+      "step": 3290
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.1502769010065667,
+      "learning_rate": 1.8305073816039492e-06,
+      "loss": 0.17,
+      "step": 3300
+    },
+    {
+      "epoch": 2.6779935275080904,
+      "grad_norm": 0.14759059904949057,
+      "learning_rate": 1.7431327688714139e-06,
+      "loss": 0.1666,
+      "step": 3310
+    },
+    {
+      "epoch": 2.686084142394822,
+      "grad_norm": 0.14884028557553985,
+      "learning_rate": 1.6578194283005804e-06,
+      "loss": 0.1684,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6941747572815533,
+      "grad_norm": 0.15011522863824736,
+      "learning_rate": 1.5745749212715794e-06,
+      "loss": 0.1679,
+      "step": 3330
+    },
+    {
+      "epoch": 2.7022653721682848,
+      "grad_norm": 0.15248852381104064,
+      "learning_rate": 1.4934066258024182e-06,
+      "loss": 0.1741,
+      "step": 3340
+    },
+    {
+      "epoch": 2.7103559870550162,
+      "grad_norm": 0.14444830557078114,
+      "learning_rate": 1.4143217358950217e-06,
+      "loss": 0.168,
+      "step": 3350
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.16097353621052152,
+      "learning_rate": 1.3373272608976668e-06,
+      "loss": 0.1722,
+      "step": 3360
+    },
+    {
+      "epoch": 2.726537216828479,
+      "grad_norm": 0.15190902224222305,
+      "learning_rate": 1.2624300248836928e-06,
+      "loss": 0.1696,
+      "step": 3370
+    },
+    {
+      "epoch": 2.73462783171521,
+      "grad_norm": 0.15777837628961622,
+      "learning_rate": 1.1896366660467173e-06,
+      "loss": 0.168,
+      "step": 3380
+    },
+    {
+      "epoch": 2.7427184466019416,
+      "grad_norm": 0.14537830984259653,
+      "learning_rate": 1.1189536361122799e-06,
+      "loss": 0.1665,
+      "step": 3390
+    },
+    {
+      "epoch": 2.750809061488673,
+      "grad_norm": 0.14842465745819863,
+      "learning_rate": 1.0503871997660036e-06,
+      "loss": 0.1623,
+      "step": 3400
+    },
+    {
+      "epoch": 2.7588996763754046,
+      "grad_norm": 0.16097421255088173,
+      "learning_rate": 9.83943434098372e-07,
+      "loss": 0.166,
+      "step": 3410
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.1486212209344348,
+      "learning_rate": 9.196282280661023e-07,
+      "loss": 0.1688,
+      "step": 3420
+    },
+    {
+      "epoch": 2.775080906148867,
+      "grad_norm": 0.14198772068734017,
+      "learning_rate": 8.574472819702029e-07,
+      "loss": 0.1652,
+      "step": 3430
+    },
+    {
+      "epoch": 2.783171521035599,
+      "grad_norm": 0.1483955979101031,
+      "learning_rate": 7.974061069507571e-07,
+      "loss": 0.1679,
+      "step": 3440
+    },
+    {
+      "epoch": 2.79126213592233,
+      "grad_norm": 0.1510691708695613,
+      "learning_rate": 7.395100244984604e-07,
+      "loss": 0.1676,
+      "step": 3450
+    },
+    {
+      "epoch": 2.7993527508090614,
+      "grad_norm": 0.14675861790646366,
+      "learning_rate": 6.837641659829807e-07,
+      "loss": 0.1662,
+      "step": 3460
+    },
+    {
+      "epoch": 2.807443365695793,
+      "grad_norm": 0.1444417747058221,
+      "learning_rate": 6.301734721981533e-07,
+      "loss": 0.1669,
+      "step": 3470
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.14445462788403754,
+      "learning_rate": 5.787426929240808e-07,
+      "loss": 0.166,
+      "step": 3480
+    },
+    {
+      "epoch": 2.823624595469256,
+      "grad_norm": 0.1614165649662563,
+      "learning_rate": 5.294763865061558e-07,
+      "loss": 0.1683,
+      "step": 3490
+    },
+    {
+      "epoch": 2.831715210355987,
+      "grad_norm": 0.15711301471816222,
+      "learning_rate": 4.823789194510514e-07,
+      "loss": 0.1706,
+      "step": 3500
+    },
+    {
+      "epoch": 2.8398058252427183,
+      "grad_norm": 0.1453400612300062,
+      "learning_rate": 4.3745446603971064e-07,
+      "loss": 0.1688,
+      "step": 3510
+    },
+    {
+      "epoch": 2.8478964401294498,
+      "grad_norm": 0.15333438971393676,
+      "learning_rate": 3.947070079573872e-07,
+      "loss": 0.1638,
+      "step": 3520
+    },
+    {
+      "epoch": 2.855987055016181,
+      "grad_norm": 0.1472420355366777,
+      "learning_rate": 3.541403339407279e-07,
+      "loss": 0.171,
+      "step": 3530
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.1486881057392244,
+      "learning_rate": 3.1575803944199624e-07,
+      "loss": 0.1698,
+      "step": 3540
+    },
+    {
+      "epoch": 2.872168284789644,
+      "grad_norm": 0.140283471672274,
+      "learning_rate": 2.7956352631038906e-07,
+      "loss": 0.1659,
+      "step": 3550
+    },
+    {
+      "epoch": 2.8802588996763756,
+      "grad_norm": 0.14768247651236527,
+      "learning_rate": 2.4556000249054133e-07,
+      "loss": 0.1705,
+      "step": 3560
+    },
+    {
+      "epoch": 2.8883495145631066,
+      "grad_norm": 0.14227761591758845,
+      "learning_rate": 2.1375048173818412e-07,
+      "loss": 0.1676,
+      "step": 3570
+    },
+    {
+      "epoch": 2.896440129449838,
+      "grad_norm": 0.1397094378867443,
+      "learning_rate": 1.8413778335305e-07,
+      "loss": 0.1639,
+      "step": 3580
+    },
+    {
+      "epoch": 2.9045307443365695,
+      "grad_norm": 0.14544808567516948,
+      "learning_rate": 1.567245319290006e-07,
+      "loss": 0.169,
+      "step": 3590
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.14376315892226174,
+      "learning_rate": 1.315131571213879e-07,
+      "loss": 0.163,
+      "step": 3600
+    },
+    {
+      "epoch": 2.9207119741100325,
+      "grad_norm": 0.14749344179261822,
+      "learning_rate": 1.0850589343172624e-07,
+      "loss": 0.1669,
+      "step": 3610
+    },
+    {
+      "epoch": 2.9288025889967635,
+      "grad_norm": 0.1452354644803129,
+      "learning_rate": 8.770478000964532e-08,
+      "loss": 0.1684,
+      "step": 3620
+    },
+    {
+      "epoch": 2.9368932038834954,
+      "grad_norm": 0.14097833660885037,
+      "learning_rate": 6.911166047215145e-08,
+      "loss": 0.1694,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9449838187702264,
+      "grad_norm": 0.14528957043064877,
+      "learning_rate": 5.272818274023872e-08,
+      "loss": 0.1675,
+      "step": 3640
+    },
+    {
+      "epoch": 2.953074433656958,
+      "grad_norm": 0.14549261409943784,
+      "learning_rate": 3.855579889282257e-08,
+      "loss": 0.1679,
+      "step": 3650
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.14557934166615819,
+      "learning_rate": 2.6595765038045507e-08,
+      "loss": 0.1677,
+      "step": 3660
+    },
+    {
+      "epoch": 2.969255663430421,
+      "grad_norm": 0.1445108267077229,
+      "learning_rate": 1.6849141201946693e-08,
+      "loss": 0.165,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9773462783171523,
+      "grad_norm": 0.1425028501060893,
+      "learning_rate": 9.316791234506572e-09,
+      "loss": 0.1713,
+      "step": 3680
+    },
+    {
+      "epoch": 2.9854368932038833,
+      "grad_norm": 0.13458537716318286,
+      "learning_rate": 3.999382733096968e-09,
+      "loss": 0.1665,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9935275080906147,
+      "grad_norm": 0.13982192859442388,
+      "learning_rate": 8.973869832895609e-10,
+      "loss": 0.1686,
+      "step": 3700
+    },
+    {
+      "epoch": 3.0,
+      "step": 3708,
+      "total_flos": 7.26626862104576e+18,
+      "train_loss": 0.24862830043482034,
+      "train_runtime": 257107.0797,
+      "train_samples_per_second": 0.461,
+      "train_steps_per_second": 0.014
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 3708,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.26626862104576e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efde72064ee18fa35cca6fd639189f25db3e52f3bbc800f20c7c6eb49dea86ca
+size 7736