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/Qwen2.5-7B
+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 [/mnt/dhwfile/raise/user/caimengzhang/huggingface/hub/Qwen2.5-7B](https://huggingface.co//mnt/dhwfile/raise/user/caimengzhang/huggingface/hub/Qwen2.5-7B) 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: /mnt/dhwfile/raise/user/caimengzhang/huggingface/hub/Qwen2.5-7B
+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 [/mnt/dhwfile/raise/user/caimengzhang/huggingface/hub/Qwen2.5-7B](https://huggingface.co//mnt/dhwfile/raise/user/caimengzhang/huggingface/hub/Qwen2.5-7B) 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 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|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": 6.364819972765516e+18,
+    "train_loss": 0.2649726264235014,
+    "train_runtime": 208629.7363,
+    "train_samples_per_second": 0.569,
+    "train_steps_per_second": 0.018
+}

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 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|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 +1,59 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 131072,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.55.0",
+  "use_cache": false,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

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 sha256:99cfc18248c17cf4247a7e9c232dcb3008b5475c479d18edd496cfcd824bdd7c
+size 4877660776

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2361d86c7cec32642d1e8f3f71efa1e1438d327c026bd797d31994e66060957e
+size 4932751008

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11f094a82831873bb2359b0d0d6380d8fb267848105e2e0b8215a9198850e80f
+size 4330865200

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

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4917239fa70d1f953df5e58c6cd8f5b4f52a2b2f7f75a4fb9d8e560000a8790f
+size 1089994880

checkpoint-3708/model.safetensors.index.json

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

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 sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

checkpoint-3708/tokenizer_config.json

@@ -0,0 +1,208 @@
+{
+  "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
+    }
+  },
+  "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": 10.671297597284719,
+      "learning_rate": 1.2129380053908356e-06,
+      "loss": 2.601,
+      "step": 10
+    },
+    {
+      "epoch": 0.016181229773462782,
+      "grad_norm": 6.311089864073009,
+      "learning_rate": 2.560646900269542e-06,
+      "loss": 2.3954,
+      "step": 20
+    },
+    {
+      "epoch": 0.024271844660194174,
+      "grad_norm": 5.215362734590839,
+      "learning_rate": 3.908355795148248e-06,
+      "loss": 1.7763,
+      "step": 30
+    },
+    {
+      "epoch": 0.032362459546925564,
+      "grad_norm": 2.0676790587143836,
+      "learning_rate": 5.256064690026954e-06,
+      "loss": 1.2086,
+      "step": 40
+    },
+    {
+      "epoch": 0.040453074433656956,
+      "grad_norm": 1.4206424846718397,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.8436,
+      "step": 50
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 1.5867176039284459,
+      "learning_rate": 7.951482479784367e-06,
+      "loss": 0.6706,
+      "step": 60
+    },
+    {
+      "epoch": 0.05663430420711974,
+      "grad_norm": 1.2138979784248856,
+      "learning_rate": 9.299191374663074e-06,
+      "loss": 0.5858,
+      "step": 70
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 0.7702442127235006,
+      "learning_rate": 1.0646900269541779e-05,
+      "loss": 0.5384,
+      "step": 80
+    },
+    {
+      "epoch": 0.07281553398058252,
+      "grad_norm": 1.0114209152057974,
+      "learning_rate": 1.1994609164420486e-05,
+      "loss": 0.4977,
+      "step": 90
+    },
+    {
+      "epoch": 0.08090614886731391,
+      "grad_norm": 0.5794843568633474,
+      "learning_rate": 1.3342318059299191e-05,
+      "loss": 0.4639,
+      "step": 100
+    },
+    {
+      "epoch": 0.0889967637540453,
+      "grad_norm": 0.696228528125689,
+      "learning_rate": 1.4690026954177898e-05,
+      "loss": 0.4589,
+      "step": 110
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 0.7033818455552552,
+      "learning_rate": 1.6037735849056604e-05,
+      "loss": 0.4511,
+      "step": 120
+    },
+    {
+      "epoch": 0.10517799352750809,
+      "grad_norm": 0.6918266955210028,
+      "learning_rate": 1.738544474393531e-05,
+      "loss": 0.4259,
+      "step": 130
+    },
+    {
+      "epoch": 0.11326860841423948,
+      "grad_norm": 0.48391302376719975,
+      "learning_rate": 1.8733153638814018e-05,
+      "loss": 0.4271,
+      "step": 140
+    },
+    {
+      "epoch": 0.12135922330097088,
+      "grad_norm": 0.6600203232524687,
+      "learning_rate": 2.0080862533692725e-05,
+      "loss": 0.415,
+      "step": 150
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 0.6491506270021533,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.4047,
+      "step": 160
+    },
+    {
+      "epoch": 0.13754045307443366,
+      "grad_norm": 1.108358955768956,
+      "learning_rate": 2.2776280323450135e-05,
+      "loss": 0.3977,
+      "step": 170
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 1.0220828997842137,
+      "learning_rate": 2.4123989218328842e-05,
+      "loss": 0.3972,
+      "step": 180
+    },
+    {
+      "epoch": 0.15372168284789645,
+      "grad_norm": 0.7866428488585541,
+      "learning_rate": 2.547169811320755e-05,
+      "loss": 0.3886,
+      "step": 190
+    },
+    {
+      "epoch": 0.16181229773462782,
+      "grad_norm": 0.7956757749575036,
+      "learning_rate": 2.6819407008086256e-05,
+      "loss": 0.3881,
+      "step": 200
+    },
+    {
+      "epoch": 0.16990291262135923,
+      "grad_norm": 0.7154298194551921,
+      "learning_rate": 2.8167115902964963e-05,
+      "loss": 0.3915,
+      "step": 210
+    },
+    {
+      "epoch": 0.1779935275080906,
+      "grad_norm": 0.7172238259413939,
+      "learning_rate": 2.9514824797843667e-05,
+      "loss": 0.384,
+      "step": 220
+    },
+    {
+      "epoch": 0.18608414239482202,
+      "grad_norm": 0.7995655410192202,
+      "learning_rate": 3.086253369272237e-05,
+      "loss": 0.3748,
+      "step": 230
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 13.800265685156136,
+      "learning_rate": 3.221024258760108e-05,
+      "loss": 0.3732,
+      "step": 240
+    },
+    {
+      "epoch": 0.2022653721682848,
+      "grad_norm": 0.9742105484202757,
+      "learning_rate": 3.355795148247979e-05,
+      "loss": 0.3778,
+      "step": 250
+    },
+    {
+      "epoch": 0.21035598705501618,
+      "grad_norm": 0.7304881990540768,
+      "learning_rate": 3.490566037735849e-05,
+      "loss": 0.3779,
+      "step": 260
+    },
+    {
+      "epoch": 0.21844660194174756,
+      "grad_norm": 0.6020503863674406,
+      "learning_rate": 3.62533692722372e-05,
+      "loss": 0.3659,
+      "step": 270
+    },
+    {
+      "epoch": 0.22653721682847897,
+      "grad_norm": 0.6249460279802084,
+      "learning_rate": 3.76010781671159e-05,
+      "loss": 0.3678,
+      "step": 280
+    },
+    {
+      "epoch": 0.23462783171521034,
+      "grad_norm": 0.7546333918045006,
+      "learning_rate": 3.894878706199461e-05,
+      "loss": 0.3608,
+      "step": 290
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 1.0594379228439452,
+      "learning_rate": 4.0296495956873316e-05,
+      "loss": 0.3623,
+      "step": 300
+    },
+    {
+      "epoch": 0.25080906148867316,
+      "grad_norm": 0.8968179435463678,
+      "learning_rate": 4.164420485175202e-05,
+      "loss": 0.361,
+      "step": 310
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 0.6785334732318974,
+      "learning_rate": 4.299191374663073e-05,
+      "loss": 0.3587,
+      "step": 320
+    },
+    {
+      "epoch": 0.2669902912621359,
+      "grad_norm": 0.7864105607470325,
+      "learning_rate": 4.433962264150944e-05,
+      "loss": 0.3641,
+      "step": 330
+    },
+    {
+      "epoch": 0.2750809061488673,
+      "grad_norm": 0.7520885489880877,
+      "learning_rate": 4.5687331536388144e-05,
+      "loss": 0.3571,
+      "step": 340
+    },
+    {
+      "epoch": 0.28317152103559873,
+      "grad_norm": 0.7154271960483066,
+      "learning_rate": 4.703504043126685e-05,
+      "loss": 0.3564,
+      "step": 350
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.5540540831665184,
+      "learning_rate": 4.838274932614555e-05,
+      "loss": 0.3499,
+      "step": 360
+    },
+    {
+      "epoch": 0.2993527508090615,
+      "grad_norm": 0.8972157863979777,
+      "learning_rate": 4.973045822102426e-05,
+      "loss": 0.3546,
+      "step": 370
+    },
+    {
+      "epoch": 0.3074433656957929,
+      "grad_norm": 0.5977035180668759,
+      "learning_rate": 4.9999290952604396e-05,
+      "loss": 0.358,
+      "step": 380
+    },
+    {
+      "epoch": 0.3155339805825243,
+      "grad_norm": 0.6794119464101197,
+      "learning_rate": 4.9996410516491115e-05,
+      "loss": 0.3442,
+      "step": 390
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 0.6445475083450791,
+      "learning_rate": 4.9991314631296585e-05,
+      "loss": 0.3504,
+      "step": 400
+    },
+    {
+      "epoch": 0.33171521035598706,
+      "grad_norm": 0.6254469656785541,
+      "learning_rate": 4.9984003748672604e-05,
+      "loss": 0.3451,
+      "step": 410
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 0.6124809359745962,
+      "learning_rate": 4.997447851658774e-05,
+      "loss": 0.3373,
+      "step": 420
+    },
+    {
+      "epoch": 0.3478964401294498,
+      "grad_norm": 0.8314626543045377,
+      "learning_rate": 4.9962739779269887e-05,
+      "loss": 0.3427,
+      "step": 430
+    },
+    {
+      "epoch": 0.3559870550161812,
+      "grad_norm": 0.7096404483846562,
+      "learning_rate": 4.9948788577131414e-05,
+      "loss": 0.3402,
+      "step": 440
+    },
+    {
+      "epoch": 0.3640776699029126,
+      "grad_norm": 0.5010872233529703,
+      "learning_rate": 4.993262614667696e-05,
+      "loss": 0.3404,
+      "step": 450
+    },
+    {
+      "epoch": 0.37216828478964403,
+      "grad_norm": 0.7096873040187438,
+      "learning_rate": 4.9914253920393884e-05,
+      "loss": 0.3374,
+      "step": 460
+    },
+    {
+      "epoch": 0.3802588996763754,
+      "grad_norm": 0.6463889915075135,
+      "learning_rate": 4.9893673526625265e-05,
+      "loss": 0.3374,
+      "step": 470
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.698463838389325,
+      "learning_rate": 4.987088678942555e-05,
+      "loss": 0.338,
+      "step": 480
+    },
+    {
+      "epoch": 0.3964401294498382,
+      "grad_norm": 0.6798285617884782,
+      "learning_rate": 4.984589572839897e-05,
+      "loss": 0.3335,
+      "step": 490
+    },
+    {
+      "epoch": 0.4045307443365696,
+      "grad_norm": 0.4599792513564179,
+      "learning_rate": 4.9818702558520485e-05,
+      "loss": 0.3299,
+      "step": 500
+    },
+    {
+      "epoch": 0.41262135922330095,
+      "grad_norm": 0.49728635683969147,
+      "learning_rate": 4.978930968993946e-05,
+      "loss": 0.3388,
+      "step": 510
+    },
+    {
+      "epoch": 0.42071197411003236,
+      "grad_norm": 0.4217113735156639,
+      "learning_rate": 4.9757719727766085e-05,
+      "loss": 0.3319,
+      "step": 520
+    },
+    {
+      "epoch": 0.42880258899676377,
+      "grad_norm": 0.48369122768285383,
+      "learning_rate": 4.972393547184046e-05,
+      "loss": 0.33,
+      "step": 530
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.43182058716411714,
+      "learning_rate": 4.968795991648446e-05,
+      "loss": 0.3252,
+      "step": 540
+    },
+    {
+      "epoch": 0.4449838187702265,
+      "grad_norm": 0.5876386529955178,
+      "learning_rate": 4.9649796250236344e-05,
+      "loss": 0.324,
+      "step": 550
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 0.5486057928498816,
+      "learning_rate": 4.960944785556814e-05,
+      "loss": 0.3254,
+      "step": 560
+    },
+    {
+      "epoch": 0.46116504854368934,
+      "grad_norm": 0.4867958049194759,
+      "learning_rate": 4.956691830858585e-05,
+      "loss": 0.3216,
+      "step": 570
+    },
+    {
+      "epoch": 0.4692556634304207,
+      "grad_norm": 0.4571043410958847,
+      "learning_rate": 4.952221137871252e-05,
+      "loss": 0.3206,
+      "step": 580
+    },
+    {
+      "epoch": 0.4773462783171521,
+      "grad_norm": 0.5063667000215466,
+      "learning_rate": 4.947533102835413e-05,
+      "loss": 0.322,
+      "step": 590
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.4522556669690265,
+      "learning_rate": 4.942628141254843e-05,
+      "loss": 0.3257,
+      "step": 600
+    },
+    {
+      "epoch": 0.4935275080906149,
+      "grad_norm": 0.4321047527271199,
+      "learning_rate": 4.937506687859666e-05,
+      "loss": 0.3197,
+      "step": 610
+    },
+    {
+      "epoch": 0.5016181229773463,
+      "grad_norm": 0.393401694292632,
+      "learning_rate": 4.932169196567824e-05,
+      "loss": 0.3209,
+      "step": 620
+    },
+    {
+      "epoch": 0.5097087378640777,
+      "grad_norm": 0.39900307971080384,
+      "learning_rate": 4.9266161404448454e-05,
+      "loss": 0.3102,
+      "step": 630
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 0.4716849647042864,
+      "learning_rate": 4.920848011661919e-05,
+      "loss": 0.3195,
+      "step": 640
+    },
+    {
+      "epoch": 0.5258899676375405,
+      "grad_norm": 0.460323217460504,
+      "learning_rate": 4.914865321452274e-05,
+      "loss": 0.3151,
+      "step": 650
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 0.45925505988077353,
+      "learning_rate": 4.908668600065862e-05,
+      "loss": 0.3151,
+      "step": 660
+    },
+    {
+      "epoch": 0.5420711974110033,
+      "grad_norm": 0.5830461135145422,
+      "learning_rate": 4.90225839672237e-05,
+      "loss": 0.3216,
+      "step": 670
+    },
+    {
+      "epoch": 0.5501618122977346,
+      "grad_norm": 0.4701717654524378,
+      "learning_rate": 4.8956352795625325e-05,
+      "loss": 0.3168,
+      "step": 680
+    },
+    {
+      "epoch": 0.558252427184466,
+      "grad_norm": 0.337481033517792,
+      "learning_rate": 4.8887998355977886e-05,
+      "loss": 0.313,
+      "step": 690
+    },
+    {
+      "epoch": 0.5663430420711975,
+      "grad_norm": 0.6269962483834282,
+      "learning_rate": 4.881752670658244e-05,
+      "loss": 0.3132,
+      "step": 700
+    },
+    {
+      "epoch": 0.5744336569579288,
+      "grad_norm": 0.4662171259348263,
+      "learning_rate": 4.87449440933898e-05,
+      "loss": 0.314,
+      "step": 710
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.5290025725406212,
+      "learning_rate": 4.867025694944698e-05,
+      "loss": 0.3109,
+      "step": 720
+    },
+    {
+      "epoch": 0.5906148867313916,
+      "grad_norm": 0.43779075196037404,
+      "learning_rate": 4.859347189432699e-05,
+      "loss": 0.3188,
+      "step": 730
+    },
+    {
+      "epoch": 0.598705501618123,
+      "grad_norm": 0.5368922851512631,
+      "learning_rate": 4.8514595733542144e-05,
+      "loss": 0.3053,
+      "step": 740
+    },
+    {
+      "epoch": 0.6067961165048543,
+      "grad_norm": 0.3739446436447005,
+      "learning_rate": 4.8433635457940915e-05,
+      "loss": 0.3095,
+      "step": 750
+    },
+    {
+      "epoch": 0.6148867313915858,
+      "grad_norm": 0.36805536772795056,
+      "learning_rate": 4.8350598243088283e-05,
+      "loss": 0.3081,
+      "step": 760
+    },
+    {
+      "epoch": 0.6229773462783171,
+      "grad_norm": 0.6081454922011427,
+      "learning_rate": 4.8265491448629804e-05,
+      "loss": 0.3096,
+      "step": 770
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.606989309832587,
+      "learning_rate": 4.817832261763928e-05,
+      "loss": 0.3064,
+      "step": 780
+    },
+    {
+      "epoch": 0.63915857605178,
+      "grad_norm": 0.374478457061797,
+      "learning_rate": 4.8089099475950257e-05,
+      "loss": 0.3075,
+      "step": 790
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 0.4495226360651449,
+      "learning_rate": 4.7997829931471225e-05,
+      "loss": 0.3055,
+      "step": 800
+    },
+    {
+      "epoch": 0.6553398058252428,
+      "grad_norm": 0.48869687342847196,
+      "learning_rate": 4.7904522073484786e-05,
+      "loss": 0.3056,
+      "step": 810
+    },
+    {
+      "epoch": 0.6634304207119741,
+      "grad_norm": 0.4244338878096003,
+      "learning_rate": 4.780918417193065e-05,
+      "loss": 0.3068,
+      "step": 820
+    },
+    {
+      "epoch": 0.6715210355987055,
+      "grad_norm": 0.8568600210635637,
+      "learning_rate": 4.7711824676672726e-05,
+      "loss": 0.311,
+      "step": 830
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.4818809073072544,
+      "learning_rate": 4.76124522167501e-05,
+      "loss": 0.3139,
+      "step": 840
+    },
+    {
+      "epoch": 0.6877022653721683,
+      "grad_norm": 0.5673965426592162,
+      "learning_rate": 4.751107559961238e-05,
+      "loss": 0.3085,
+      "step": 850
+    },
+    {
+      "epoch": 0.6957928802588996,
+      "grad_norm": 0.42296772741332755,
+      "learning_rate": 4.740770381033894e-05,
+      "loss": 0.3129,
+      "step": 860
+    },
+    {
+      "epoch": 0.7038834951456311,
+      "grad_norm": 0.4358296754132214,
+      "learning_rate": 4.730234601084268e-05,
+      "loss": 0.3058,
+      "step": 870
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 0.4373965191621123,
+      "learning_rate": 4.719501153905793e-05,
+      "loss": 0.3025,
+      "step": 880
+    },
+    {
+      "epoch": 0.7200647249190939,
+      "grad_norm": 0.3787352736875979,
+      "learning_rate": 4.7085709908112866e-05,
+      "loss": 0.3034,
+      "step": 890
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.469843837143792,
+      "learning_rate": 4.6974450805486305e-05,
+      "loss": 0.303,
+      "step": 900
+    },
+    {
+      "epoch": 0.7362459546925566,
+      "grad_norm": 0.4394456386334091,
+      "learning_rate": 4.686124409214917e-05,
+      "loss": 0.2975,
+      "step": 910
+    },
+    {
+      "epoch": 0.7443365695792881,
+      "grad_norm": 0.37340718943669743,
+      "learning_rate": 4.674609980169042e-05,
+      "loss": 0.3002,
+      "step": 920
+    },
+    {
+      "epoch": 0.7524271844660194,
+      "grad_norm": 0.47566576341803307,
+      "learning_rate": 4.662902813942784e-05,
+      "loss": 0.3051,
+      "step": 930
+    },
+    {
+      "epoch": 0.7605177993527508,
+      "grad_norm": 0.4410622981522137,
+      "learning_rate": 4.651003948150349e-05,
+      "loss": 0.2962,
+      "step": 940
+    },
+    {
+      "epoch": 0.7686084142394822,
+      "grad_norm": 0.41727057267800893,
+      "learning_rate": 4.638914437396408e-05,
+      "loss": 0.2959,
+      "step": 950
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.3766577031650596,
+      "learning_rate": 4.626635353182626e-05,
+      "loss": 0.2951,
+      "step": 960
+    },
+    {
+      "epoch": 0.7847896440129449,
+      "grad_norm": 0.36286446424405566,
+      "learning_rate": 4.614167783812694e-05,
+      "loss": 0.294,
+      "step": 970
+    },
+    {
+      "epoch": 0.7928802588996764,
+      "grad_norm": 0.4051751512540439,
+      "learning_rate": 4.601512834295874e-05,
+      "loss": 0.2944,
+      "step": 980
+    },
+    {
+      "epoch": 0.8009708737864077,
+      "grad_norm": 0.37027627803204927,
+      "learning_rate": 4.588671626249057e-05,
+      "loss": 0.2952,
+      "step": 990
+    },
+    {
+      "epoch": 0.8090614886731392,
+      "grad_norm": 0.464872048858869,
+      "learning_rate": 4.5756452977973585e-05,
+      "loss": 0.292,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8171521035598706,
+      "grad_norm": 0.4408350560704109,
+      "learning_rate": 4.56243500347324e-05,
+      "loss": 0.3009,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.49398039939609895,
+      "learning_rate": 4.549041914114188e-05,
+      "loss": 0.2969,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.5592891850787267,
+      "learning_rate": 4.535467216758936e-05,
+      "loss": 0.2981,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 0.4230495195351826,
+      "learning_rate": 4.5217121145422616e-05,
+      "loss": 0.2883,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8495145631067961,
+      "grad_norm": 0.33133186454457314,
+      "learning_rate": 4.5077778265883477e-05,
+      "loss": 0.2927,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8576051779935275,
+      "grad_norm": 0.3936509630212508,
+      "learning_rate": 4.4936655879027336e-05,
+      "loss": 0.2948,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8656957928802589,
+      "grad_norm": 0.4500916019115437,
+      "learning_rate": 4.479376649262855e-05,
+      "loss": 0.2943,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.3668748277040676,
+      "learning_rate": 4.464912277107185e-05,
+      "loss": 0.2928,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8818770226537217,
+      "grad_norm": 0.320494366403153,
+      "learning_rate": 4.450273753422992e-05,
+      "loss": 0.2909,
+      "step": 1090
+    },
+    {
+      "epoch": 0.889967637540453,
+      "grad_norm": 0.37987923612254676,
+      "learning_rate": 4.435462375632711e-05,
+      "loss": 0.2964,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8980582524271845,
+      "grad_norm": 0.39981221876873035,
+      "learning_rate": 4.420479456478957e-05,
+      "loss": 0.2994,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 0.3091126654651453,
+      "learning_rate": 4.405326323908178e-05,
+      "loss": 0.2851,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9142394822006472,
+      "grad_norm": 0.47407567538309503,
+      "learning_rate": 4.390004320952947e-05,
+      "loss": 0.2914,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.4473436088105401,
+      "learning_rate": 4.374514805612942e-05,
+      "loss": 0.2958,
+      "step": 1140
+    },
+    {
+      "epoch": 0.93042071197411,
+      "grad_norm": 0.3335572127939302,
+      "learning_rate": 4.358859150734576e-05,
+      "loss": 0.2915,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9385113268608414,
+      "grad_norm": 0.4410700993823537,
+      "learning_rate": 4.343038743889324e-05,
+      "loss": 0.2936,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9466019417475728,
+      "grad_norm": 0.4566532222343424,
+      "learning_rate": 4.3270549872507415e-05,
+      "loss": 0.2921,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9546925566343042,
+      "grad_norm": 0.34831612960636166,
+      "learning_rate": 4.3109092974701895e-05,
+      "loss": 0.2961,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9627831715210357,
+      "grad_norm": 0.33970324826345877,
+      "learning_rate": 4.2946031055512733e-05,
+      "loss": 0.2887,
+      "step": 1190
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.46192580192198573,
+      "learning_rate": 4.2781378567230145e-05,
+      "loss": 0.29,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9789644012944984,
+      "grad_norm": 0.4615119904861507,
+      "learning_rate": 4.2615150103117576e-05,
+      "loss": 0.2886,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9870550161812298,
+      "grad_norm": 0.2981731488057052,
+      "learning_rate": 4.24473603961183e-05,
+      "loss": 0.2908,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9951456310679612,
+      "grad_norm": 0.3632890670918485,
+      "learning_rate": 4.227802431754961e-05,
+      "loss": 0.2924,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0032362459546926,
+      "grad_norm": 0.4150403194043393,
+      "learning_rate": 4.210715687578481e-05,
+      "loss": 0.2727,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0113268608414239,
+      "grad_norm": 0.34869821901507675,
+      "learning_rate": 4.193477321492293e-05,
+      "loss": 0.2453,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.3601010037088075,
+      "learning_rate": 4.176088861344657e-05,
+      "loss": 0.2414,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0275080906148868,
+      "grad_norm": 0.32325115826296424,
+      "learning_rate": 4.158551848286773e-05,
+      "loss": 0.2425,
+      "step": 1270
+    },
+    {
+      "epoch": 1.035598705501618,
+      "grad_norm": 0.38691862676257593,
+      "learning_rate": 4.140867836636189e-05,
+      "loss": 0.247,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0436893203883495,
+      "grad_norm": 0.34882276619380004,
+      "learning_rate": 4.1230383937390374e-05,
+      "loss": 0.239,
+      "step": 1290
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 0.3951880114595209,
+      "learning_rate": 4.1050650998311215e-05,
+      "loss": 0.2375,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0598705501618122,
+      "grad_norm": 0.3736593550545292,
+      "learning_rate": 4.086949547897862e-05,
+      "loss": 0.2479,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.34774896387912796,
+      "learning_rate": 4.068693343533103e-05,
+      "loss": 0.2372,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0760517799352751,
+      "grad_norm": 0.3330973525666551,
+      "learning_rate": 4.050298104796812e-05,
+      "loss": 0.2413,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0841423948220066,
+      "grad_norm": 0.27924340181820345,
+      "learning_rate": 4.0317654620716704e-05,
+      "loss": 0.2454,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0922330097087378,
+      "grad_norm": 0.3807786754448861,
+      "learning_rate": 4.013097057918566e-05,
+      "loss": 0.2457,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1003236245954693,
+      "grad_norm": 0.34402611290470836,
+      "learning_rate": 3.9942945469310175e-05,
+      "loss": 0.2406,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1084142394822007,
+      "grad_norm": 0.3087094208360471,
+      "learning_rate": 3.9753595955885264e-05,
+      "loss": 0.2403,
+      "step": 1370
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.334687554837771,
+      "learning_rate": 3.9562938821088705e-05,
+      "loss": 0.2443,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1245954692556634,
+      "grad_norm": 0.34101610586326964,
+      "learning_rate": 3.9370990962993695e-05,
+      "loss": 0.2455,
+      "step": 1390
+    },
+    {
+      "epoch": 1.132686084142395,
+      "grad_norm": 0.36304196735659044,
+      "learning_rate": 3.9177769394071086e-05,
+      "loss": 0.2423,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1407766990291262,
+      "grad_norm": 0.32455755430896477,
+      "learning_rate": 3.898329123968163e-05,
+      "loss": 0.2424,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1488673139158576,
+      "grad_norm": 0.29393617658060145,
+      "learning_rate": 3.87875737365581e-05,
+      "loss": 0.2383,
+      "step": 1420
+    },
+    {
+      "epoch": 1.156957928802589,
+      "grad_norm": 0.40951987232018516,
+      "learning_rate": 3.8590634231277626e-05,
+      "loss": 0.2605,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.3441977909919604,
+      "learning_rate": 3.8392490178724184e-05,
+      "loss": 0.2455,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1731391585760518,
+      "grad_norm": 0.3112960711261136,
+      "learning_rate": 3.8193159140541645e-05,
+      "loss": 0.2419,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 0.3032973624770194,
+      "learning_rate": 3.7992658783577215e-05,
+      "loss": 0.2422,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1893203883495145,
+      "grad_norm": 0.29109047481785005,
+      "learning_rate": 3.779100687831563e-05,
+      "loss": 0.2462,
+      "step": 1470
+    },
+    {
+      "epoch": 1.197411003236246,
+      "grad_norm": 0.2948022665787199,
+      "learning_rate": 3.758822129730415e-05,
+      "loss": 0.245,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2055016181229774,
+      "grad_norm": 0.31258495126302516,
+      "learning_rate": 3.738432001356851e-05,
+      "loss": 0.2455,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.27328969134267245,
+      "learning_rate": 3.7179321099019916e-05,
+      "loss": 0.2369,
+      "step": 1500
+    },
+    {
+      "epoch": 1.22168284789644,
+      "grad_norm": 0.2847184801501813,
+      "learning_rate": 3.6973242722853365e-05,
+      "loss": 0.2402,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2297734627831716,
+      "grad_norm": 0.29812191017952583,
+      "learning_rate": 3.6766103149937295e-05,
+      "loss": 0.2427,
+      "step": 1520
+    },
+    {
+      "epoch": 1.237864077669903,
+      "grad_norm": 0.3017615376479855,
+      "learning_rate": 3.655792073919471e-05,
+      "loss": 0.2363,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 0.27567364203177047,
+      "learning_rate": 3.634871394197607e-05,
+      "loss": 0.2388,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2540453074433657,
+      "grad_norm": 0.2685513406212581,
+      "learning_rate": 3.6138501300423934e-05,
+      "loss": 0.2378,
+      "step": 1550
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.25546275734682755,
+      "learning_rate": 3.592730144582948e-05,
+      "loss": 0.2341,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2702265372168284,
+      "grad_norm": 0.2841532225060496,
+      "learning_rate": 3.571513309698131e-05,
+      "loss": 0.2366,
+      "step": 1570
+    },
+    {
+      "epoch": 1.27831715210356,
+      "grad_norm": 0.30541995269870376,
+      "learning_rate": 3.5502015058506335e-05,
+      "loss": 0.2375,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2864077669902914,
+      "grad_norm": 0.34844869264740597,
+      "learning_rate": 3.528796621920307e-05,
+      "loss": 0.239,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2944983818770226,
+      "grad_norm": 0.30708349077357805,
+      "learning_rate": 3.50730055503676e-05,
+      "loss": 0.2356,
+      "step": 1600
+    },
+    {
+      "epoch": 1.302588996763754,
+      "grad_norm": 0.3175696506236882,
+      "learning_rate": 3.485715210411204e-05,
+      "loss": 0.2358,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.29197347345983926,
+      "learning_rate": 3.4640425011676034e-05,
+      "loss": 0.2408,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3187702265372168,
+      "grad_norm": 0.28169129894596506,
+      "learning_rate": 3.442284348173106e-05,
+      "loss": 0.2395,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3268608414239482,
+      "grad_norm": 0.3992186155056125,
+      "learning_rate": 3.420442679867796e-05,
+      "loss": 0.2391,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3349514563106797,
+      "grad_norm": 0.35636813012626095,
+      "learning_rate": 3.398519432093782e-05,
+      "loss": 0.2374,
+      "step": 1650
+    },
+    {
+      "epoch": 1.343042071197411,
+      "grad_norm": 0.27742962322683806,
+      "learning_rate": 3.376516547923614e-05,
+      "loss": 0.2336,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3511326860841424,
+      "grad_norm": 0.28660763778902115,
+      "learning_rate": 3.3544359774880714e-05,
+      "loss": 0.2371,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.40153445123463827,
+      "learning_rate": 3.3322796778033204e-05,
+      "loss": 0.2376,
+      "step": 1680
+    },
+    {
+      "epoch": 1.367313915857605,
+      "grad_norm": 0.3035749815074104,
+      "learning_rate": 3.3100496125974624e-05,
+      "loss": 0.2377,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 0.342174636479568,
+      "learning_rate": 3.2877477521364895e-05,
+      "loss": 0.2347,
+      "step": 1700
+    },
+    {
+      "epoch": 1.383495145631068,
+      "grad_norm": 0.3199103842497454,
+      "learning_rate": 3.2653760730496555e-05,
+      "loss": 0.2297,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3915857605177995,
+      "grad_norm": 0.27362978201436644,
+      "learning_rate": 3.242936558154285e-05,
+      "loss": 0.2356,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3996763754045307,
+      "grad_norm": 0.27057113114600895,
+      "learning_rate": 3.2204311962800426e-05,
+      "loss": 0.2322,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.26643871030324334,
+      "learning_rate": 3.197861982092651e-05,
+      "loss": 0.2384,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4158576051779934,
+      "grad_norm": 0.31497718220761595,
+      "learning_rate": 3.175230915917108e-05,
+      "loss": 0.2427,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4239482200647249,
+      "grad_norm": 0.3267383691103009,
+      "learning_rate": 3.152540003560398e-05,
+      "loss": 0.2417,
+      "step": 1760
+    },
+    {
+      "epoch": 1.4320388349514563,
+      "grad_norm": 0.29663197558136717,
+      "learning_rate": 3.129791256133712e-05,
+      "loss": 0.2288,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 0.32445565003875587,
+      "learning_rate": 3.106986689874204e-05,
+      "loss": 0.2347,
+      "step": 1780
+    },
+    {
+      "epoch": 1.448220064724919,
+      "grad_norm": 0.29535887592241783,
+      "learning_rate": 3.0841283259662875e-05,
+      "loss": 0.2385,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.2733621998571008,
+      "learning_rate": 3.0612181903625014e-05,
+      "loss": 0.2359,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4644012944983817,
+      "grad_norm": 0.2698891118245767,
+      "learning_rate": 3.0382583136039444e-05,
+      "loss": 0.2339,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4724919093851132,
+      "grad_norm": 0.30735748022487547,
+      "learning_rate": 3.015250730640308e-05,
+      "loss": 0.237,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4805825242718447,
+      "grad_norm": 0.35024703112347744,
+      "learning_rate": 2.9921974806495178e-05,
+      "loss": 0.2301,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4886731391585761,
+      "grad_norm": 0.27111655738392937,
+      "learning_rate": 2.969100606856998e-05,
+      "loss": 0.2339,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4967637540453074,
+      "grad_norm": 0.2787522117740128,
+      "learning_rate": 2.9459621563545825e-05,
+      "loss": 0.2385,
+      "step": 1850
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.2969133211792456,
+      "learning_rate": 2.9227841799190775e-05,
+      "loss": 0.2305,
+      "step": 1860
+    },
+    {
+      "epoch": 1.51294498381877,
+      "grad_norm": 0.2613286640167969,
+      "learning_rate": 2.8995687318304975e-05,
+      "loss": 0.2328,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5210355987055015,
+      "grad_norm": 0.27216163489187184,
+      "learning_rate": 2.8763178696899995e-05,
+      "loss": 0.2373,
+      "step": 1880
+    },
+    {
+      "epoch": 1.529126213592233,
+      "grad_norm": 0.2964461145886047,
+      "learning_rate": 2.853033654237507e-05,
+      "loss": 0.2289,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5372168284789645,
+      "grad_norm": 0.32949202305010017,
+      "learning_rate": 2.8297181491690756e-05,
+      "loss": 0.23,
+      "step": 1900
+    },
+    {
+      "epoch": 1.545307443365696,
+      "grad_norm": 0.3071880700004742,
+      "learning_rate": 2.8063734209539773e-05,
+      "loss": 0.2367,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 0.2529468819780906,
+      "learning_rate": 2.783001538651554e-05,
+      "loss": 0.2292,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5614886731391586,
+      "grad_norm": 0.2773840151703638,
+      "learning_rate": 2.7596045737278336e-05,
+      "loss": 0.2398,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5695792880258899,
+      "grad_norm": 0.295185746563609,
+      "learning_rate": 2.7361845998719315e-05,
+      "loss": 0.2325,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5776699029126213,
+      "grad_norm": 0.25593073129115584,
+      "learning_rate": 2.7127436928122612e-05,
+      "loss": 0.2367,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5857605177993528,
+      "grad_norm": 0.3516341035154705,
+      "learning_rate": 2.6892839301325623e-05,
+      "loss": 0.239,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5938511326860842,
+      "grad_norm": 0.27153437539635045,
+      "learning_rate": 2.6658073910877603e-05,
+      "loss": 0.2288,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.26490849237503283,
+      "learning_rate": 2.6423161564196803e-05,
+      "loss": 0.231,
+      "step": 1980
+    },
+    {
+      "epoch": 1.610032362459547,
+      "grad_norm": 0.25415769405974487,
+      "learning_rate": 2.6188123081726306e-05,
+      "loss": 0.2341,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6181229773462782,
+      "grad_norm": 0.2566774871023634,
+      "learning_rate": 2.5952979295088714e-05,
+      "loss": 0.2303,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6262135922330097,
+      "grad_norm": 0.28720215998836357,
+      "learning_rate": 2.57177510452398e-05,
+      "loss": 0.2297,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6343042071197411,
+      "grad_norm": 0.26535154825569424,
+      "learning_rate": 2.5482459180621377e-05,
+      "loss": 0.2336,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6423948220064726,
+      "grad_norm": 0.2348030131858575,
+      "learning_rate": 2.524712455531347e-05,
+      "loss": 0.2283,
+      "step": 2030
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.2981340903697484,
+      "learning_rate": 2.501176802718599e-05,
+      "loss": 0.2367,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6585760517799353,
+      "grad_norm": 0.2508609786813777,
+      "learning_rate": 2.4776410456050165e-05,
+      "loss": 0.232,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.2760175375931361,
+      "learning_rate": 2.4541072701809624e-05,
+      "loss": 0.2348,
+      "step": 2060
+    },
+    {
+      "epoch": 1.674757281553398,
+      "grad_norm": 0.2681609470474934,
+      "learning_rate": 2.4305775622611627e-05,
+      "loss": 0.2285,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6828478964401294,
+      "grad_norm": 0.3323878551187027,
+      "learning_rate": 2.4070540072998372e-05,
+      "loss": 0.2272,
+      "step": 2080
+    },
+    {
+      "epoch": 1.690938511326861,
+      "grad_norm": 0.24794381758629244,
+      "learning_rate": 2.3835386902058637e-05,
+      "loss": 0.2292,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.26234035679605566,
+      "learning_rate": 2.360033695157995e-05,
+      "loss": 0.2337,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7071197411003236,
+      "grad_norm": 0.3095409200594481,
+      "learning_rate": 2.3365411054201315e-05,
+      "loss": 0.2265,
+      "step": 2110
+    },
+    {
+      "epoch": 1.715210355987055,
+      "grad_norm": 0.24814239367844307,
+      "learning_rate": 2.3130630031566818e-05,
+      "loss": 0.2269,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7233009708737863,
+      "grad_norm": 0.2503027719931105,
+      "learning_rate": 2.2896014692480226e-05,
+      "loss": 0.231,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7313915857605178,
+      "grad_norm": 0.24658598678553392,
+      "learning_rate": 2.266158583106063e-05,
+      "loss": 0.228,
+      "step": 2140
+    },
+    {
+      "epoch": 1.7394822006472492,
+      "grad_norm": 0.23999491418212948,
+      "learning_rate": 2.2427364224899502e-05,
+      "loss": 0.2289,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.24294562795428556,
+      "learning_rate": 2.2193370633219115e-05,
+      "loss": 0.2295,
+      "step": 2160
+    },
+    {
+      "epoch": 1.755663430420712,
+      "grad_norm": 0.27195306057038027,
+      "learning_rate": 2.1959625795032664e-05,
+      "loss": 0.2312,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7637540453074434,
+      "grad_norm": 0.31612469238407165,
+      "learning_rate": 2.1726150427306182e-05,
+      "loss": 0.2286,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7718446601941746,
+      "grad_norm": 0.29752263873224494,
+      "learning_rate": 2.1492965223122305e-05,
+      "loss": 0.2267,
+      "step": 2190
+    },
+    {
+      "epoch": 1.779935275080906,
+      "grad_norm": 0.2603667246009949,
+      "learning_rate": 2.126009084984629e-05,
+      "loss": 0.2251,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7880258899676376,
+      "grad_norm": 0.26289071039247,
+      "learning_rate": 2.102754794729426e-05,
+      "loss": 0.2246,
+      "step": 2210
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.263903471162293,
+      "learning_rate": 2.079535712590382e-05,
+      "loss": 0.2282,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8042071197411005,
+      "grad_norm": 0.27035700806263924,
+      "learning_rate": 2.056353896490742e-05,
+      "loss": 0.2231,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8122977346278317,
+      "grad_norm": 0.2654192689718002,
+      "learning_rate": 2.0332114010508334e-05,
+      "loss": 0.2268,
+      "step": 2240
+    },
+    {
+      "epoch": 1.820388349514563,
+      "grad_norm": 0.23959916600696418,
+      "learning_rate": 2.010110277405966e-05,
+      "loss": 0.2274,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8284789644012944,
+      "grad_norm": 0.24872872826123787,
+      "learning_rate": 1.9870525730246424e-05,
+      "loss": 0.2293,
+      "step": 2260
+    },
+    {
+      "epoch": 1.8365695792880259,
+      "grad_norm": 0.2645954361424342,
+      "learning_rate": 1.9640403315270824e-05,
+      "loss": 0.2286,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.24420673184598343,
+      "learning_rate": 1.9410755925041006e-05,
+      "loss": 0.2257,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8527508090614888,
+      "grad_norm": 0.24705849748195569,
+      "learning_rate": 1.918160391336335e-05,
+      "loss": 0.2259,
+      "step": 2290
+    },
+    {
+      "epoch": 1.86084142394822,
+      "grad_norm": 0.2951271763878442,
+      "learning_rate": 1.8952967590138472e-05,
+      "loss": 0.2248,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8689320388349513,
+      "grad_norm": 0.2520418480110052,
+      "learning_rate": 1.8724867219561203e-05,
+      "loss": 0.2287,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8770226537216828,
+      "grad_norm": 0.294901863474176,
+      "learning_rate": 1.8497323018324476e-05,
+      "loss": 0.2252,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8851132686084142,
+      "grad_norm": 0.2620675111989084,
+      "learning_rate": 1.8270355153827598e-05,
+      "loss": 0.2243,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.21729342901905718,
+      "learning_rate": 1.804398374238872e-05,
+      "loss": 0.2231,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9012944983818771,
+      "grad_norm": 0.26438351131145754,
+      "learning_rate": 1.781822884746196e-05,
+      "loss": 0.2203,
+      "step": 2350
+    },
+    {
+      "epoch": 1.9093851132686084,
+      "grad_norm": 0.22358504292564055,
+      "learning_rate": 1.7593110477859153e-05,
+      "loss": 0.223,
+      "step": 2360
+    },
+    {
+      "epoch": 1.9174757281553398,
+      "grad_norm": 0.24990353950246044,
+      "learning_rate": 1.736864858597645e-05,
+      "loss": 0.2233,
+      "step": 2370
+    },
+    {
+      "epoch": 1.925566343042071,
+      "grad_norm": 0.24537211238465853,
+      "learning_rate": 1.7144863066025955e-05,
+      "loss": 0.2193,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9336569579288025,
+      "grad_norm": 0.2394304662544921,
+      "learning_rate": 1.692177375227242e-05,
+      "loss": 0.2251,
+      "step": 2390
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.2785110439501238,
+      "learning_rate": 1.669940041727538e-05,
+      "loss": 0.218,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9498381877022655,
+      "grad_norm": 0.23030205395765063,
+      "learning_rate": 1.6477762770136707e-05,
+      "loss": 0.225,
+      "step": 2410
+    },
+    {
+      "epoch": 1.9579288025889967,
+      "grad_norm": 0.24458486846119534,
+      "learning_rate": 1.625688045475371e-05,
+      "loss": 0.2208,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9660194174757282,
+      "grad_norm": 0.223552780447482,
+      "learning_rate": 1.603677304807815e-05,
+      "loss": 0.2223,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9741100323624594,
+      "grad_norm": 0.23262918496301524,
+      "learning_rate": 1.5817460058381088e-05,
+      "loss": 0.2274,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9822006472491909,
+      "grad_norm": 0.2210251232882193,
+      "learning_rate": 1.5598960923523842e-05,
+      "loss": 0.2248,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.2543751678692885,
+      "learning_rate": 1.5381295009235262e-05,
+      "loss": 0.2277,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9983818770226538,
+      "grad_norm": 0.2767394174239906,
+      "learning_rate": 1.5164481607395238e-05,
+      "loss": 0.2243,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0064724919093853,
+      "grad_norm": 0.2831027468460383,
+      "learning_rate": 1.4948539934324923e-05,
+      "loss": 0.1818,
+      "step": 2480
+    },
+    {
+      "epoch": 2.0145631067961167,
+      "grad_norm": 0.23514037047255174,
+      "learning_rate": 1.4733489129083534e-05,
+      "loss": 0.1718,
+      "step": 2490
+    },
+    {
+      "epoch": 2.0226537216828477,
+      "grad_norm": 0.23543211368476746,
+      "learning_rate": 1.4519348251772058e-05,
+      "loss": 0.1696,
+      "step": 2500
+    },
+    {
+      "epoch": 2.030744336569579,
+      "grad_norm": 0.24256191701199467,
+      "learning_rate": 1.4306136281843962e-05,
+      "loss": 0.1691,
+      "step": 2510
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.22234069623742508,
+      "learning_rate": 1.4093872116422979e-05,
+      "loss": 0.169,
+      "step": 2520
+    },
+    {
+      "epoch": 2.046925566343042,
+      "grad_norm": 0.23132835680535235,
+      "learning_rate": 1.3882574568628315e-05,
+      "loss": 0.168,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0550161812297736,
+      "grad_norm": 0.23134456189741553,
+      "learning_rate": 1.3672262365907163e-05,
+      "loss": 0.1684,
+      "step": 2540
+    },
+    {
+      "epoch": 2.063106796116505,
+      "grad_norm": 0.22760921973646828,
+      "learning_rate": 1.3462954148374899e-05,
+      "loss": 0.1661,
+      "step": 2550
+    },
+    {
+      "epoch": 2.071197411003236,
+      "grad_norm": 0.21228629003960797,
+      "learning_rate": 1.3254668467163029e-05,
+      "loss": 0.1696,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0792880258899675,
+      "grad_norm": 0.2182291343034792,
+      "learning_rate": 1.3047423782774937e-05,
+      "loss": 0.1649,
+      "step": 2570
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.21718735716811796,
+      "learning_rate": 1.2841238463449743e-05,
+      "loss": 0.1719,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0954692556634305,
+      "grad_norm": 0.22472401035201942,
+      "learning_rate": 1.2636130783534319e-05,
+      "loss": 0.1703,
+      "step": 2590
+    },
+    {
+      "epoch": 2.103559870550162,
+      "grad_norm": 0.21069125465698005,
+      "learning_rate": 1.2432118921863604e-05,
+      "loss": 0.168,
+      "step": 2600
+    },
+    {
+      "epoch": 2.1116504854368934,
+      "grad_norm": 0.21446866638864276,
+      "learning_rate": 1.2229220960149431e-05,
+      "loss": 0.1695,
+      "step": 2610
+    },
+    {
+      "epoch": 2.1197411003236244,
+      "grad_norm": 0.21738468937149366,
+      "learning_rate": 1.2027454881377889e-05,
+      "loss": 0.1675,
+      "step": 2620
+    },
+    {
+      "epoch": 2.127831715210356,
+      "grad_norm": 0.22154845083846222,
+      "learning_rate": 1.1826838568215526e-05,
+      "loss": 0.1685,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.20713221800601425,
+      "learning_rate": 1.1627389801424351e-05,
+      "loss": 0.1706,
+      "step": 2640
+    },
+    {
+      "epoch": 2.144012944983819,
+      "grad_norm": 0.20802282917927592,
+      "learning_rate": 1.1429126258285946e-05,
+      "loss": 0.1661,
+      "step": 2650
+    },
+    {
+      "epoch": 2.1521035598705502,
+      "grad_norm": 0.20646189593102915,
+      "learning_rate": 1.1232065511034696e-05,
+      "loss": 0.1663,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1601941747572817,
+      "grad_norm": 0.21341477579858295,
+      "learning_rate": 1.1036225025300357e-05,
+      "loss": 0.1687,
+      "step": 2670
+    },
+    {
+      "epoch": 2.168284789644013,
+      "grad_norm": 0.21356709357188863,
+      "learning_rate": 1.0841622158560085e-05,
+      "loss": 0.1668,
+      "step": 2680
+    },
+    {
+      "epoch": 2.176375404530744,
+      "grad_norm": 0.20685337347705235,
+      "learning_rate": 1.0648274158599994e-05,
+      "loss": 0.1698,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.22156407557595678,
+      "learning_rate": 1.0456198161986489e-05,
+      "loss": 0.1753,
+      "step": 2700
+    },
+    {
+      "epoch": 2.192556634304207,
+      "grad_norm": 0.2035462284179136,
+      "learning_rate": 1.0265411192547462e-05,
+      "loss": 0.1718,
+      "step": 2710
+    },
+    {
+      "epoch": 2.2006472491909386,
+      "grad_norm": 0.20189924705262216,
+      "learning_rate": 1.0075930159863416e-05,
+      "loss": 0.1661,
+      "step": 2720
+    },
+    {
+      "epoch": 2.20873786407767,
+      "grad_norm": 0.20956549431298424,
+      "learning_rate": 9.887771857768796e-06,
+      "loss": 0.1683,
+      "step": 2730
+    },
+    {
+      "epoch": 2.2168284789644015,
+      "grad_norm": 0.2140205114087363,
+      "learning_rate": 9.700952962863513e-06,
+      "loss": 0.1701,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2249190938511325,
+      "grad_norm": 0.20398172720239158,
+      "learning_rate": 9.515490033034893e-06,
+      "loss": 0.1667,
+      "step": 2750
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.20165594228560696,
+      "learning_rate": 9.331399505990168e-06,
+      "loss": 0.1655,
+      "step": 2760
+    },
+    {
+      "epoch": 2.2411003236245954,
+      "grad_norm": 0.20182970094436348,
+      "learning_rate": 9.148697697799533e-06,
+      "loss": 0.1694,
+      "step": 2770
+    },
+    {
+      "epoch": 2.249190938511327,
+      "grad_norm": 0.2291167648037075,
+      "learning_rate": 8.967400801450105e-06,
+      "loss": 0.1685,
+      "step": 2780
+    },
+    {
+      "epoch": 2.2572815533980584,
+      "grad_norm": 0.20375204743694134,
+      "learning_rate": 8.787524885410678e-06,
+      "loss": 0.1638,
+      "step": 2790
+    },
+    {
+      "epoch": 2.26537216828479,
+      "grad_norm": 0.214007858233087,
+      "learning_rate": 8.60908589220758e-06,
+      "loss": 0.1676,
+      "step": 2800
+    },
+    {
+      "epoch": 2.273462783171521,
+      "grad_norm": 0.2025950468246571,
+      "learning_rate": 8.432099637011693e-06,
+      "loss": 0.1657,
+      "step": 2810
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.21480260045098348,
+      "learning_rate": 8.256581806236704e-06,
+      "loss": 0.1649,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2896440129449838,
+      "grad_norm": 0.2052262400110326,
+      "learning_rate": 8.082547956148873e-06,
+      "loss": 0.1663,
+      "step": 2830
+    },
+    {
+      "epoch": 2.2977346278317152,
+      "grad_norm": 0.20019987270861134,
+      "learning_rate": 7.91001351148819e-06,
+      "loss": 0.1653,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3058252427184467,
+      "grad_norm": 0.207551520801799,
+      "learning_rate": 7.738993764101324e-06,
+      "loss": 0.1677,
+      "step": 2850
+    },
+    {
+      "epoch": 2.313915857605178,
+      "grad_norm": 0.19280865204847908,
+      "learning_rate": 7.569503871586292e-06,
+      "loss": 0.1674,
+      "step": 2860
+    },
+    {
+      "epoch": 2.3220064724919096,
+      "grad_norm": 0.20122645680053192,
+      "learning_rate": 7.401558855949004e-06,
+      "loss": 0.1678,
+      "step": 2870
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.1945489837836613,
+      "learning_rate": 7.235173602271875e-06,
+      "loss": 0.162,
+      "step": 2880
+    },
+    {
+      "epoch": 2.338187702265372,
+      "grad_norm": 0.2051511298220981,
+      "learning_rate": 7.070362857394538e-06,
+      "loss": 0.1712,
+      "step": 2890
+    },
+    {
+      "epoch": 2.3462783171521036,
+      "grad_norm": 0.2038615765888219,
+      "learning_rate": 6.907141228606831e-06,
+      "loss": 0.1664,
+      "step": 2900
+    },
+    {
+      "epoch": 2.354368932038835,
+      "grad_norm": 0.19824356024446557,
+      "learning_rate": 6.745523182354147e-06,
+      "loss": 0.1634,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3624595469255665,
+      "grad_norm": 0.1897897800458373,
+      "learning_rate": 6.585523042955233e-06,
+      "loss": 0.1633,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3705501618122975,
+      "grad_norm": 0.19361276633294466,
+      "learning_rate": 6.427154991332665e-06,
+      "loss": 0.1665,
+      "step": 2930
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.19269698264895235,
+      "learning_rate": 6.2704330637559315e-06,
+      "loss": 0.1632,
+      "step": 2940
+    },
+    {
+      "epoch": 2.3867313915857604,
+      "grad_norm": 0.19487091156668288,
+      "learning_rate": 6.115371150597413e-06,
+      "loss": 0.1657,
+      "step": 2950
+    },
+    {
+      "epoch": 2.394822006472492,
+      "grad_norm": 0.17726783434092147,
+      "learning_rate": 5.961982995101301e-06,
+      "loss": 0.1652,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4029126213592233,
+      "grad_norm": 0.19174116985719267,
+      "learning_rate": 5.810282192165442e-06,
+      "loss": 0.1674,
+      "step": 2970
+    },
+    {
+      "epoch": 2.411003236245955,
+      "grad_norm": 0.20196362222659836,
+      "learning_rate": 5.660282187136507e-06,
+      "loss": 0.1648,
+      "step": 2980
+    },
+    {
+      "epoch": 2.4190938511326863,
+      "grad_norm": 0.2037293127188244,
+      "learning_rate": 5.511996274618253e-06,
+      "loss": 0.1673,
+      "step": 2990
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.2075997573199353,
+      "learning_rate": 5.365437597293238e-06,
+      "loss": 0.1681,
+      "step": 3000
+    },
+    {
+      "epoch": 2.4352750809061487,
+      "grad_norm": 0.19485346480204271,
+      "learning_rate": 5.220619144757996e-06,
+      "loss": 0.168,
+      "step": 3010
+    },
+    {
+      "epoch": 2.44336569579288,
+      "grad_norm": 0.20032468931249656,
+      "learning_rate": 5.077553752371708e-06,
+      "loss": 0.1695,
+      "step": 3020
+    },
+    {
+      "epoch": 2.4514563106796117,
+      "grad_norm": 0.18780814063723766,
+      "learning_rate": 4.936254100118656e-06,
+      "loss": 0.1626,
+      "step": 3030
+    },
+    {
+      "epoch": 2.459546925566343,
+      "grad_norm": 0.20597954746449326,
+      "learning_rate": 4.796732711484342e-06,
+      "loss": 0.1688,
+      "step": 3040
+    },
+    {
+      "epoch": 2.4676375404530746,
+      "grad_norm": 0.1884986064298571,
+      "learning_rate": 4.659001952345538e-06,
+      "loss": 0.1618,
+      "step": 3050
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.19330264761303187,
+      "learning_rate": 4.523074029874291e-06,
+      "loss": 0.1611,
+      "step": 3060
+    },
+    {
+      "epoch": 2.483818770226537,
+      "grad_norm": 0.1899995190829889,
+      "learning_rate": 4.388960991455998e-06,
+      "loss": 0.1642,
+      "step": 3070
+    },
+    {
+      "epoch": 2.4919093851132685,
+      "grad_norm": 0.20352479818055758,
+      "learning_rate": 4.256674723621621e-06,
+      "loss": 0.165,
+      "step": 3080
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.19705480301688083,
+      "learning_rate": 4.126226950994211e-06,
+      "loss": 0.1633,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5080906148867315,
+      "grad_norm": 0.19973033430929824,
+      "learning_rate": 3.997629235249692e-06,
+      "loss": 0.1645,
+      "step": 3100
+    },
+    {
+      "epoch": 2.516181229773463,
+      "grad_norm": 0.19227297342228458,
+      "learning_rate": 3.870892974092197e-06,
+      "loss": 0.1657,
+      "step": 3110
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.19226389494086693,
+      "learning_rate": 3.7460294002438444e-06,
+      "loss": 0.1649,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5323624595469254,
+      "grad_norm": 0.23947157434034713,
+      "learning_rate": 3.6230495804491864e-06,
+      "loss": 0.1638,
+      "step": 3130
+    },
+    {
+      "epoch": 2.540453074433657,
+      "grad_norm": 0.18606261818175104,
+      "learning_rate": 3.5019644144943576e-06,
+      "loss": 0.1657,
+      "step": 3140
+    },
+    {
+      "epoch": 2.5485436893203883,
+      "grad_norm": 0.18558275603040017,
+      "learning_rate": 3.382784634241015e-06,
+      "loss": 0.1626,
+      "step": 3150
+    },
+    {
+      "epoch": 2.55663430420712,
+      "grad_norm": 0.1838874793051254,
+      "learning_rate": 3.2655208026751816e-06,
+      "loss": 0.1645,
+      "step": 3160
+    },
+    {
+      "epoch": 2.5647249190938513,
+      "grad_norm": 0.18161462708162493,
+      "learning_rate": 3.150183312971014e-06,
+      "loss": 0.1625,
+      "step": 3170
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.18098413098000662,
+      "learning_rate": 3.036782387569659e-06,
+      "loss": 0.1623,
+      "step": 3180
+    },
+    {
+      "epoch": 2.5809061488673137,
+      "grad_norm": 0.19242866333812558,
+      "learning_rate": 2.9253280772732595e-06,
+      "loss": 0.1627,
+      "step": 3190
+    },
+    {
+      "epoch": 2.588996763754045,
+      "grad_norm": 0.17889937009144094,
+      "learning_rate": 2.8158302603540965e-06,
+      "loss": 0.1609,
+      "step": 3200
+    },
+    {
+      "epoch": 2.5970873786407767,
+      "grad_norm": 0.1879805259242369,
+      "learning_rate": 2.708298641679105e-06,
+      "loss": 0.1642,
+      "step": 3210
+    },
+    {
+      "epoch": 2.605177993527508,
+      "grad_norm": 0.18048882079511047,
+      "learning_rate": 2.6027427518497153e-06,
+      "loss": 0.1634,
+      "step": 3220
+    },
+    {
+      "epoch": 2.6132686084142396,
+      "grad_norm": 0.18903574393083356,
+      "learning_rate": 2.49917194635714e-06,
+      "loss": 0.1607,
+      "step": 3230
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.18631019977805036,
+      "learning_rate": 2.397595404753225e-06,
+      "loss": 0.1589,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6294498381877025,
+      "grad_norm": 0.18583873327326217,
+      "learning_rate": 2.2980221298367995e-06,
+      "loss": 0.1679,
+      "step": 3250
+    },
+    {
+      "epoch": 2.6375404530744335,
+      "grad_norm": 0.1872827952915852,
+      "learning_rate": 2.2004609468558175e-06,
+      "loss": 0.1648,
+      "step": 3260
+    },
+    {
+      "epoch": 2.645631067961165,
+      "grad_norm": 0.18646198991723772,
+      "learning_rate": 2.1049205027251216e-06,
+      "loss": 0.1648,
+      "step": 3270
+    },
+    {
+      "epoch": 2.6537216828478964,
+      "grad_norm": 0.18510117286188082,
+      "learning_rate": 2.0114092652600806e-06,
+      "loss": 0.1669,
+      "step": 3280
+    },
+    {
+      "epoch": 2.661812297734628,
+      "grad_norm": 0.1813478635510933,
+      "learning_rate": 1.919935522426081e-06,
+      "loss": 0.1601,
+      "step": 3290
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.18995739003728715,
+      "learning_rate": 1.8305073816039492e-06,
+      "loss": 0.165,
+      "step": 3300
+    },
+    {
+      "epoch": 2.6779935275080904,
+      "grad_norm": 0.18850255636775962,
+      "learning_rate": 1.7431327688714139e-06,
+      "loss": 0.1686,
+      "step": 3310
+    },
+    {
+      "epoch": 2.686084142394822,
+      "grad_norm": 0.19405450386616901,
+      "learning_rate": 1.6578194283005804e-06,
+      "loss": 0.1586,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6941747572815533,
+      "grad_norm": 0.18168641096398497,
+      "learning_rate": 1.5745749212715794e-06,
+      "loss": 0.1669,
+      "step": 3330
+    },
+    {
+      "epoch": 2.7022653721682848,
+      "grad_norm": 0.18390467428127966,
+      "learning_rate": 1.4934066258024182e-06,
+      "loss": 0.1681,
+      "step": 3340
+    },
+    {
+      "epoch": 2.7103559870550162,
+      "grad_norm": 0.18623903954773796,
+      "learning_rate": 1.4143217358950217e-06,
+      "loss": 0.1648,
+      "step": 3350
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.18708444577479785,
+      "learning_rate": 1.3373272608976668e-06,
+      "loss": 0.1635,
+      "step": 3360
+    },
+    {
+      "epoch": 2.726537216828479,
+      "grad_norm": 0.1850383082145996,
+      "learning_rate": 1.2624300248836928e-06,
+      "loss": 0.1607,
+      "step": 3370
+    },
+    {
+      "epoch": 2.73462783171521,
+      "grad_norm": 0.19358386622393567,
+      "learning_rate": 1.1896366660467173e-06,
+      "loss": 0.1645,
+      "step": 3380
+    },
+    {
+      "epoch": 2.7427184466019416,
+      "grad_norm": 0.17960690916881664,
+      "learning_rate": 1.1189536361122799e-06,
+      "loss": 0.1628,
+      "step": 3390
+    },
+    {
+      "epoch": 2.750809061488673,
+      "grad_norm": 0.20127134763525154,
+      "learning_rate": 1.0503871997660036e-06,
+      "loss": 0.168,
+      "step": 3400
+    },
+    {
+      "epoch": 2.7588996763754046,
+      "grad_norm": 0.19658936043581707,
+      "learning_rate": 9.83943434098372e-07,
+      "loss": 0.1633,
+      "step": 3410
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.18650679653778446,
+      "learning_rate": 9.196282280661023e-07,
+      "loss": 0.1673,
+      "step": 3420
+    },
+    {
+      "epoch": 2.775080906148867,
+      "grad_norm": 0.1733323921140541,
+      "learning_rate": 8.574472819702029e-07,
+      "loss": 0.1652,
+      "step": 3430
+    },
+    {
+      "epoch": 2.783171521035599,
+      "grad_norm": 0.17972872694953243,
+      "learning_rate": 7.974061069507571e-07,
+      "loss": 0.1636,
+      "step": 3440
+    },
+    {
+      "epoch": 2.79126213592233,
+      "grad_norm": 0.1777461753836034,
+      "learning_rate": 7.395100244984604e-07,
+      "loss": 0.1634,
+      "step": 3450
+    },
+    {
+      "epoch": 2.7993527508090614,
+      "grad_norm": 0.18989630729322288,
+      "learning_rate": 6.837641659829807e-07,
+      "loss": 0.1661,
+      "step": 3460
+    },
+    {
+      "epoch": 2.807443365695793,
+      "grad_norm": 0.17625775327505652,
+      "learning_rate": 6.301734721981533e-07,
+      "loss": 0.1643,
+      "step": 3470
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.1862820489099266,
+      "learning_rate": 5.787426929240808e-07,
+      "loss": 0.1643,
+      "step": 3480
+    },
+    {
+      "epoch": 2.823624595469256,
+      "grad_norm": 0.1790622651864573,
+      "learning_rate": 5.294763865061558e-07,
+      "loss": 0.1626,
+      "step": 3490
+    },
+    {
+      "epoch": 2.831715210355987,
+      "grad_norm": 0.173713637483726,
+      "learning_rate": 4.823789194510514e-07,
+      "loss": 0.1639,
+      "step": 3500
+    },
+    {
+      "epoch": 2.8398058252427183,
+      "grad_norm": 0.1866652309901256,
+      "learning_rate": 4.3745446603971064e-07,
+      "loss": 0.1602,
+      "step": 3510
+    },
+    {
+      "epoch": 2.8478964401294498,
+      "grad_norm": 0.17910113776598907,
+      "learning_rate": 3.947070079573872e-07,
+      "loss": 0.1642,
+      "step": 3520
+    },
+    {
+      "epoch": 2.855987055016181,
+      "grad_norm": 0.18065751924055337,
+      "learning_rate": 3.541403339407279e-07,
+      "loss": 0.1656,
+      "step": 3530
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.18007167117331419,
+      "learning_rate": 3.1575803944199624e-07,
+      "loss": 0.1651,
+      "step": 3540
+    },
+    {
+      "epoch": 2.872168284789644,
+      "grad_norm": 0.18247762658055755,
+      "learning_rate": 2.7956352631038906e-07,
+      "loss": 0.1623,
+      "step": 3550
+    },
+    {
+      "epoch": 2.8802588996763756,
+      "grad_norm": 0.17849381652959184,
+      "learning_rate": 2.4556000249054133e-07,
+      "loss": 0.1653,
+      "step": 3560
+    },
+    {
+      "epoch": 2.8883495145631066,
+      "grad_norm": 0.1846373828579316,
+      "learning_rate": 2.1375048173818412e-07,
+      "loss": 0.1691,
+      "step": 3570
+    },
+    {
+      "epoch": 2.896440129449838,
+      "grad_norm": 0.17151731227871805,
+      "learning_rate": 1.8413778335305e-07,
+      "loss": 0.1606,
+      "step": 3580
+    },
+    {
+      "epoch": 2.9045307443365695,
+      "grad_norm": 0.22901848451176904,
+      "learning_rate": 1.567245319290006e-07,
+      "loss": 0.1607,
+      "step": 3590
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.1841472999781577,
+      "learning_rate": 1.315131571213879e-07,
+      "loss": 0.1659,
+      "step": 3600
+    },
+    {
+      "epoch": 2.9207119741100325,
+      "grad_norm": 0.17620735263373793,
+      "learning_rate": 1.0850589343172624e-07,
+      "loss": 0.163,
+      "step": 3610
+    },
+    {
+      "epoch": 2.9288025889967635,
+      "grad_norm": 0.1732366383013906,
+      "learning_rate": 8.770478000964532e-08,
+      "loss": 0.1603,
+      "step": 3620
+    },
+    {
+      "epoch": 2.9368932038834954,
+      "grad_norm": 0.18106989776343552,
+      "learning_rate": 6.911166047215145e-08,
+      "loss": 0.1639,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9449838187702264,
+      "grad_norm": 0.18480994602766684,
+      "learning_rate": 5.272818274023872e-08,
+      "loss": 0.163,
+      "step": 3640
+    },
+    {
+      "epoch": 2.953074433656958,
+      "grad_norm": 0.17298187343963853,
+      "learning_rate": 3.855579889282257e-08,
+      "loss": 0.1607,
+      "step": 3650
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.1826761908488716,
+      "learning_rate": 2.6595765038045507e-08,
+      "loss": 0.1598,
+      "step": 3660
+    },
+    {
+      "epoch": 2.969255663430421,
+      "grad_norm": 0.18104206371371273,
+      "learning_rate": 1.6849141201946693e-08,
+      "loss": 0.1656,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9773462783171523,
+      "grad_norm": 0.17420913687043998,
+      "learning_rate": 9.316791234506572e-09,
+      "loss": 0.1589,
+      "step": 3680
+    },
+    {
+      "epoch": 2.9854368932038833,
+      "grad_norm": 0.177305619544985,
+      "learning_rate": 3.999382733096968e-09,
+      "loss": 0.1633,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9935275080906147,
+      "grad_norm": 0.17610657137080274,
+      "learning_rate": 8.973869832895609e-10,
+      "loss": 0.1623,
+      "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": 6.364819972765516e+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:cea2cc34d068b43f5f15fcf4b3b00ad0f535091c89422c3523e5bd862f073f10
+size 7736

config.json

@@ -0,0 +1,59 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 131072,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.55.0",
+  "use_cache": false,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "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 sha256:99cfc18248c17cf4247a7e9c232dcb3008b5475c479d18edd496cfcd824bdd7c
+size 4877660776

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2361d86c7cec32642d1e8f3f71efa1e1438d327c026bd797d31994e66060957e
+size 4932751008

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11f094a82831873bb2359b0d0d6380d8fb267848105e2e0b8215a9198850e80f
+size 4330865200

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4917239fa70d1f953df5e58c6cd8f5b4f52a2b2f7f75a4fb9d8e560000a8790f
+size 1089994880

model.safetensors.index.json

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

runs/Oct30_20-29-34_SH-IDCA1404-10-140-54-56/events.out.tfevents.1761827539.SH-IDCA1404-10-140-54-56.105170.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bec124ce4e6d4a06b58d9b21e02a5b9e1abec5017ead942e701e4d8f860e6852
+size 84614

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 sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

tokenizer_config.json

@@ -0,0 +1,209 @@
+{
+  "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
+    }
+  },
+  "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": 6.364819972765516e+18,
+    "train_loss": 0.2649726264235014,
+    "train_runtime": 208629.7363,
+    "train_samples_per_second": 0.569,
+    "train_steps_per_second": 0.018
+}

trainer_log.jsonl

@@ -0,0 +1,371 @@
+{"current_steps": 10, "total_steps": 3708, "loss": 2.601, "lr": 1.2129380053908356e-06, "epoch": 0.008090614886731391, "percentage": 0.27, "elapsed_time": "0:09:26", "remaining_time": "2 days, 10:12:11"}
+{"current_steps": 20, "total_steps": 3708, "loss": 2.3954, "lr": 2.560646900269542e-06, "epoch": 0.016181229773462782, "percentage": 0.54, "elapsed_time": "0:18:48", "remaining_time": "2 days, 9:48:56"}
+{"current_steps": 30, "total_steps": 3708, "loss": 1.7763, "lr": 3.908355795148248e-06, "epoch": 0.024271844660194174, "percentage": 0.81, "elapsed_time": "0:28:10", "remaining_time": "2 days, 9:34:50"}
+{"current_steps": 40, "total_steps": 3708, "loss": 1.2086, "lr": 5.256064690026954e-06, "epoch": 0.032362459546925564, "percentage": 1.08, "elapsed_time": "0:37:33", "remaining_time": "2 days, 9:23:39"}
+{"current_steps": 50, "total_steps": 3708, "loss": 0.8436, "lr": 6.60377358490566e-06, "epoch": 0.040453074433656956, "percentage": 1.35, "elapsed_time": "0:46:56", "remaining_time": "2 days, 9:13:48"}
+{"current_steps": 60, "total_steps": 3708, "loss": 0.6706, "lr": 7.951482479784367e-06, "epoch": 0.04854368932038835, "percentage": 1.62, "elapsed_time": "0:56:18", "remaining_time": "2 days, 9:03:33"}
+{"current_steps": 70, "total_steps": 3708, "loss": 0.5858, "lr": 9.299191374663074e-06, "epoch": 0.05663430420711974, "percentage": 1.89, "elapsed_time": "1:05:40", "remaining_time": "2 days, 8:53:32"}
+{"current_steps": 80, "total_steps": 3708, "loss": 0.5384, "lr": 1.0646900269541779e-05, "epoch": 0.06472491909385113, "percentage": 2.16, "elapsed_time": "1:15:02", "remaining_time": "2 days, 8:43:20"}
+{"current_steps": 90, "total_steps": 3708, "loss": 0.4977, "lr": 1.1994609164420486e-05, "epoch": 0.07281553398058252, "percentage": 2.43, "elapsed_time": "1:24:25", "remaining_time": "2 days, 8:33:57"}
+{"current_steps": 100, "total_steps": 3708, "loss": 0.4639, "lr": 1.3342318059299191e-05, "epoch": 0.08090614886731391, "percentage": 2.7, "elapsed_time": "1:33:47", "remaining_time": "2 days, 8:24:15"}
+{"current_steps": 110, "total_steps": 3708, "loss": 0.4589, "lr": 1.4690026954177898e-05, "epoch": 0.0889967637540453, "percentage": 2.97, "elapsed_time": "1:43:10", "remaining_time": "2 days, 8:14:51"}
+{"current_steps": 120, "total_steps": 3708, "loss": 0.4511, "lr": 1.6037735849056604e-05, "epoch": 0.0970873786407767, "percentage": 3.24, "elapsed_time": "1:52:33", "remaining_time": "2 days, 8:05:23"}
+{"current_steps": 130, "total_steps": 3708, "loss": 0.4259, "lr": 1.738544474393531e-05, "epoch": 0.10517799352750809, "percentage": 3.51, "elapsed_time": "2:01:55", "remaining_time": "2 days, 7:55:57"}
+{"current_steps": 140, "total_steps": 3708, "loss": 0.4271, "lr": 1.8733153638814018e-05, "epoch": 0.11326860841423948, "percentage": 3.78, "elapsed_time": "2:11:18", "remaining_time": "2 days, 7:46:33"}
+{"current_steps": 150, "total_steps": 3708, "loss": 0.415, "lr": 2.0080862533692725e-05, "epoch": 0.12135922330097088, "percentage": 4.05, "elapsed_time": "2:20:41", "remaining_time": "2 days, 7:37:10"}
+{"current_steps": 160, "total_steps": 3708, "loss": 0.4047, "lr": 2.1428571428571428e-05, "epoch": 0.12944983818770225, "percentage": 4.31, "elapsed_time": "2:30:03", "remaining_time": "2 days, 7:27:41"}
+{"current_steps": 170, "total_steps": 3708, "loss": 0.3977, "lr": 2.2776280323450135e-05, "epoch": 0.13754045307443366, "percentage": 4.58, "elapsed_time": "2:39:26", "remaining_time": "2 days, 7:18:20"}
+{"current_steps": 180, "total_steps": 3708, "loss": 0.3972, "lr": 2.4123989218328842e-05, "epoch": 0.14563106796116504, "percentage": 4.85, "elapsed_time": "2:48:48", "remaining_time": "2 days, 7:08:45"}
+{"current_steps": 190, "total_steps": 3708, "loss": 0.3886, "lr": 2.547169811320755e-05, "epoch": 0.15372168284789645, "percentage": 5.12, "elapsed_time": "2:58:11", "remaining_time": "2 days, 6:59:15"}
+{"current_steps": 200, "total_steps": 3708, "loss": 0.3881, "lr": 2.6819407008086256e-05, "epoch": 0.16181229773462782, "percentage": 5.39, "elapsed_time": "3:07:32", "remaining_time": "2 days, 6:49:36"}
+{"current_steps": 210, "total_steps": 3708, "loss": 0.3915, "lr": 2.8167115902964963e-05, "epoch": 0.16990291262135923, "percentage": 5.66, "elapsed_time": "3:16:55", "remaining_time": "2 days, 6:40:04"}
+{"current_steps": 220, "total_steps": 3708, "loss": 0.384, "lr": 2.9514824797843667e-05, "epoch": 0.1779935275080906, "percentage": 5.93, "elapsed_time": "3:26:17", "remaining_time": "2 days, 6:30:37"}
+{"current_steps": 230, "total_steps": 3708, "loss": 0.3748, "lr": 3.086253369272237e-05, "epoch": 0.18608414239482202, "percentage": 6.2, "elapsed_time": "3:35:39", "remaining_time": "2 days, 6:21:12"}
+{"current_steps": 240, "total_steps": 3708, "loss": 0.3732, "lr": 3.221024258760108e-05, "epoch": 0.1941747572815534, "percentage": 6.47, "elapsed_time": "3:45:02", "remaining_time": "2 days, 6:11:45"}
+{"current_steps": 250, "total_steps": 3708, "loss": 0.3778, "lr": 3.355795148247979e-05, "epoch": 0.2022653721682848, "percentage": 6.74, "elapsed_time": "3:54:24", "remaining_time": "2 days, 6:02:24"}
+{"current_steps": 260, "total_steps": 3708, "loss": 0.3779, "lr": 3.490566037735849e-05, "epoch": 0.21035598705501618, "percentage": 7.01, "elapsed_time": "4:03:47", "remaining_time": "2 days, 5:53:02"}
+{"current_steps": 270, "total_steps": 3708, "loss": 0.3659, "lr": 3.62533692722372e-05, "epoch": 0.21844660194174756, "percentage": 7.28, "elapsed_time": "4:13:09", "remaining_time": "2 days, 5:43:35"}
+{"current_steps": 280, "total_steps": 3708, "loss": 0.3678, "lr": 3.76010781671159e-05, "epoch": 0.22653721682847897, "percentage": 7.55, "elapsed_time": "4:22:32", "remaining_time": "2 days, 5:34:12"}
+{"current_steps": 290, "total_steps": 3708, "loss": 0.3608, "lr": 3.894878706199461e-05, "epoch": 0.23462783171521034, "percentage": 7.82, "elapsed_time": "4:31:55", "remaining_time": "2 days, 5:24:56"}
+{"current_steps": 300, "total_steps": 3708, "loss": 0.3623, "lr": 4.0296495956873316e-05, "epoch": 0.24271844660194175, "percentage": 8.09, "elapsed_time": "4:41:17", "remaining_time": "2 days, 5:15:28"}
+{"current_steps": 310, "total_steps": 3708, "loss": 0.361, "lr": 4.164420485175202e-05, "epoch": 0.25080906148867316, "percentage": 8.36, "elapsed_time": "4:50:39", "remaining_time": "2 days, 5:06:01"}
+{"current_steps": 320, "total_steps": 3708, "loss": 0.3587, "lr": 4.299191374663073e-05, "epoch": 0.2588996763754045, "percentage": 8.63, "elapsed_time": "5:00:01", "remaining_time": "2 days, 4:56:33"}
+{"current_steps": 330, "total_steps": 3708, "loss": 0.3641, "lr": 4.433962264150944e-05, "epoch": 0.2669902912621359, "percentage": 8.9, "elapsed_time": "5:09:24", "remaining_time": "2 days, 4:47:09"}
+{"current_steps": 340, "total_steps": 3708, "loss": 0.3571, "lr": 4.5687331536388144e-05, "epoch": 0.2750809061488673, "percentage": 9.17, "elapsed_time": "5:18:46", "remaining_time": "2 days, 4:37:46"}
+{"current_steps": 350, "total_steps": 3708, "loss": 0.3564, "lr": 4.703504043126685e-05, "epoch": 0.28317152103559873, "percentage": 9.44, "elapsed_time": "5:28:08", "remaining_time": "2 days, 4:28:19"}
+{"current_steps": 360, "total_steps": 3708, "loss": 0.3499, "lr": 4.838274932614555e-05, "epoch": 0.2912621359223301, "percentage": 9.71, "elapsed_time": "5:37:30", "remaining_time": "2 days, 4:18:50"}
+{"current_steps": 370, "total_steps": 3708, "loss": 0.3546, "lr": 4.973045822102426e-05, "epoch": 0.2993527508090615, "percentage": 9.98, "elapsed_time": "5:46:53", "remaining_time": "2 days, 4:09:29"}
+{"current_steps": 380, "total_steps": 3708, "loss": 0.358, "lr": 4.9999290952604396e-05, "epoch": 0.3074433656957929, "percentage": 10.25, "elapsed_time": "5:56:15", "remaining_time": "2 days, 4:00:06"}
+{"current_steps": 390, "total_steps": 3708, "loss": 0.3442, "lr": 4.9996410516491115e-05, "epoch": 0.3155339805825243, "percentage": 10.52, "elapsed_time": "6:05:38", "remaining_time": "2 days, 3:50:41"}
+{"current_steps": 400, "total_steps": 3708, "loss": 0.3504, "lr": 4.9991314631296585e-05, "epoch": 0.32362459546925565, "percentage": 10.79, "elapsed_time": "6:15:00", "remaining_time": "2 days, 3:41:18"}
+{"current_steps": 410, "total_steps": 3708, "loss": 0.3451, "lr": 4.9984003748672604e-05, "epoch": 0.33171521035598706, "percentage": 11.06, "elapsed_time": "6:24:22", "remaining_time": "2 days, 3:31:52"}
+{"current_steps": 420, "total_steps": 3708, "loss": 0.3373, "lr": 4.997447851658774e-05, "epoch": 0.33980582524271846, "percentage": 11.33, "elapsed_time": "6:33:44", "remaining_time": "2 days, 3:22:28"}
+{"current_steps": 430, "total_steps": 3708, "loss": 0.3427, "lr": 4.9962739779269887e-05, "epoch": 0.3478964401294498, "percentage": 11.6, "elapsed_time": "6:43:07", "remaining_time": "2 days, 3:13:06"}
+{"current_steps": 440, "total_steps": 3708, "loss": 0.3402, "lr": 4.9948788577131414e-05, "epoch": 0.3559870550161812, "percentage": 11.87, "elapsed_time": "6:52:29", "remaining_time": "2 days, 3:03:42"}
+{"current_steps": 450, "total_steps": 3708, "loss": 0.3404, "lr": 4.993262614667696e-05, "epoch": 0.3640776699029126, "percentage": 12.14, "elapsed_time": "7:01:52", "remaining_time": "2 days, 2:54:23"}
+{"current_steps": 460, "total_steps": 3708, "loss": 0.3374, "lr": 4.9914253920393884e-05, "epoch": 0.37216828478964403, "percentage": 12.41, "elapsed_time": "7:11:14", "remaining_time": "2 days, 2:44:55"}
+{"current_steps": 470, "total_steps": 3708, "loss": 0.3374, "lr": 4.9893673526625265e-05, "epoch": 0.3802588996763754, "percentage": 12.68, "elapsed_time": "7:20:36", "remaining_time": "2 days, 2:35:33"}
+{"current_steps": 480, "total_steps": 3708, "loss": 0.338, "lr": 4.987088678942555e-05, "epoch": 0.3883495145631068, "percentage": 12.94, "elapsed_time": "7:29:59", "remaining_time": "2 days, 2:26:09"}
+{"current_steps": 490, "total_steps": 3708, "loss": 0.3335, "lr": 4.984589572839897e-05, "epoch": 0.3964401294498382, "percentage": 13.21, "elapsed_time": "7:39:21", "remaining_time": "2 days, 2:16:43"}
+{"current_steps": 500, "total_steps": 3708, "loss": 0.3299, "lr": 4.9818702558520485e-05, "epoch": 0.4045307443365696, "percentage": 13.48, "elapsed_time": "7:48:43", "remaining_time": "2 days, 2:07:20"}
+{"current_steps": 510, "total_steps": 3708, "loss": 0.3388, "lr": 4.978930968993946e-05, "epoch": 0.41262135922330095, "percentage": 13.75, "elapsed_time": "7:58:05", "remaining_time": "2 days, 1:57:56"}
+{"current_steps": 520, "total_steps": 3708, "loss": 0.3319, "lr": 4.9757719727766085e-05, "epoch": 0.42071197411003236, "percentage": 14.02, "elapsed_time": "8:07:27", "remaining_time": "2 days, 1:48:29"}
+{"current_steps": 530, "total_steps": 3708, "loss": 0.33, "lr": 4.972393547184046e-05, "epoch": 0.42880258899676377, "percentage": 14.29, "elapsed_time": "8:16:49", "remaining_time": "2 days, 1:39:04"}
+{"current_steps": 540, "total_steps": 3708, "loss": 0.3252, "lr": 4.968795991648446e-05, "epoch": 0.4368932038834951, "percentage": 14.56, "elapsed_time": "8:26:11", "remaining_time": "2 days, 1:29:37"}
+{"current_steps": 550, "total_steps": 3708, "loss": 0.324, "lr": 4.9649796250236344e-05, "epoch": 0.4449838187702265, "percentage": 14.83, "elapsed_time": "8:35:33", "remaining_time": "2 days, 1:20:14"}
+{"current_steps": 560, "total_steps": 3708, "loss": 0.3254, "lr": 4.960944785556814e-05, "epoch": 0.45307443365695793, "percentage": 15.1, "elapsed_time": "8:44:55", "remaining_time": "2 days, 1:10:48"}
+{"current_steps": 570, "total_steps": 3708, "loss": 0.3216, "lr": 4.956691830858585e-05, "epoch": 0.46116504854368934, "percentage": 15.37, "elapsed_time": "8:54:17", "remaining_time": "2 days, 1:01:26"}
+{"current_steps": 580, "total_steps": 3708, "loss": 0.3206, "lr": 4.952221137871252e-05, "epoch": 0.4692556634304207, "percentage": 15.64, "elapsed_time": "9:03:40", "remaining_time": "2 days, 0:52:03"}
+{"current_steps": 590, "total_steps": 3708, "loss": 0.322, "lr": 4.947533102835413e-05, "epoch": 0.4773462783171521, "percentage": 15.91, "elapsed_time": "9:13:02", "remaining_time": "2 days, 0:42:40"}
+{"current_steps": 600, "total_steps": 3708, "loss": 0.3257, "lr": 4.942628141254843e-05, "epoch": 0.4854368932038835, "percentage": 16.18, "elapsed_time": "9:22:24", "remaining_time": "2 days, 0:33:16"}
+{"current_steps": 610, "total_steps": 3708, "loss": 0.3197, "lr": 4.937506687859666e-05, "epoch": 0.4935275080906149, "percentage": 16.45, "elapsed_time": "9:31:46", "remaining_time": "2 days, 0:23:52"}
+{"current_steps": 620, "total_steps": 3708, "loss": 0.3209, "lr": 4.932169196567824e-05, "epoch": 0.5016181229773463, "percentage": 16.72, "elapsed_time": "9:41:08", "remaining_time": "2 days, 0:14:29"}
+{"current_steps": 630, "total_steps": 3708, "loss": 0.3102, "lr": 4.9266161404448454e-05, "epoch": 0.5097087378640777, "percentage": 16.99, "elapsed_time": "9:50:31", "remaining_time": "2 days, 0:05:06"}
+{"current_steps": 640, "total_steps": 3708, "loss": 0.3195, "lr": 4.920848011661919e-05, "epoch": 0.517799352750809, "percentage": 17.26, "elapsed_time": "9:59:53", "remaining_time": "1 day, 23:55:41"}
+{"current_steps": 650, "total_steps": 3708, "loss": 0.3151, "lr": 4.914865321452274e-05, "epoch": 0.5258899676375405, "percentage": 17.53, "elapsed_time": "10:09:14", "remaining_time": "1 day, 23:46:16"}
+{"current_steps": 660, "total_steps": 3708, "loss": 0.3151, "lr": 4.908668600065862e-05, "epoch": 0.5339805825242718, "percentage": 17.8, "elapsed_time": "10:18:36", "remaining_time": "1 day, 23:36:51"}
+{"current_steps": 670, "total_steps": 3708, "loss": 0.3216, "lr": 4.90225839672237e-05, "epoch": 0.5420711974110033, "percentage": 18.07, "elapsed_time": "10:27:59", "remaining_time": "1 day, 23:27:29"}
+{"current_steps": 680, "total_steps": 3708, "loss": 0.3168, "lr": 4.8956352795625325e-05, "epoch": 0.5501618122977346, "percentage": 18.34, "elapsed_time": "10:37:21", "remaining_time": "1 day, 23:18:05"}
+{"current_steps": 690, "total_steps": 3708, "loss": 0.313, "lr": 4.8887998355977886e-05, "epoch": 0.558252427184466, "percentage": 18.61, "elapsed_time": "10:46:43", "remaining_time": "1 day, 23:08:43"}
+{"current_steps": 700, "total_steps": 3708, "loss": 0.3132, "lr": 4.881752670658244e-05, "epoch": 0.5663430420711975, "percentage": 18.88, "elapsed_time": "10:56:04", "remaining_time": "1 day, 22:59:17"}
+{"current_steps": 710, "total_steps": 3708, "loss": 0.314, "lr": 4.87449440933898e-05, "epoch": 0.5744336569579288, "percentage": 19.15, "elapsed_time": "11:05:27", "remaining_time": "1 day, 22:49:53"}
+{"current_steps": 720, "total_steps": 3708, "loss": 0.3109, "lr": 4.867025694944698e-05, "epoch": 0.5825242718446602, "percentage": 19.42, "elapsed_time": "11:14:48", "remaining_time": "1 day, 22:40:28"}
+{"current_steps": 730, "total_steps": 3708, "loss": 0.3188, "lr": 4.859347189432699e-05, "epoch": 0.5906148867313916, "percentage": 19.69, "elapsed_time": "11:24:11", "remaining_time": "1 day, 22:31:06"}
+{"current_steps": 740, "total_steps": 3708, "loss": 0.3053, "lr": 4.8514595733542144e-05, "epoch": 0.598705501618123, "percentage": 19.96, "elapsed_time": "11:33:33", "remaining_time": "1 day, 22:21:42"}
+{"current_steps": 750, "total_steps": 3708, "loss": 0.3095, "lr": 4.8433635457940915e-05, "epoch": 0.6067961165048543, "percentage": 20.23, "elapsed_time": "11:42:55", "remaining_time": "1 day, 22:12:18"}
+{"current_steps": 760, "total_steps": 3708, "loss": 0.3081, "lr": 4.8350598243088283e-05, "epoch": 0.6148867313915858, "percentage": 20.5, "elapsed_time": "11:52:16", "remaining_time": "1 day, 22:02:54"}
+{"current_steps": 770, "total_steps": 3708, "loss": 0.3096, "lr": 4.8265491448629804e-05, "epoch": 0.6229773462783171, "percentage": 20.77, "elapsed_time": "12:01:39", "remaining_time": "1 day, 21:53:31"}
+{"current_steps": 780, "total_steps": 3708, "loss": 0.3064, "lr": 4.817832261763928e-05, "epoch": 0.6310679611650486, "percentage": 21.04, "elapsed_time": "12:11:00", "remaining_time": "1 day, 21:44:06"}
+{"current_steps": 790, "total_steps": 3708, "loss": 0.3075, "lr": 4.8089099475950257e-05, "epoch": 0.63915857605178, "percentage": 21.31, "elapsed_time": "12:20:22", "remaining_time": "1 day, 21:34:42"}
+{"current_steps": 800, "total_steps": 3708, "loss": 0.3055, "lr": 4.7997829931471225e-05, "epoch": 0.6472491909385113, "percentage": 21.57, "elapsed_time": "12:29:44", "remaining_time": "1 day, 21:25:18"}
+{"current_steps": 810, "total_steps": 3708, "loss": 0.3056, "lr": 4.7904522073484786e-05, "epoch": 0.6553398058252428, "percentage": 21.84, "elapsed_time": "12:39:06", "remaining_time": "1 day, 21:15:55"}
+{"current_steps": 820, "total_steps": 3708, "loss": 0.3068, "lr": 4.780918417193065e-05, "epoch": 0.6634304207119741, "percentage": 22.11, "elapsed_time": "12:48:28", "remaining_time": "1 day, 21:06:31"}
+{"current_steps": 830, "total_steps": 3708, "loss": 0.311, "lr": 4.7711824676672726e-05, "epoch": 0.6715210355987055, "percentage": 22.38, "elapsed_time": "12:57:50", "remaining_time": "1 day, 20:57:08"}
+{"current_steps": 840, "total_steps": 3708, "loss": 0.3139, "lr": 4.76124522167501e-05, "epoch": 0.6796116504854369, "percentage": 22.65, "elapsed_time": "13:07:12", "remaining_time": "1 day, 20:47:45"}
+{"current_steps": 850, "total_steps": 3708, "loss": 0.3085, "lr": 4.751107559961238e-05, "epoch": 0.6877022653721683, "percentage": 22.92, "elapsed_time": "13:16:34", "remaining_time": "1 day, 20:38:23"}
+{"current_steps": 860, "total_steps": 3708, "loss": 0.3129, "lr": 4.740770381033894e-05, "epoch": 0.6957928802588996, "percentage": 23.19, "elapsed_time": "13:25:56", "remaining_time": "1 day, 20:29:00"}
+{"current_steps": 870, "total_steps": 3708, "loss": 0.3058, "lr": 4.730234601084268e-05, "epoch": 0.7038834951456311, "percentage": 23.46, "elapsed_time": "13:35:18", "remaining_time": "1 day, 20:19:36"}
+{"current_steps": 880, "total_steps": 3708, "loss": 0.3025, "lr": 4.719501153905793e-05, "epoch": 0.7119741100323624, "percentage": 23.73, "elapsed_time": "13:44:40", "remaining_time": "1 day, 20:10:12"}
+{"current_steps": 890, "total_steps": 3708, "loss": 0.3034, "lr": 4.7085709908112866e-05, "epoch": 0.7200647249190939, "percentage": 24.0, "elapsed_time": "13:54:02", "remaining_time": "1 day, 20:00:50"}
+{"current_steps": 900, "total_steps": 3708, "loss": 0.303, "lr": 4.6974450805486305e-05, "epoch": 0.7281553398058253, "percentage": 24.27, "elapsed_time": "14:03:24", "remaining_time": "1 day, 19:51:27"}
+{"current_steps": 910, "total_steps": 3708, "loss": 0.2975, "lr": 4.686124409214917e-05, "epoch": 0.7362459546925566, "percentage": 24.54, "elapsed_time": "14:12:46", "remaining_time": "1 day, 19:42:03"}
+{"current_steps": 920, "total_steps": 3708, "loss": 0.3002, "lr": 4.674609980169042e-05, "epoch": 0.7443365695792881, "percentage": 24.81, "elapsed_time": "14:22:08", "remaining_time": "1 day, 19:32:40"}
+{"current_steps": 930, "total_steps": 3708, "loss": 0.3051, "lr": 4.662902813942784e-05, "epoch": 0.7524271844660194, "percentage": 25.08, "elapsed_time": "14:31:30", "remaining_time": "1 day, 19:23:17"}
+{"current_steps": 940, "total_steps": 3708, "loss": 0.2962, "lr": 4.651003948150349e-05, "epoch": 0.7605177993527508, "percentage": 25.35, "elapsed_time": "14:40:52", "remaining_time": "1 day, 19:13:53"}
+{"current_steps": 950, "total_steps": 3708, "loss": 0.2959, "lr": 4.638914437396408e-05, "epoch": 0.7686084142394822, "percentage": 25.62, "elapsed_time": "14:50:14", "remaining_time": "1 day, 19:04:30"}
+{"current_steps": 960, "total_steps": 3708, "loss": 0.2951, "lr": 4.626635353182626e-05, "epoch": 0.7766990291262136, "percentage": 25.89, "elapsed_time": "14:59:36", "remaining_time": "1 day, 18:55:08"}
+{"current_steps": 970, "total_steps": 3708, "loss": 0.294, "lr": 4.614167783812694e-05, "epoch": 0.7847896440129449, "percentage": 26.16, "elapsed_time": "15:08:59", "remaining_time": "1 day, 18:45:46"}
+{"current_steps": 980, "total_steps": 3708, "loss": 0.2944, "lr": 4.601512834295874e-05, "epoch": 0.7928802588996764, "percentage": 26.43, "elapsed_time": "15:18:21", "remaining_time": "1 day, 18:36:23"}
+{"current_steps": 990, "total_steps": 3708, "loss": 0.2952, "lr": 4.588671626249057e-05, "epoch": 0.8009708737864077, "percentage": 26.7, "elapsed_time": "15:27:42", "remaining_time": "1 day, 18:27:00"}
+{"current_steps": 1000, "total_steps": 3708, "loss": 0.292, "lr": 4.5756452977973585e-05, "epoch": 0.8090614886731392, "percentage": 26.97, "elapsed_time": "15:37:04", "remaining_time": "1 day, 18:17:36"}
+{"current_steps": 1010, "total_steps": 3708, "loss": 0.3009, "lr": 4.56243500347324e-05, "epoch": 0.8171521035598706, "percentage": 27.24, "elapsed_time": "15:46:26", "remaining_time": "1 day, 18:08:13"}
+{"current_steps": 1020, "total_steps": 3708, "loss": 0.2969, "lr": 4.549041914114188e-05, "epoch": 0.8252427184466019, "percentage": 27.51, "elapsed_time": "15:55:48", "remaining_time": "1 day, 17:58:49"}
+{"current_steps": 1030, "total_steps": 3708, "loss": 0.2981, "lr": 4.535467216758936e-05, "epoch": 0.8333333333333334, "percentage": 27.78, "elapsed_time": "16:05:10", "remaining_time": "1 day, 17:49:27"}
+{"current_steps": 1040, "total_steps": 3708, "loss": 0.2883, "lr": 4.5217121145422616e-05, "epoch": 0.8414239482200647, "percentage": 28.05, "elapsed_time": "16:14:32", "remaining_time": "1 day, 17:40:04"}
+{"current_steps": 1050, "total_steps": 3708, "loss": 0.2927, "lr": 4.5077778265883477e-05, "epoch": 0.8495145631067961, "percentage": 28.32, "elapsed_time": "16:23:54", "remaining_time": "1 day, 17:30:41"}
+{"current_steps": 1060, "total_steps": 3708, "loss": 0.2948, "lr": 4.4936655879027336e-05, "epoch": 0.8576051779935275, "percentage": 28.59, "elapsed_time": "16:33:16", "remaining_time": "1 day, 17:21:18"}
+{"current_steps": 1070, "total_steps": 3708, "loss": 0.2943, "lr": 4.479376649262855e-05, "epoch": 0.8656957928802589, "percentage": 28.86, "elapsed_time": "16:42:38", "remaining_time": "1 day, 17:11:55"}
+{"current_steps": 1080, "total_steps": 3708, "loss": 0.2928, "lr": 4.464912277107185e-05, "epoch": 0.8737864077669902, "percentage": 29.13, "elapsed_time": "16:52:00", "remaining_time": "1 day, 17:02:33"}
+{"current_steps": 1090, "total_steps": 3708, "loss": 0.2909, "lr": 4.450273753422992e-05, "epoch": 0.8818770226537217, "percentage": 29.4, "elapsed_time": "17:01:22", "remaining_time": "1 day, 16:53:09"}
+{"current_steps": 1100, "total_steps": 3708, "loss": 0.2964, "lr": 4.435462375632711e-05, "epoch": 0.889967637540453, "percentage": 29.67, "elapsed_time": "17:10:44", "remaining_time": "1 day, 16:43:46"}
+{"current_steps": 1110, "total_steps": 3708, "loss": 0.2994, "lr": 4.420479456478957e-05, "epoch": 0.8980582524271845, "percentage": 29.94, "elapsed_time": "17:20:06", "remaining_time": "1 day, 16:34:24"}
+{"current_steps": 1120, "total_steps": 3708, "loss": 0.2851, "lr": 4.405326323908178e-05, "epoch": 0.9061488673139159, "percentage": 30.2, "elapsed_time": "17:29:28", "remaining_time": "1 day, 16:25:01"}
+{"current_steps": 1130, "total_steps": 3708, "loss": 0.2914, "lr": 4.390004320952947e-05, "epoch": 0.9142394822006472, "percentage": 30.47, "elapsed_time": "17:38:50", "remaining_time": "1 day, 16:15:39"}
+{"current_steps": 1140, "total_steps": 3708, "loss": 0.2958, "lr": 4.374514805612942e-05, "epoch": 0.9223300970873787, "percentage": 30.74, "elapsed_time": "17:48:12", "remaining_time": "1 day, 16:06:16"}
+{"current_steps": 1150, "total_steps": 3708, "loss": 0.2915, "lr": 4.358859150734576e-05, "epoch": 0.93042071197411, "percentage": 31.01, "elapsed_time": "17:57:34", "remaining_time": "1 day, 15:56:53"}
+{"current_steps": 1160, "total_steps": 3708, "loss": 0.2936, "lr": 4.343038743889324e-05, "epoch": 0.9385113268608414, "percentage": 31.28, "elapsed_time": "18:06:56", "remaining_time": "1 day, 15:47:30"}
+{"current_steps": 1170, "total_steps": 3708, "loss": 0.2921, "lr": 4.3270549872507415e-05, "epoch": 0.9466019417475728, "percentage": 31.55, "elapsed_time": "18:16:18", "remaining_time": "1 day, 15:38:08"}
+{"current_steps": 1180, "total_steps": 3708, "loss": 0.2961, "lr": 4.3109092974701895e-05, "epoch": 0.9546925566343042, "percentage": 31.82, "elapsed_time": "18:25:40", "remaining_time": "1 day, 15:28:45"}
+{"current_steps": 1190, "total_steps": 3708, "loss": 0.2887, "lr": 4.2946031055512733e-05, "epoch": 0.9627831715210357, "percentage": 32.09, "elapsed_time": "18:35:02", "remaining_time": "1 day, 15:19:22"}
+{"current_steps": 1200, "total_steps": 3708, "loss": 0.29, "lr": 4.2781378567230145e-05, "epoch": 0.970873786407767, "percentage": 32.36, "elapsed_time": "18:44:23", "remaining_time": "1 day, 15:09:58"}
+{"current_steps": 1210, "total_steps": 3708, "loss": 0.2886, "lr": 4.2615150103117576e-05, "epoch": 0.9789644012944984, "percentage": 32.63, "elapsed_time": "18:53:45", "remaining_time": "1 day, 15:00:36"}
+{"current_steps": 1220, "total_steps": 3708, "loss": 0.2908, "lr": 4.24473603961183e-05, "epoch": 0.9870550161812298, "percentage": 32.9, "elapsed_time": "19:03:07", "remaining_time": "1 day, 14:51:13"}
+{"current_steps": 1230, "total_steps": 3708, "loss": 0.2924, "lr": 4.227802431754961e-05, "epoch": 0.9951456310679612, "percentage": 33.17, "elapsed_time": "19:12:29", "remaining_time": "1 day, 14:41:50"}
+{"current_steps": 1240, "total_steps": 3708, "loss": 0.2727, "lr": 4.210715687578481e-05, "epoch": 1.0032362459546926, "percentage": 33.44, "elapsed_time": "19:22:26", "remaining_time": "1 day, 14:33:38"}
+{"current_steps": 1250, "total_steps": 3708, "loss": 0.2453, "lr": 4.193477321492293e-05, "epoch": 1.0113268608414239, "percentage": 33.71, "elapsed_time": "19:31:49", "remaining_time": "1 day, 14:24:15"}
+{"current_steps": 1260, "total_steps": 3708, "loss": 0.2414, "lr": 4.176088861344657e-05, "epoch": 1.0194174757281553, "percentage": 33.98, "elapsed_time": "19:41:11", "remaining_time": "1 day, 14:14:53"}
+{"current_steps": 1270, "total_steps": 3708, "loss": 0.2425, "lr": 4.158551848286773e-05, "epoch": 1.0275080906148868, "percentage": 34.25, "elapsed_time": "19:50:33", "remaining_time": "1 day, 14:05:30"}
+{"current_steps": 1280, "total_steps": 3708, "loss": 0.247, "lr": 4.140867836636189e-05, "epoch": 1.035598705501618, "percentage": 34.52, "elapsed_time": "19:59:56", "remaining_time": "1 day, 13:56:07"}
+{"current_steps": 1290, "total_steps": 3708, "loss": 0.239, "lr": 4.1230383937390374e-05, "epoch": 1.0436893203883495, "percentage": 34.79, "elapsed_time": "20:09:17", "remaining_time": "1 day, 13:46:43"}
+{"current_steps": 1300, "total_steps": 3708, "loss": 0.2375, "lr": 4.1050650998311215e-05, "epoch": 1.051779935275081, "percentage": 35.06, "elapsed_time": "20:18:40", "remaining_time": "1 day, 13:37:20"}
+{"current_steps": 1310, "total_steps": 3708, "loss": 0.2479, "lr": 4.086949547897862e-05, "epoch": 1.0598705501618122, "percentage": 35.33, "elapsed_time": "20:28:02", "remaining_time": "1 day, 13:27:57"}
+{"current_steps": 1320, "total_steps": 3708, "loss": 0.2372, "lr": 4.068693343533103e-05, "epoch": 1.0679611650485437, "percentage": 35.6, "elapsed_time": "20:37:23", "remaining_time": "1 day, 13:18:33"}
+{"current_steps": 1330, "total_steps": 3708, "loss": 0.2413, "lr": 4.050298104796812e-05, "epoch": 1.0760517799352751, "percentage": 35.87, "elapsed_time": "20:46:45", "remaining_time": "1 day, 13:09:09"}
+{"current_steps": 1340, "total_steps": 3708, "loss": 0.2454, "lr": 4.0317654620716704e-05, "epoch": 1.0841423948220066, "percentage": 36.14, "elapsed_time": "20:56:07", "remaining_time": "1 day, 12:59:46"}
+{"current_steps": 1350, "total_steps": 3708, "loss": 0.2457, "lr": 4.013097057918566e-05, "epoch": 1.0922330097087378, "percentage": 36.41, "elapsed_time": "21:05:29", "remaining_time": "1 day, 12:50:23"}
+{"current_steps": 1360, "total_steps": 3708, "loss": 0.2406, "lr": 3.9942945469310175e-05, "epoch": 1.1003236245954693, "percentage": 36.68, "elapsed_time": "21:14:51", "remaining_time": "1 day, 12:41:00"}
+{"current_steps": 1370, "total_steps": 3708, "loss": 0.2403, "lr": 3.9753595955885264e-05, "epoch": 1.1084142394822007, "percentage": 36.95, "elapsed_time": "21:24:13", "remaining_time": "1 day, 12:31:37"}
+{"current_steps": 1380, "total_steps": 3708, "loss": 0.2443, "lr": 3.9562938821088705e-05, "epoch": 1.116504854368932, "percentage": 37.22, "elapsed_time": "21:33:35", "remaining_time": "1 day, 12:22:13"}
+{"current_steps": 1390, "total_steps": 3708, "loss": 0.2455, "lr": 3.9370990962993695e-05, "epoch": 1.1245954692556634, "percentage": 37.49, "elapsed_time": "21:42:57", "remaining_time": "1 day, 12:12:50"}
+{"current_steps": 1400, "total_steps": 3708, "loss": 0.2423, "lr": 3.9177769394071086e-05, "epoch": 1.132686084142395, "percentage": 37.76, "elapsed_time": "21:52:18", "remaining_time": "1 day, 12:03:26"}
+{"current_steps": 1410, "total_steps": 3708, "loss": 0.2424, "lr": 3.898329123968163e-05, "epoch": 1.1407766990291262, "percentage": 38.03, "elapsed_time": "22:01:41", "remaining_time": "1 day, 11:54:03"}
+{"current_steps": 1420, "total_steps": 3708, "loss": 0.2383, "lr": 3.87875737365581e-05, "epoch": 1.1488673139158576, "percentage": 38.3, "elapsed_time": "22:11:02", "remaining_time": "1 day, 11:44:40"}
+{"current_steps": 1430, "total_steps": 3708, "loss": 0.2605, "lr": 3.8590634231277626e-05, "epoch": 1.156957928802589, "percentage": 38.57, "elapsed_time": "22:20:24", "remaining_time": "1 day, 11:35:17"}
+{"current_steps": 1440, "total_steps": 3708, "loss": 0.2455, "lr": 3.8392490178724184e-05, "epoch": 1.1650485436893203, "percentage": 38.83, "elapsed_time": "22:29:47", "remaining_time": "1 day, 11:25:55"}
+{"current_steps": 1450, "total_steps": 3708, "loss": 0.2419, "lr": 3.8193159140541645e-05, "epoch": 1.1731391585760518, "percentage": 39.1, "elapsed_time": "22:39:09", "remaining_time": "1 day, 11:16:32"}
+{"current_steps": 1460, "total_steps": 3708, "loss": 0.2422, "lr": 3.7992658783577215e-05, "epoch": 1.1812297734627832, "percentage": 39.37, "elapsed_time": "22:48:31", "remaining_time": "1 day, 11:07:09"}
+{"current_steps": 1470, "total_steps": 3708, "loss": 0.2462, "lr": 3.779100687831563e-05, "epoch": 1.1893203883495145, "percentage": 39.64, "elapsed_time": "22:57:53", "remaining_time": "1 day, 10:57:46"}
+{"current_steps": 1480, "total_steps": 3708, "loss": 0.245, "lr": 3.758822129730415e-05, "epoch": 1.197411003236246, "percentage": 39.91, "elapsed_time": "23:07:15", "remaining_time": "1 day, 10:48:23"}
+{"current_steps": 1490, "total_steps": 3708, "loss": 0.2455, "lr": 3.738432001356851e-05, "epoch": 1.2055016181229774, "percentage": 40.18, "elapsed_time": "23:16:37", "remaining_time": "1 day, 10:39:00"}
+{"current_steps": 1500, "total_steps": 3708, "loss": 0.2369, "lr": 3.7179321099019916e-05, "epoch": 1.2135922330097086, "percentage": 40.45, "elapsed_time": "23:25:59", "remaining_time": "1 day, 10:29:37"}
+{"current_steps": 1510, "total_steps": 3708, "loss": 0.2402, "lr": 3.6973242722853365e-05, "epoch": 1.22168284789644, "percentage": 40.72, "elapsed_time": "23:35:21", "remaining_time": "1 day, 10:20:14"}
+{"current_steps": 1520, "total_steps": 3708, "loss": 0.2427, "lr": 3.6766103149937295e-05, "epoch": 1.2297734627831716, "percentage": 40.99, "elapsed_time": "23:44:43", "remaining_time": "1 day, 10:10:51"}
+{"current_steps": 1530, "total_steps": 3708, "loss": 0.2363, "lr": 3.655792073919471e-05, "epoch": 1.237864077669903, "percentage": 41.26, "elapsed_time": "23:54:05", "remaining_time": "1 day, 10:01:28"}
+{"current_steps": 1540, "total_steps": 3708, "loss": 0.2388, "lr": 3.634871394197607e-05, "epoch": 1.2459546925566343, "percentage": 41.53, "elapsed_time": "1 day, 0:03:27", "remaining_time": "1 day, 9:52:05"}
+{"current_steps": 1550, "total_steps": 3708, "loss": 0.2378, "lr": 3.6138501300423934e-05, "epoch": 1.2540453074433657, "percentage": 41.8, "elapsed_time": "1 day, 0:12:50", "remaining_time": "1 day, 9:42:43"}
+{"current_steps": 1560, "total_steps": 3708, "loss": 0.2341, "lr": 3.592730144582948e-05, "epoch": 1.262135922330097, "percentage": 42.07, "elapsed_time": "1 day, 0:22:11", "remaining_time": "1 day, 9:33:20"}
+{"current_steps": 1570, "total_steps": 3708, "loss": 0.2366, "lr": 3.571513309698131e-05, "epoch": 1.2702265372168284, "percentage": 42.34, "elapsed_time": "1 day, 0:31:34", "remaining_time": "1 day, 9:23:57"}
+{"current_steps": 1580, "total_steps": 3708, "loss": 0.2375, "lr": 3.5502015058506335e-05, "epoch": 1.27831715210356, "percentage": 42.61, "elapsed_time": "1 day, 0:40:56", "remaining_time": "1 day, 9:14:34"}
+{"current_steps": 1590, "total_steps": 3708, "loss": 0.239, "lr": 3.528796621920307e-05, "epoch": 1.2864077669902914, "percentage": 42.88, "elapsed_time": "1 day, 0:50:18", "remaining_time": "1 day, 9:05:11"}
+{"current_steps": 1600, "total_steps": 3708, "loss": 0.2356, "lr": 3.50730055503676e-05, "epoch": 1.2944983818770226, "percentage": 43.15, "elapsed_time": "1 day, 0:59:40", "remaining_time": "1 day, 8:55:49"}
+{"current_steps": 1610, "total_steps": 3708, "loss": 0.2358, "lr": 3.485715210411204e-05, "epoch": 1.302588996763754, "percentage": 43.42, "elapsed_time": "1 day, 1:09:02", "remaining_time": "1 day, 8:46:26"}
+{"current_steps": 1620, "total_steps": 3708, "loss": 0.2408, "lr": 3.4640425011676034e-05, "epoch": 1.3106796116504853, "percentage": 43.69, "elapsed_time": "1 day, 1:18:24", "remaining_time": "1 day, 8:37:03"}
+{"current_steps": 1630, "total_steps": 3708, "loss": 0.2395, "lr": 3.442284348173106e-05, "epoch": 1.3187702265372168, "percentage": 43.96, "elapsed_time": "1 day, 1:27:47", "remaining_time": "1 day, 8:27:41"}
+{"current_steps": 1640, "total_steps": 3708, "loss": 0.2391, "lr": 3.420442679867796e-05, "epoch": 1.3268608414239482, "percentage": 44.23, "elapsed_time": "1 day, 1:37:08", "remaining_time": "1 day, 8:18:18"}
+{"current_steps": 1650, "total_steps": 3708, "loss": 0.2374, "lr": 3.398519432093782e-05, "epoch": 1.3349514563106797, "percentage": 44.5, "elapsed_time": "1 day, 1:46:31", "remaining_time": "1 day, 8:08:55"}
+{"current_steps": 1660, "total_steps": 3708, "loss": 0.2336, "lr": 3.376516547923614e-05, "epoch": 1.343042071197411, "percentage": 44.77, "elapsed_time": "1 day, 1:55:52", "remaining_time": "1 day, 7:59:32"}
+{"current_steps": 1670, "total_steps": 3708, "loss": 0.2371, "lr": 3.3544359774880714e-05, "epoch": 1.3511326860841424, "percentage": 45.04, "elapsed_time": "1 day, 2:05:14", "remaining_time": "1 day, 7:50:09"}
+{"current_steps": 1680, "total_steps": 3708, "loss": 0.2376, "lr": 3.3322796778033204e-05, "epoch": 1.3592233009708738, "percentage": 45.31, "elapsed_time": "1 day, 2:14:37", "remaining_time": "1 day, 7:40:47"}
+{"current_steps": 1690, "total_steps": 3708, "loss": 0.2377, "lr": 3.3100496125974624e-05, "epoch": 1.367313915857605, "percentage": 45.58, "elapsed_time": "1 day, 2:23:58", "remaining_time": "1 day, 7:31:24"}
+{"current_steps": 1700, "total_steps": 3708, "loss": 0.2347, "lr": 3.2877477521364895e-05, "epoch": 1.3754045307443366, "percentage": 45.85, "elapsed_time": "1 day, 2:33:21", "remaining_time": "1 day, 7:22:01"}
+{"current_steps": 1710, "total_steps": 3708, "loss": 0.2297, "lr": 3.2653760730496555e-05, "epoch": 1.383495145631068, "percentage": 46.12, "elapsed_time": "1 day, 2:42:43", "remaining_time": "1 day, 7:12:39"}
+{"current_steps": 1720, "total_steps": 3708, "loss": 0.2356, "lr": 3.242936558154285e-05, "epoch": 1.3915857605177995, "percentage": 46.39, "elapsed_time": "1 day, 2:52:05", "remaining_time": "1 day, 7:03:17"}
+{"current_steps": 1730, "total_steps": 3708, "loss": 0.2322, "lr": 3.2204311962800426e-05, "epoch": 1.3996763754045307, "percentage": 46.66, "elapsed_time": "1 day, 3:01:28", "remaining_time": "1 day, 6:53:54"}
+{"current_steps": 1740, "total_steps": 3708, "loss": 0.2384, "lr": 3.197861982092651e-05, "epoch": 1.4077669902912622, "percentage": 46.93, "elapsed_time": "1 day, 3:10:50", "remaining_time": "1 day, 6:44:31"}
+{"current_steps": 1750, "total_steps": 3708, "loss": 0.2427, "lr": 3.175230915917108e-05, "epoch": 1.4158576051779934, "percentage": 47.2, "elapsed_time": "1 day, 3:20:12", "remaining_time": "1 day, 6:35:09"}
+{"current_steps": 1760, "total_steps": 3708, "loss": 0.2417, "lr": 3.152540003560398e-05, "epoch": 1.4239482200647249, "percentage": 47.46, "elapsed_time": "1 day, 3:29:33", "remaining_time": "1 day, 6:25:46"}
+{"current_steps": 1770, "total_steps": 3708, "loss": 0.2288, "lr": 3.129791256133712e-05, "epoch": 1.4320388349514563, "percentage": 47.73, "elapsed_time": "1 day, 3:38:56", "remaining_time": "1 day, 6:16:23"}
+{"current_steps": 1780, "total_steps": 3708, "loss": 0.2347, "lr": 3.106986689874204e-05, "epoch": 1.4401294498381878, "percentage": 48.0, "elapsed_time": "1 day, 3:48:18", "remaining_time": "1 day, 6:07:01"}
+{"current_steps": 1790, "total_steps": 3708, "loss": 0.2385, "lr": 3.0841283259662875e-05, "epoch": 1.448220064724919, "percentage": 48.27, "elapsed_time": "1 day, 3:57:41", "remaining_time": "1 day, 5:57:39"}
+{"current_steps": 1800, "total_steps": 3708, "loss": 0.2359, "lr": 3.0612181903625014e-05, "epoch": 1.4563106796116505, "percentage": 48.54, "elapsed_time": "1 day, 4:07:02", "remaining_time": "1 day, 5:48:16"}
+{"current_steps": 1810, "total_steps": 3708, "loss": 0.2339, "lr": 3.0382583136039444e-05, "epoch": 1.4644012944983817, "percentage": 48.81, "elapsed_time": "1 day, 4:16:24", "remaining_time": "1 day, 5:38:53"}
+{"current_steps": 1820, "total_steps": 3708, "loss": 0.237, "lr": 3.015250730640308e-05, "epoch": 1.4724919093851132, "percentage": 49.08, "elapsed_time": "1 day, 4:25:46", "remaining_time": "1 day, 5:29:30"}
+{"current_steps": 1830, "total_steps": 3708, "loss": 0.2301, "lr": 2.9921974806495178e-05, "epoch": 1.4805825242718447, "percentage": 49.35, "elapsed_time": "1 day, 4:35:08", "remaining_time": "1 day, 5:20:08"}
+{"current_steps": 1840, "total_steps": 3708, "loss": 0.2339, "lr": 2.969100606856998e-05, "epoch": 1.4886731391585761, "percentage": 49.62, "elapsed_time": "1 day, 4:44:31", "remaining_time": "1 day, 5:10:45"}
+{"current_steps": 1850, "total_steps": 3708, "loss": 0.2385, "lr": 2.9459621563545825e-05, "epoch": 1.4967637540453074, "percentage": 49.89, "elapsed_time": "1 day, 4:53:53", "remaining_time": "1 day, 5:01:23"}
+{"current_steps": 1860, "total_steps": 3708, "loss": 0.2305, "lr": 2.9227841799190775e-05, "epoch": 1.5048543689320388, "percentage": 50.16, "elapsed_time": "1 day, 5:03:15", "remaining_time": "1 day, 4:52:00"}
+{"current_steps": 1870, "total_steps": 3708, "loss": 0.2328, "lr": 2.8995687318304975e-05, "epoch": 1.51294498381877, "percentage": 50.43, "elapsed_time": "1 day, 5:12:37", "remaining_time": "1 day, 4:42:37"}
+{"current_steps": 1880, "total_steps": 3708, "loss": 0.2373, "lr": 2.8763178696899995e-05, "epoch": 1.5210355987055015, "percentage": 50.7, "elapsed_time": "1 day, 5:21:59", "remaining_time": "1 day, 4:33:15"}
+{"current_steps": 1890, "total_steps": 3708, "loss": 0.2289, "lr": 2.853033654237507e-05, "epoch": 1.529126213592233, "percentage": 50.97, "elapsed_time": "1 day, 5:31:22", "remaining_time": "1 day, 4:23:53"}
+{"current_steps": 1900, "total_steps": 3708, "loss": 0.23, "lr": 2.8297181491690756e-05, "epoch": 1.5372168284789645, "percentage": 51.24, "elapsed_time": "1 day, 5:40:44", "remaining_time": "1 day, 4:14:30"}
+{"current_steps": 1910, "total_steps": 3708, "loss": 0.2367, "lr": 2.8063734209539773e-05, "epoch": 1.545307443365696, "percentage": 51.51, "elapsed_time": "1 day, 5:50:06", "remaining_time": "1 day, 4:05:08"}
+{"current_steps": 1920, "total_steps": 3708, "loss": 0.2292, "lr": 2.783001538651554e-05, "epoch": 1.5533980582524272, "percentage": 51.78, "elapsed_time": "1 day, 5:59:28", "remaining_time": "1 day, 3:55:45"}
+{"current_steps": 1930, "total_steps": 3708, "loss": 0.2398, "lr": 2.7596045737278336e-05, "epoch": 1.5614886731391586, "percentage": 52.05, "elapsed_time": "1 day, 6:08:51", "remaining_time": "1 day, 3:46:23"}
+{"current_steps": 1940, "total_steps": 3708, "loss": 0.2325, "lr": 2.7361845998719315e-05, "epoch": 1.5695792880258899, "percentage": 52.32, "elapsed_time": "1 day, 6:18:13", "remaining_time": "1 day, 3:37:00"}
+{"current_steps": 1950, "total_steps": 3708, "loss": 0.2367, "lr": 2.7127436928122612e-05, "epoch": 1.5776699029126213, "percentage": 52.59, "elapsed_time": "1 day, 6:27:35", "remaining_time": "1 day, 3:27:38"}
+{"current_steps": 1960, "total_steps": 3708, "loss": 0.239, "lr": 2.6892839301325623e-05, "epoch": 1.5857605177993528, "percentage": 52.86, "elapsed_time": "1 day, 6:36:57", "remaining_time": "1 day, 3:18:15"}
+{"current_steps": 1970, "total_steps": 3708, "loss": 0.2288, "lr": 2.6658073910877603e-05, "epoch": 1.5938511326860842, "percentage": 53.13, "elapsed_time": "1 day, 6:46:18", "remaining_time": "1 day, 3:08:52"}
+{"current_steps": 1980, "total_steps": 3708, "loss": 0.231, "lr": 2.6423161564196803e-05, "epoch": 1.6019417475728155, "percentage": 53.4, "elapsed_time": "1 day, 6:55:40", "remaining_time": "1 day, 2:59:30"}
+{"current_steps": 1990, "total_steps": 3708, "loss": 0.2341, "lr": 2.6188123081726306e-05, "epoch": 1.610032362459547, "percentage": 53.67, "elapsed_time": "1 day, 7:05:03", "remaining_time": "1 day, 2:50:07"}
+{"current_steps": 2000, "total_steps": 3708, "loss": 0.2303, "lr": 2.5952979295088714e-05, "epoch": 1.6181229773462782, "percentage": 53.94, "elapsed_time": "1 day, 7:14:25", "remaining_time": "1 day, 2:40:45"}
+{"current_steps": 2010, "total_steps": 3708, "loss": 0.2297, "lr": 2.57177510452398e-05, "epoch": 1.6262135922330097, "percentage": 54.21, "elapsed_time": "1 day, 7:23:47", "remaining_time": "1 day, 2:31:22"}
+{"current_steps": 2020, "total_steps": 3708, "loss": 0.2336, "lr": 2.5482459180621377e-05, "epoch": 1.6343042071197411, "percentage": 54.48, "elapsed_time": "1 day, 7:33:09", "remaining_time": "1 day, 2:22:00"}
+{"current_steps": 2030, "total_steps": 3708, "loss": 0.2283, "lr": 2.524712455531347e-05, "epoch": 1.6423948220064726, "percentage": 54.75, "elapsed_time": "1 day, 7:42:31", "remaining_time": "1 day, 2:12:38"}
+{"current_steps": 2040, "total_steps": 3708, "loss": 0.2367, "lr": 2.501176802718599e-05, "epoch": 1.650485436893204, "percentage": 55.02, "elapsed_time": "1 day, 7:51:54", "remaining_time": "1 day, 2:03:15"}
+{"current_steps": 2050, "total_steps": 3708, "loss": 0.232, "lr": 2.4776410456050165e-05, "epoch": 1.6585760517799353, "percentage": 55.29, "elapsed_time": "1 day, 8:01:16", "remaining_time": "1 day, 1:53:53"}
+{"current_steps": 2060, "total_steps": 3708, "loss": 0.2348, "lr": 2.4541072701809624e-05, "epoch": 1.6666666666666665, "percentage": 55.56, "elapsed_time": "1 day, 8:10:38", "remaining_time": "1 day, 1:44:30"}
+{"current_steps": 2070, "total_steps": 3708, "loss": 0.2285, "lr": 2.4305775622611627e-05, "epoch": 1.674757281553398, "percentage": 55.83, "elapsed_time": "1 day, 8:20:00", "remaining_time": "1 day, 1:35:08"}
+{"current_steps": 2080, "total_steps": 3708, "loss": 0.2272, "lr": 2.4070540072998372e-05, "epoch": 1.6828478964401294, "percentage": 56.09, "elapsed_time": "1 day, 8:29:22", "remaining_time": "1 day, 1:25:45"}
+{"current_steps": 2090, "total_steps": 3708, "loss": 0.2292, "lr": 2.3835386902058637e-05, "epoch": 1.690938511326861, "percentage": 56.36, "elapsed_time": "1 day, 8:38:44", "remaining_time": "1 day, 1:16:23"}
+{"current_steps": 2100, "total_steps": 3708, "loss": 0.2337, "lr": 2.360033695157995e-05, "epoch": 1.6990291262135924, "percentage": 56.63, "elapsed_time": "1 day, 8:48:06", "remaining_time": "1 day, 1:07:00"}
+{"current_steps": 2110, "total_steps": 3708, "loss": 0.2265, "lr": 2.3365411054201315e-05, "epoch": 1.7071197411003236, "percentage": 56.9, "elapsed_time": "1 day, 8:57:27", "remaining_time": "1 day, 0:57:37"}
+{"current_steps": 2120, "total_steps": 3708, "loss": 0.2269, "lr": 2.3130630031566818e-05, "epoch": 1.715210355987055, "percentage": 57.17, "elapsed_time": "1 day, 9:06:49", "remaining_time": "1 day, 0:48:15"}
+{"current_steps": 2130, "total_steps": 3708, "loss": 0.231, "lr": 2.2896014692480226e-05, "epoch": 1.7233009708737863, "percentage": 57.44, "elapsed_time": "1 day, 9:16:12", "remaining_time": "1 day, 0:38:52"}
+{"current_steps": 2140, "total_steps": 3708, "loss": 0.228, "lr": 2.266158583106063e-05, "epoch": 1.7313915857605178, "percentage": 57.71, "elapsed_time": "1 day, 9:25:34", "remaining_time": "1 day, 0:29:30"}
+{"current_steps": 2150, "total_steps": 3708, "loss": 0.2289, "lr": 2.2427364224899502e-05, "epoch": 1.7394822006472492, "percentage": 57.98, "elapsed_time": "1 day, 9:34:56", "remaining_time": "1 day, 0:20:07"}
+{"current_steps": 2160, "total_steps": 3708, "loss": 0.2295, "lr": 2.2193370633219115e-05, "epoch": 1.7475728155339807, "percentage": 58.25, "elapsed_time": "1 day, 9:44:18", "remaining_time": "1 day, 0:10:45"}
+{"current_steps": 2170, "total_steps": 3708, "loss": 0.2312, "lr": 2.1959625795032664e-05, "epoch": 1.755663430420712, "percentage": 58.52, "elapsed_time": "1 day, 9:53:41", "remaining_time": "1 day, 0:01:23"}
+{"current_steps": 2180, "total_steps": 3708, "loss": 0.2286, "lr": 2.1726150427306182e-05, "epoch": 1.7637540453074434, "percentage": 58.79, "elapsed_time": "1 day, 10:03:03", "remaining_time": "23:52:00"}
+{"current_steps": 2190, "total_steps": 3708, "loss": 0.2267, "lr": 2.1492965223122305e-05, "epoch": 1.7718446601941746, "percentage": 59.06, "elapsed_time": "1 day, 10:12:25", "remaining_time": "23:42:38"}
+{"current_steps": 2200, "total_steps": 3708, "loss": 0.2251, "lr": 2.126009084984629e-05, "epoch": 1.779935275080906, "percentage": 59.33, "elapsed_time": "1 day, 10:21:47", "remaining_time": "23:33:16"}
+{"current_steps": 2210, "total_steps": 3708, "loss": 0.2246, "lr": 2.102754794729426e-05, "epoch": 1.7880258899676376, "percentage": 59.6, "elapsed_time": "1 day, 10:31:09", "remaining_time": "23:23:53"}
+{"current_steps": 2220, "total_steps": 3708, "loss": 0.2282, "lr": 2.079535712590382e-05, "epoch": 1.796116504854369, "percentage": 59.87, "elapsed_time": "1 day, 10:40:31", "remaining_time": "23:14:31"}
+{"current_steps": 2230, "total_steps": 3708, "loss": 0.2231, "lr": 2.056353896490742e-05, "epoch": 1.8042071197411005, "percentage": 60.14, "elapsed_time": "1 day, 10:49:54", "remaining_time": "23:05:08"}
+{"current_steps": 2240, "total_steps": 3708, "loss": 0.2268, "lr": 2.0332114010508334e-05, "epoch": 1.8122977346278317, "percentage": 60.41, "elapsed_time": "1 day, 10:59:16", "remaining_time": "22:55:46"}
+{"current_steps": 2250, "total_steps": 3708, "loss": 0.2274, "lr": 2.010110277405966e-05, "epoch": 1.820388349514563, "percentage": 60.68, "elapsed_time": "1 day, 11:08:38", "remaining_time": "22:46:24"}
+{"current_steps": 2260, "total_steps": 3708, "loss": 0.2293, "lr": 1.9870525730246424e-05, "epoch": 1.8284789644012944, "percentage": 60.95, "elapsed_time": "1 day, 11:18:01", "remaining_time": "22:37:01"}
+{"current_steps": 2270, "total_steps": 3708, "loss": 0.2286, "lr": 1.9640403315270824e-05, "epoch": 1.8365695792880259, "percentage": 61.22, "elapsed_time": "1 day, 11:27:23", "remaining_time": "22:27:39"}
+{"current_steps": 2280, "total_steps": 3708, "loss": 0.2257, "lr": 1.9410755925041006e-05, "epoch": 1.8446601941747574, "percentage": 61.49, "elapsed_time": "1 day, 11:36:45", "remaining_time": "22:18:17"}
+{"current_steps": 2290, "total_steps": 3708, "loss": 0.2259, "lr": 1.918160391336335e-05, "epoch": 1.8527508090614888, "percentage": 61.76, "elapsed_time": "1 day, 11:46:07", "remaining_time": "22:08:54"}
+{"current_steps": 2300, "total_steps": 3708, "loss": 0.2248, "lr": 1.8952967590138472e-05, "epoch": 1.86084142394822, "percentage": 62.03, "elapsed_time": "1 day, 11:55:29", "remaining_time": "21:59:31"}
+{"current_steps": 2310, "total_steps": 3708, "loss": 0.2287, "lr": 1.8724867219561203e-05, "epoch": 1.8689320388349513, "percentage": 62.3, "elapsed_time": "1 day, 12:04:51", "remaining_time": "21:50:09"}
+{"current_steps": 2320, "total_steps": 3708, "loss": 0.2252, "lr": 1.8497323018324476e-05, "epoch": 1.8770226537216828, "percentage": 62.57, "elapsed_time": "1 day, 12:14:13", "remaining_time": "21:40:47"}
+{"current_steps": 2330, "total_steps": 3708, "loss": 0.2243, "lr": 1.8270355153827598e-05, "epoch": 1.8851132686084142, "percentage": 62.84, "elapsed_time": "1 day, 12:23:35", "remaining_time": "21:31:24"}
+{"current_steps": 2340, "total_steps": 3708, "loss": 0.2231, "lr": 1.804398374238872e-05, "epoch": 1.8932038834951457, "percentage": 63.11, "elapsed_time": "1 day, 12:32:57", "remaining_time": "21:22:02"}
+{"current_steps": 2350, "total_steps": 3708, "loss": 0.2203, "lr": 1.781822884746196e-05, "epoch": 1.9012944983818771, "percentage": 63.38, "elapsed_time": "1 day, 12:42:20", "remaining_time": "21:12:40"}
+{"current_steps": 2360, "total_steps": 3708, "loss": 0.223, "lr": 1.7593110477859153e-05, "epoch": 1.9093851132686084, "percentage": 63.65, "elapsed_time": "1 day, 12:51:42", "remaining_time": "21:03:17"}
+{"current_steps": 2370, "total_steps": 3708, "loss": 0.2233, "lr": 1.736864858597645e-05, "epoch": 1.9174757281553398, "percentage": 63.92, "elapsed_time": "1 day, 13:01:04", "remaining_time": "20:53:55"}
+{"current_steps": 2380, "total_steps": 3708, "loss": 0.2193, "lr": 1.7144863066025955e-05, "epoch": 1.925566343042071, "percentage": 64.19, "elapsed_time": "1 day, 13:10:26", "remaining_time": "20:44:32"}
+{"current_steps": 2390, "total_steps": 3708, "loss": 0.2251, "lr": 1.692177375227242e-05, "epoch": 1.9336569579288025, "percentage": 64.46, "elapsed_time": "1 day, 13:19:48", "remaining_time": "20:35:10"}
+{"current_steps": 2400, "total_steps": 3708, "loss": 0.218, "lr": 1.669940041727538e-05, "epoch": 1.941747572815534, "percentage": 64.72, "elapsed_time": "1 day, 13:29:10", "remaining_time": "20:25:48"}
+{"current_steps": 2410, "total_steps": 3708, "loss": 0.225, "lr": 1.6477762770136707e-05, "epoch": 1.9498381877022655, "percentage": 64.99, "elapsed_time": "1 day, 13:38:32", "remaining_time": "20:16:25"}
+{"current_steps": 2420, "total_steps": 3708, "loss": 0.2208, "lr": 1.625688045475371e-05, "epoch": 1.9579288025889967, "percentage": 65.26, "elapsed_time": "1 day, 13:47:54", "remaining_time": "20:07:03"}
+{"current_steps": 2430, "total_steps": 3708, "loss": 0.2223, "lr": 1.603677304807815e-05, "epoch": 1.9660194174757282, "percentage": 65.53, "elapsed_time": "1 day, 13:57:16", "remaining_time": "19:57:40"}
+{"current_steps": 2440, "total_steps": 3708, "loss": 0.2274, "lr": 1.5817460058381088e-05, "epoch": 1.9741100323624594, "percentage": 65.8, "elapsed_time": "1 day, 14:06:39", "remaining_time": "19:48:18"}
+{"current_steps": 2450, "total_steps": 3708, "loss": 0.2248, "lr": 1.5598960923523842e-05, "epoch": 1.9822006472491909, "percentage": 66.07, "elapsed_time": "1 day, 14:16:01", "remaining_time": "19:38:56"}
+{"current_steps": 2460, "total_steps": 3708, "loss": 0.2277, "lr": 1.5381295009235262e-05, "epoch": 1.9902912621359223, "percentage": 66.34, "elapsed_time": "1 day, 14:25:23", "remaining_time": "19:29:33"}
+{"current_steps": 2470, "total_steps": 3708, "loss": 0.2243, "lr": 1.5164481607395238e-05, "epoch": 1.9983818770226538, "percentage": 66.61, "elapsed_time": "1 day, 14:34:45", "remaining_time": "19:20:11"}
+{"current_steps": 2480, "total_steps": 3708, "loss": 0.1818, "lr": 1.4948539934324923e-05, "epoch": 2.0064724919093853, "percentage": 66.88, "elapsed_time": "1 day, 14:44:43", "remaining_time": "19:11:06"}
+{"current_steps": 2490, "total_steps": 3708, "loss": 0.1718, "lr": 1.4733489129083534e-05, "epoch": 2.0145631067961167, "percentage": 67.15, "elapsed_time": "1 day, 14:54:06", "remaining_time": "19:01:44"}
+{"current_steps": 2500, "total_steps": 3708, "loss": 0.1696, "lr": 1.4519348251772058e-05, "epoch": 2.0226537216828477, "percentage": 67.42, "elapsed_time": "1 day, 15:03:28", "remaining_time": "18:52:21"}
+{"current_steps": 2510, "total_steps": 3708, "loss": 0.1691, "lr": 1.4306136281843962e-05, "epoch": 2.030744336569579, "percentage": 67.69, "elapsed_time": "1 day, 15:12:50", "remaining_time": "18:42:59"}
+{"current_steps": 2520, "total_steps": 3708, "loss": 0.169, "lr": 1.4093872116422979e-05, "epoch": 2.0388349514563107, "percentage": 67.96, "elapsed_time": "1 day, 15:22:12", "remaining_time": "18:33:36"}
+{"current_steps": 2530, "total_steps": 3708, "loss": 0.168, "lr": 1.3882574568628315e-05, "epoch": 2.046925566343042, "percentage": 68.23, "elapsed_time": "1 day, 15:31:34", "remaining_time": "18:24:14"}
+{"current_steps": 2540, "total_steps": 3708, "loss": 0.1684, "lr": 1.3672262365907163e-05, "epoch": 2.0550161812297736, "percentage": 68.5, "elapsed_time": "1 day, 15:40:57", "remaining_time": "18:14:51"}
+{"current_steps": 2550, "total_steps": 3708, "loss": 0.1661, "lr": 1.3462954148374899e-05, "epoch": 2.063106796116505, "percentage": 68.77, "elapsed_time": "1 day, 15:50:19", "remaining_time": "18:05:29"}
+{"current_steps": 2560, "total_steps": 3708, "loss": 0.1696, "lr": 1.3254668467163029e-05, "epoch": 2.071197411003236, "percentage": 69.04, "elapsed_time": "1 day, 15:59:41", "remaining_time": "17:56:06"}
+{"current_steps": 2570, "total_steps": 3708, "loss": 0.1649, "lr": 1.3047423782774937e-05, "epoch": 2.0792880258899675, "percentage": 69.31, "elapsed_time": "1 day, 16:09:03", "remaining_time": "17:46:44"}
+{"current_steps": 2580, "total_steps": 3708, "loss": 0.1719, "lr": 1.2841238463449743e-05, "epoch": 2.087378640776699, "percentage": 69.58, "elapsed_time": "1 day, 16:18:25", "remaining_time": "17:37:21"}
+{"current_steps": 2590, "total_steps": 3708, "loss": 0.1703, "lr": 1.2636130783534319e-05, "epoch": 2.0954692556634305, "percentage": 69.85, "elapsed_time": "1 day, 16:27:47", "remaining_time": "17:27:58"}
+{"current_steps": 2600, "total_steps": 3708, "loss": 0.168, "lr": 1.2432118921863604e-05, "epoch": 2.103559870550162, "percentage": 70.12, "elapsed_time": "1 day, 16:37:09", "remaining_time": "17:18:36"}
+{"current_steps": 2610, "total_steps": 3708, "loss": 0.1695, "lr": 1.2229220960149431e-05, "epoch": 2.1116504854368934, "percentage": 70.39, "elapsed_time": "1 day, 16:46:31", "remaining_time": "17:09:13"}
+{"current_steps": 2620, "total_steps": 3708, "loss": 0.1675, "lr": 1.2027454881377889e-05, "epoch": 2.1197411003236244, "percentage": 70.66, "elapsed_time": "1 day, 16:55:53", "remaining_time": "16:59:51"}
+{"current_steps": 2630, "total_steps": 3708, "loss": 0.1685, "lr": 1.1826838568215526e-05, "epoch": 2.127831715210356, "percentage": 70.93, "elapsed_time": "1 day, 17:05:16", "remaining_time": "16:50:28"}
+{"current_steps": 2640, "total_steps": 3708, "loss": 0.1706, "lr": 1.1627389801424351e-05, "epoch": 2.1359223300970873, "percentage": 71.2, "elapsed_time": "1 day, 17:14:38", "remaining_time": "16:41:06"}
+{"current_steps": 2650, "total_steps": 3708, "loss": 0.1661, "lr": 1.1429126258285946e-05, "epoch": 2.144012944983819, "percentage": 71.47, "elapsed_time": "1 day, 17:24:00", "remaining_time": "16:31:43"}
+{"current_steps": 2660, "total_steps": 3708, "loss": 0.1663, "lr": 1.1232065511034696e-05, "epoch": 2.1521035598705502, "percentage": 71.74, "elapsed_time": "1 day, 17:33:21", "remaining_time": "16:22:20"}
+{"current_steps": 2670, "total_steps": 3708, "loss": 0.1687, "lr": 1.1036225025300357e-05, "epoch": 2.1601941747572817, "percentage": 72.01, "elapsed_time": "1 day, 17:42:43", "remaining_time": "16:12:58"}
+{"current_steps": 2680, "total_steps": 3708, "loss": 0.1668, "lr": 1.0841622158560085e-05, "epoch": 2.168284789644013, "percentage": 72.28, "elapsed_time": "1 day, 17:52:05", "remaining_time": "16:03:35"}
+{"current_steps": 2690, "total_steps": 3708, "loss": 0.1698, "lr": 1.0648274158599994e-05, "epoch": 2.176375404530744, "percentage": 72.55, "elapsed_time": "1 day, 18:01:27", "remaining_time": "15:54:13"}
+{"current_steps": 2700, "total_steps": 3708, "loss": 0.1753, "lr": 1.0456198161986489e-05, "epoch": 2.1844660194174756, "percentage": 72.82, "elapsed_time": "1 day, 18:10:49", "remaining_time": "15:44:50"}
+{"current_steps": 2710, "total_steps": 3708, "loss": 0.1718, "lr": 1.0265411192547462e-05, "epoch": 2.192556634304207, "percentage": 73.09, "elapsed_time": "1 day, 18:20:10", "remaining_time": "15:35:27"}
+{"current_steps": 2720, "total_steps": 3708, "loss": 0.1661, "lr": 1.0075930159863416e-05, "epoch": 2.2006472491909386, "percentage": 73.35, "elapsed_time": "1 day, 18:29:33", "remaining_time": "15:26:05"}
+{"current_steps": 2730, "total_steps": 3708, "loss": 0.1683, "lr": 9.887771857768796e-06, "epoch": 2.20873786407767, "percentage": 73.62, "elapsed_time": "1 day, 18:38:55", "remaining_time": "15:16:42"}
+{"current_steps": 2740, "total_steps": 3708, "loss": 0.1701, "lr": 9.700952962863513e-06, "epoch": 2.2168284789644015, "percentage": 73.89, "elapsed_time": "1 day, 18:48:17", "remaining_time": "15:07:20"}
+{"current_steps": 2750, "total_steps": 3708, "loss": 0.1667, "lr": 9.515490033034893e-06, "epoch": 2.2249190938511325, "percentage": 74.16, "elapsed_time": "1 day, 18:57:38", "remaining_time": "14:57:57"}
+{"current_steps": 2760, "total_steps": 3708, "loss": 0.1655, "lr": 9.331399505990168e-06, "epoch": 2.233009708737864, "percentage": 74.43, "elapsed_time": "1 day, 19:07:00", "remaining_time": "14:48:34"}
+{"current_steps": 2770, "total_steps": 3708, "loss": 0.1694, "lr": 9.148697697799533e-06, "epoch": 2.2411003236245954, "percentage": 74.7, "elapsed_time": "1 day, 19:16:22", "remaining_time": "14:39:12"}
+{"current_steps": 2780, "total_steps": 3708, "loss": 0.1685, "lr": 8.967400801450105e-06, "epoch": 2.249190938511327, "percentage": 74.97, "elapsed_time": "1 day, 19:25:44", "remaining_time": "14:29:49"}
+{"current_steps": 2790, "total_steps": 3708, "loss": 0.1638, "lr": 8.787524885410678e-06, "epoch": 2.2572815533980584, "percentage": 75.24, "elapsed_time": "1 day, 19:35:06", "remaining_time": "14:20:27"}
+{"current_steps": 2800, "total_steps": 3708, "loss": 0.1676, "lr": 8.60908589220758e-06, "epoch": 2.26537216828479, "percentage": 75.51, "elapsed_time": "1 day, 19:44:28", "remaining_time": "14:11:04"}
+{"current_steps": 2810, "total_steps": 3708, "loss": 0.1657, "lr": 8.432099637011693e-06, "epoch": 2.273462783171521, "percentage": 75.78, "elapsed_time": "1 day, 19:53:50", "remaining_time": "14:01:42"}
+{"current_steps": 2820, "total_steps": 3708, "loss": 0.1649, "lr": 8.256581806236704e-06, "epoch": 2.2815533980582523, "percentage": 76.05, "elapsed_time": "1 day, 20:03:12", "remaining_time": "13:52:19"}
+{"current_steps": 2830, "total_steps": 3708, "loss": 0.1663, "lr": 8.082547956148873e-06, "epoch": 2.2896440129449838, "percentage": 76.32, "elapsed_time": "1 day, 20:12:35", "remaining_time": "13:42:57"}
+{"current_steps": 2840, "total_steps": 3708, "loss": 0.1653, "lr": 7.91001351148819e-06, "epoch": 2.2977346278317152, "percentage": 76.59, "elapsed_time": "1 day, 20:21:57", "remaining_time": "13:33:35"}
+{"current_steps": 2850, "total_steps": 3708, "loss": 0.1677, "lr": 7.738993764101324e-06, "epoch": 2.3058252427184467, "percentage": 76.86, "elapsed_time": "1 day, 20:31:19", "remaining_time": "13:24:12"}
+{"current_steps": 2860, "total_steps": 3708, "loss": 0.1674, "lr": 7.569503871586292e-06, "epoch": 2.313915857605178, "percentage": 77.13, "elapsed_time": "1 day, 20:40:41", "remaining_time": "13:14:50"}
+{"current_steps": 2870, "total_steps": 3708, "loss": 0.1678, "lr": 7.401558855949004e-06, "epoch": 2.3220064724919096, "percentage": 77.4, "elapsed_time": "1 day, 20:50:03", "remaining_time": "13:05:27"}
+{"current_steps": 2880, "total_steps": 3708, "loss": 0.162, "lr": 7.235173602271875e-06, "epoch": 2.3300970873786406, "percentage": 77.67, "elapsed_time": "1 day, 20:59:26", "remaining_time": "12:56:05"}
+{"current_steps": 2890, "total_steps": 3708, "loss": 0.1712, "lr": 7.070362857394538e-06, "epoch": 2.338187702265372, "percentage": 77.94, "elapsed_time": "1 day, 21:08:49", "remaining_time": "12:46:43"}
+{"current_steps": 2900, "total_steps": 3708, "loss": 0.1664, "lr": 6.907141228606831e-06, "epoch": 2.3462783171521036, "percentage": 78.21, "elapsed_time": "1 day, 21:18:11", "remaining_time": "12:37:20"}
+{"current_steps": 2910, "total_steps": 3708, "loss": 0.1634, "lr": 6.745523182354147e-06, "epoch": 2.354368932038835, "percentage": 78.48, "elapsed_time": "1 day, 21:27:33", "remaining_time": "12:27:58"}
+{"current_steps": 2920, "total_steps": 3708, "loss": 0.1633, "lr": 6.585523042955233e-06, "epoch": 2.3624595469255665, "percentage": 78.75, "elapsed_time": "1 day, 21:36:55", "remaining_time": "12:18:35"}
+{"current_steps": 2930, "total_steps": 3708, "loss": 0.1665, "lr": 6.427154991332665e-06, "epoch": 2.3705501618122975, "percentage": 79.02, "elapsed_time": "1 day, 21:46:17", "remaining_time": "12:09:13"}
+{"current_steps": 2940, "total_steps": 3708, "loss": 0.1632, "lr": 6.2704330637559315e-06, "epoch": 2.378640776699029, "percentage": 79.29, "elapsed_time": "1 day, 21:55:39", "remaining_time": "11:59:50"}
+{"current_steps": 2950, "total_steps": 3708, "loss": 0.1657, "lr": 6.115371150597413e-06, "epoch": 2.3867313915857604, "percentage": 79.56, "elapsed_time": "1 day, 22:05:02", "remaining_time": "11:50:28"}
+{"current_steps": 2960, "total_steps": 3708, "loss": 0.1652, "lr": 5.961982995101301e-06, "epoch": 2.394822006472492, "percentage": 79.83, "elapsed_time": "1 day, 22:14:24", "remaining_time": "11:41:06"}
+{"current_steps": 2970, "total_steps": 3708, "loss": 0.1674, "lr": 5.810282192165442e-06, "epoch": 2.4029126213592233, "percentage": 80.1, "elapsed_time": "1 day, 22:23:47", "remaining_time": "11:31:43"}
+{"current_steps": 2980, "total_steps": 3708, "loss": 0.1648, "lr": 5.660282187136507e-06, "epoch": 2.411003236245955, "percentage": 80.37, "elapsed_time": "1 day, 22:33:10", "remaining_time": "11:22:21"}
+{"current_steps": 2990, "total_steps": 3708, "loss": 0.1673, "lr": 5.511996274618253e-06, "epoch": 2.4190938511326863, "percentage": 80.64, "elapsed_time": "1 day, 22:42:32", "remaining_time": "11:12:59"}
+{"current_steps": 3000, "total_steps": 3708, "loss": 0.1681, "lr": 5.365437597293238e-06, "epoch": 2.4271844660194173, "percentage": 80.91, "elapsed_time": "1 day, 22:51:55", "remaining_time": "11:03:36"}
+{"current_steps": 3010, "total_steps": 3708, "loss": 0.168, "lr": 5.220619144757996e-06, "epoch": 2.4352750809061487, "percentage": 81.18, "elapsed_time": "1 day, 23:01:19", "remaining_time": "10:54:14"}
+{"current_steps": 3020, "total_steps": 3708, "loss": 0.1695, "lr": 5.077553752371708e-06, "epoch": 2.44336569579288, "percentage": 81.45, "elapsed_time": "1 day, 23:10:43", "remaining_time": "10:44:52"}
+{"current_steps": 3030, "total_steps": 3708, "loss": 0.1626, "lr": 4.936254100118656e-06, "epoch": 2.4514563106796117, "percentage": 81.72, "elapsed_time": "1 day, 23:20:05", "remaining_time": "10:35:30"}
+{"current_steps": 3040, "total_steps": 3708, "loss": 0.1688, "lr": 4.796732711484342e-06, "epoch": 2.459546925566343, "percentage": 81.98, "elapsed_time": "1 day, 23:29:29", "remaining_time": "10:26:08"}
+{"current_steps": 3050, "total_steps": 3708, "loss": 0.1618, "lr": 4.659001952345538e-06, "epoch": 2.4676375404530746, "percentage": 82.25, "elapsed_time": "1 day, 23:38:52", "remaining_time": "10:16:46"}
+{"current_steps": 3060, "total_steps": 3708, "loss": 0.1611, "lr": 4.523074029874291e-06, "epoch": 2.475728155339806, "percentage": 82.52, "elapsed_time": "1 day, 23:48:16", "remaining_time": "10:07:23"}
+{"current_steps": 3070, "total_steps": 3708, "loss": 0.1642, "lr": 4.388960991455998e-06, "epoch": 2.483818770226537, "percentage": 82.79, "elapsed_time": "1 day, 23:57:39", "remaining_time": "9:58:01"}
+{"current_steps": 3080, "total_steps": 3708, "loss": 0.165, "lr": 4.256674723621621e-06, "epoch": 2.4919093851132685, "percentage": 83.06, "elapsed_time": "2 days, 0:07:02", "remaining_time": "9:48:39"}
+{"current_steps": 3090, "total_steps": 3708, "loss": 0.1633, "lr": 4.126226950994211e-06, "epoch": 2.5, "percentage": 83.33, "elapsed_time": "2 days, 0:16:24", "remaining_time": "9:39:16"}
+{"current_steps": 3100, "total_steps": 3708, "loss": 0.1645, "lr": 3.997629235249692e-06, "epoch": 2.5080906148867315, "percentage": 83.6, "elapsed_time": "2 days, 0:25:46", "remaining_time": "9:29:54"}
+{"current_steps": 3110, "total_steps": 3708, "loss": 0.1657, "lr": 3.870892974092197e-06, "epoch": 2.516181229773463, "percentage": 83.87, "elapsed_time": "2 days, 0:35:08", "remaining_time": "9:20:31"}
+{"current_steps": 3120, "total_steps": 3708, "loss": 0.1649, "lr": 3.7460294002438444e-06, "epoch": 2.524271844660194, "percentage": 84.14, "elapsed_time": "2 days, 0:44:30", "remaining_time": "9:11:09"}
+{"current_steps": 3130, "total_steps": 3708, "loss": 0.1638, "lr": 3.6230495804491864e-06, "epoch": 2.5323624595469254, "percentage": 84.41, "elapsed_time": "2 days, 0:53:52", "remaining_time": "9:01:46"}
+{"current_steps": 3140, "total_steps": 3708, "loss": 0.1657, "lr": 3.5019644144943576e-06, "epoch": 2.540453074433657, "percentage": 84.68, "elapsed_time": "2 days, 1:03:14", "remaining_time": "8:52:24"}
+{"current_steps": 3150, "total_steps": 3708, "loss": 0.1626, "lr": 3.382784634241015e-06, "epoch": 2.5485436893203883, "percentage": 84.95, "elapsed_time": "2 days, 1:12:36", "remaining_time": "8:43:02"}
+{"current_steps": 3160, "total_steps": 3708, "loss": 0.1645, "lr": 3.2655208026751816e-06, "epoch": 2.55663430420712, "percentage": 85.22, "elapsed_time": "2 days, 1:21:59", "remaining_time": "8:33:39"}
+{"current_steps": 3170, "total_steps": 3708, "loss": 0.1625, "lr": 3.150183312971014e-06, "epoch": 2.5647249190938513, "percentage": 85.49, "elapsed_time": "2 days, 1:31:23", "remaining_time": "8:24:17"}
+{"current_steps": 3180, "total_steps": 3708, "loss": 0.1623, "lr": 3.036782387569659e-06, "epoch": 2.5728155339805827, "percentage": 85.76, "elapsed_time": "2 days, 1:40:45", "remaining_time": "8:14:55"}
+{"current_steps": 3190, "total_steps": 3708, "loss": 0.1627, "lr": 2.9253280772732595e-06, "epoch": 2.5809061488673137, "percentage": 86.03, "elapsed_time": "2 days, 1:50:10", "remaining_time": "8:05:33"}
+{"current_steps": 3200, "total_steps": 3708, "loss": 0.1609, "lr": 2.8158302603540965e-06, "epoch": 2.588996763754045, "percentage": 86.3, "elapsed_time": "2 days, 1:59:34", "remaining_time": "7:56:10"}
+{"current_steps": 3210, "total_steps": 3708, "loss": 0.1642, "lr": 2.708298641679105e-06, "epoch": 2.5970873786407767, "percentage": 86.57, "elapsed_time": "2 days, 2:08:56", "remaining_time": "7:46:48"}
+{"current_steps": 3220, "total_steps": 3708, "loss": 0.1634, "lr": 2.6027427518497153e-06, "epoch": 2.605177993527508, "percentage": 86.84, "elapsed_time": "2 days, 2:18:20", "remaining_time": "7:37:26"}
+{"current_steps": 3230, "total_steps": 3708, "loss": 0.1607, "lr": 2.49917194635714e-06, "epoch": 2.6132686084142396, "percentage": 87.11, "elapsed_time": "2 days, 2:27:44", "remaining_time": "7:28:04"}
+{"current_steps": 3240, "total_steps": 3708, "loss": 0.1589, "lr": 2.397595404753225e-06, "epoch": 2.6213592233009706, "percentage": 87.38, "elapsed_time": "2 days, 2:37:08", "remaining_time": "7:18:41"}
+{"current_steps": 3250, "total_steps": 3708, "loss": 0.1679, "lr": 2.2980221298367995e-06, "epoch": 2.6294498381877025, "percentage": 87.65, "elapsed_time": "2 days, 2:46:32", "remaining_time": "7:09:19"}
+{"current_steps": 3260, "total_steps": 3708, "loss": 0.1648, "lr": 2.2004609468558175e-06, "epoch": 2.6375404530744335, "percentage": 87.92, "elapsed_time": "2 days, 2:55:56", "remaining_time": "6:59:57"}
+{"current_steps": 3270, "total_steps": 3708, "loss": 0.1648, "lr": 2.1049205027251216e-06, "epoch": 2.645631067961165, "percentage": 88.19, "elapsed_time": "2 days, 3:05:20", "remaining_time": "6:50:35"}
+{"current_steps": 3280, "total_steps": 3708, "loss": 0.1669, "lr": 2.0114092652600806e-06, "epoch": 2.6537216828478964, "percentage": 88.46, "elapsed_time": "2 days, 3:14:42", "remaining_time": "6:41:12"}
+{"current_steps": 3290, "total_steps": 3708, "loss": 0.1601, "lr": 1.919935522426081e-06, "epoch": 2.661812297734628, "percentage": 88.73, "elapsed_time": "2 days, 3:24:04", "remaining_time": "6:31:50"}
+{"current_steps": 3300, "total_steps": 3708, "loss": 0.165, "lr": 1.8305073816039492e-06, "epoch": 2.6699029126213594, "percentage": 89.0, "elapsed_time": "2 days, 3:33:26", "remaining_time": "6:22:27"}
+{"current_steps": 3310, "total_steps": 3708, "loss": 0.1686, "lr": 1.7431327688714139e-06, "epoch": 2.6779935275080904, "percentage": 89.27, "elapsed_time": "2 days, 3:42:49", "remaining_time": "6:13:05"}
+{"current_steps": 3320, "total_steps": 3708, "loss": 0.1586, "lr": 1.6578194283005804e-06, "epoch": 2.686084142394822, "percentage": 89.54, "elapsed_time": "2 days, 3:52:11", "remaining_time": "6:03:42"}
+{"current_steps": 3330, "total_steps": 3708, "loss": 0.1669, "lr": 1.5745749212715794e-06, "epoch": 2.6941747572815533, "percentage": 89.81, "elapsed_time": "2 days, 4:01:34", "remaining_time": "5:54:20"}
+{"current_steps": 3340, "total_steps": 3708, "loss": 0.1681, "lr": 1.4934066258024182e-06, "epoch": 2.7022653721682848, "percentage": 90.08, "elapsed_time": "2 days, 4:10:57", "remaining_time": "5:44:58"}
+{"current_steps": 3350, "total_steps": 3708, "loss": 0.1648, "lr": 1.4143217358950217e-06, "epoch": 2.7103559870550162, "percentage": 90.35, "elapsed_time": "2 days, 4:20:21", "remaining_time": "5:35:35"}
+{"current_steps": 3360, "total_steps": 3708, "loss": 0.1635, "lr": 1.3373272608976668e-06, "epoch": 2.7184466019417477, "percentage": 90.61, "elapsed_time": "2 days, 4:29:43", "remaining_time": "5:26:13"}
+{"current_steps": 3370, "total_steps": 3708, "loss": 0.1607, "lr": 1.2624300248836928e-06, "epoch": 2.726537216828479, "percentage": 90.88, "elapsed_time": "2 days, 4:39:08", "remaining_time": "5:16:51"}
+{"current_steps": 3380, "total_steps": 3708, "loss": 0.1645, "lr": 1.1896366660467173e-06, "epoch": 2.73462783171521, "percentage": 91.15, "elapsed_time": "2 days, 4:48:30", "remaining_time": "5:07:28"}
+{"current_steps": 3390, "total_steps": 3708, "loss": 0.1628, "lr": 1.1189536361122799e-06, "epoch": 2.7427184466019416, "percentage": 91.42, "elapsed_time": "2 days, 4:57:52", "remaining_time": "4:58:06"}
+{"current_steps": 3400, "total_steps": 3708, "loss": 0.168, "lr": 1.0503871997660036e-06, "epoch": 2.750809061488673, "percentage": 91.69, "elapsed_time": "2 days, 5:07:14", "remaining_time": "4:48:43"}
+{"current_steps": 3410, "total_steps": 3708, "loss": 0.1633, "lr": 9.83943434098372e-07, "epoch": 2.7588996763754046, "percentage": 91.96, "elapsed_time": "2 days, 5:16:37", "remaining_time": "4:39:21"}
+{"current_steps": 3420, "total_steps": 3708, "loss": 0.1673, "lr": 9.196282280661023e-07, "epoch": 2.766990291262136, "percentage": 92.23, "elapsed_time": "2 days, 5:26:01", "remaining_time": "4:29:58"}
+{"current_steps": 3430, "total_steps": 3708, "loss": 0.1652, "lr": 8.574472819702029e-07, "epoch": 2.775080906148867, "percentage": 92.5, "elapsed_time": "2 days, 5:35:24", "remaining_time": "4:20:36"}
+{"current_steps": 3440, "total_steps": 3708, "loss": 0.1636, "lr": 7.974061069507571e-07, "epoch": 2.783171521035599, "percentage": 92.77, "elapsed_time": "2 days, 5:44:45", "remaining_time": "4:11:13"}
+{"current_steps": 3450, "total_steps": 3708, "loss": 0.1634, "lr": 7.395100244984604e-07, "epoch": 2.79126213592233, "percentage": 93.04, "elapsed_time": "2 days, 5:54:07", "remaining_time": "4:01:51"}
+{"current_steps": 3460, "total_steps": 3708, "loss": 0.1661, "lr": 6.837641659829807e-07, "epoch": 2.7993527508090614, "percentage": 93.31, "elapsed_time": "2 days, 6:03:29", "remaining_time": "3:52:28"}
+{"current_steps": 3470, "total_steps": 3708, "loss": 0.1643, "lr": 6.301734721981533e-07, "epoch": 2.807443365695793, "percentage": 93.58, "elapsed_time": "2 days, 6:12:54", "remaining_time": "3:43:06"}
+{"current_steps": 3480, "total_steps": 3708, "loss": 0.1643, "lr": 5.787426929240808e-07, "epoch": 2.8155339805825244, "percentage": 93.85, "elapsed_time": "2 days, 6:22:16", "remaining_time": "3:33:44"}
+{"current_steps": 3490, "total_steps": 3708, "loss": 0.1626, "lr": 5.294763865061558e-07, "epoch": 2.823624595469256, "percentage": 94.12, "elapsed_time": "2 days, 6:31:39", "remaining_time": "3:24:21"}
+{"current_steps": 3500, "total_steps": 3708, "loss": 0.1639, "lr": 4.823789194510514e-07, "epoch": 2.831715210355987, "percentage": 94.39, "elapsed_time": "2 days, 6:41:03", "remaining_time": "3:14:59"}
+{"current_steps": 3510, "total_steps": 3708, "loss": 0.1602, "lr": 4.3745446603971064e-07, "epoch": 2.8398058252427183, "percentage": 94.66, "elapsed_time": "2 days, 6:50:27", "remaining_time": "3:05:36"}
+{"current_steps": 3520, "total_steps": 3708, "loss": 0.1642, "lr": 3.947070079573872e-07, "epoch": 2.8478964401294498, "percentage": 94.93, "elapsed_time": "2 days, 6:59:50", "remaining_time": "2:56:14"}
+{"current_steps": 3530, "total_steps": 3708, "loss": 0.1656, "lr": 3.541403339407279e-07, "epoch": 2.855987055016181, "percentage": 95.2, "elapsed_time": "2 days, 7:09:14", "remaining_time": "2:46:52"}
+{"current_steps": 3540, "total_steps": 3708, "loss": 0.1651, "lr": 3.1575803944199624e-07, "epoch": 2.8640776699029127, "percentage": 95.47, "elapsed_time": "2 days, 7:18:36", "remaining_time": "2:37:29"}
+{"current_steps": 3550, "total_steps": 3708, "loss": 0.1623, "lr": 2.7956352631038906e-07, "epoch": 2.872168284789644, "percentage": 95.74, "elapsed_time": "2 days, 7:27:58", "remaining_time": "2:28:07"}
+{"current_steps": 3560, "total_steps": 3708, "loss": 0.1653, "lr": 2.4556000249054133e-07, "epoch": 2.8802588996763756, "percentage": 96.01, "elapsed_time": "2 days, 7:37:21", "remaining_time": "2:18:44"}
+{"current_steps": 3570, "total_steps": 3708, "loss": 0.1691, "lr": 2.1375048173818412e-07, "epoch": 2.8883495145631066, "percentage": 96.28, "elapsed_time": "2 days, 7:46:45", "remaining_time": "2:09:22"}
+{"current_steps": 3580, "total_steps": 3708, "loss": 0.1606, "lr": 1.8413778335305e-07, "epoch": 2.896440129449838, "percentage": 96.55, "elapsed_time": "2 days, 7:56:08", "remaining_time": "1:59:59"}
+{"current_steps": 3590, "total_steps": 3708, "loss": 0.1607, "lr": 1.567245319290006e-07, "epoch": 2.9045307443365695, "percentage": 96.82, "elapsed_time": "2 days, 8:05:30", "remaining_time": "1:50:37"}
+{"current_steps": 3600, "total_steps": 3708, "loss": 0.1659, "lr": 1.315131571213879e-07, "epoch": 2.912621359223301, "percentage": 97.09, "elapsed_time": "2 days, 8:14:52", "remaining_time": "1:41:14"}
+{"current_steps": 3610, "total_steps": 3708, "loss": 0.163, "lr": 1.0850589343172624e-07, "epoch": 2.9207119741100325, "percentage": 97.36, "elapsed_time": "2 days, 8:24:15", "remaining_time": "1:31:52"}
+{"current_steps": 3620, "total_steps": 3708, "loss": 0.1603, "lr": 8.770478000964532e-08, "epoch": 2.9288025889967635, "percentage": 97.63, "elapsed_time": "2 days, 8:33:37", "remaining_time": "1:22:29"}
+{"current_steps": 3630, "total_steps": 3708, "loss": 0.1639, "lr": 6.911166047215145e-08, "epoch": 2.9368932038834954, "percentage": 97.9, "elapsed_time": "2 days, 8:42:59", "remaining_time": "1:13:07"}
+{"current_steps": 3640, "total_steps": 3708, "loss": 0.163, "lr": 5.272818274023872e-08, "epoch": 2.9449838187702264, "percentage": 98.17, "elapsed_time": "2 days, 8:52:21", "remaining_time": "1:03:44"}
+{"current_steps": 3650, "total_steps": 3708, "loss": 0.1607, "lr": 3.855579889282257e-08, "epoch": 2.953074433656958, "percentage": 98.44, "elapsed_time": "2 days, 9:01:43", "remaining_time": "0:54:22"}
+{"current_steps": 3660, "total_steps": 3708, "loss": 0.1598, "lr": 2.6595765038045507e-08, "epoch": 2.9611650485436893, "percentage": 98.71, "elapsed_time": "2 days, 9:11:05", "remaining_time": "0:44:59"}
+{"current_steps": 3670, "total_steps": 3708, "loss": 0.1656, "lr": 1.6849141201946693e-08, "epoch": 2.969255663430421, "percentage": 98.98, "elapsed_time": "2 days, 9:20:28", "remaining_time": "0:35:37"}
+{"current_steps": 3680, "total_steps": 3708, "loss": 0.1589, "lr": 9.316791234506572e-09, "epoch": 2.9773462783171523, "percentage": 99.24, "elapsed_time": "2 days, 9:29:52", "remaining_time": "0:26:14"}
+{"current_steps": 3690, "total_steps": 3708, "loss": 0.1633, "lr": 3.999382733096968e-09, "epoch": 2.9854368932038833, "percentage": 99.51, "elapsed_time": "2 days, 9:39:14", "remaining_time": "0:16:52"}
+{"current_steps": 3700, "total_steps": 3708, "loss": 0.1623, "lr": 8.973869832895609e-10, "epoch": 2.9935275080906147, "percentage": 99.78, "elapsed_time": "2 days, 9:48:37", "remaining_time": "0:07:29"}
+{"current_steps": 3708, "total_steps": 3708, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "2 days, 9:57:09", "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": 10.671297597284719,
+      "learning_rate": 1.2129380053908356e-06,
+      "loss": 2.601,
+      "step": 10
+    },
+    {
+      "epoch": 0.016181229773462782,
+      "grad_norm": 6.311089864073009,
+      "learning_rate": 2.560646900269542e-06,
+      "loss": 2.3954,
+      "step": 20
+    },
+    {
+      "epoch": 0.024271844660194174,
+      "grad_norm": 5.215362734590839,
+      "learning_rate": 3.908355795148248e-06,
+      "loss": 1.7763,
+      "step": 30
+    },
+    {
+      "epoch": 0.032362459546925564,
+      "grad_norm": 2.0676790587143836,
+      "learning_rate": 5.256064690026954e-06,
+      "loss": 1.2086,
+      "step": 40
+    },
+    {
+      "epoch": 0.040453074433656956,
+      "grad_norm": 1.4206424846718397,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.8436,
+      "step": 50
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 1.5867176039284459,
+      "learning_rate": 7.951482479784367e-06,
+      "loss": 0.6706,
+      "step": 60
+    },
+    {
+      "epoch": 0.05663430420711974,
+      "grad_norm": 1.2138979784248856,
+      "learning_rate": 9.299191374663074e-06,
+      "loss": 0.5858,
+      "step": 70
+    },
+    {
+      "epoch": 0.06472491909385113,
+      "grad_norm": 0.7702442127235006,
+      "learning_rate": 1.0646900269541779e-05,
+      "loss": 0.5384,
+      "step": 80
+    },
+    {
+      "epoch": 0.07281553398058252,
+      "grad_norm": 1.0114209152057974,
+      "learning_rate": 1.1994609164420486e-05,
+      "loss": 0.4977,
+      "step": 90
+    },
+    {
+      "epoch": 0.08090614886731391,
+      "grad_norm": 0.5794843568633474,
+      "learning_rate": 1.3342318059299191e-05,
+      "loss": 0.4639,
+      "step": 100
+    },
+    {
+      "epoch": 0.0889967637540453,
+      "grad_norm": 0.696228528125689,
+      "learning_rate": 1.4690026954177898e-05,
+      "loss": 0.4589,
+      "step": 110
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 0.7033818455552552,
+      "learning_rate": 1.6037735849056604e-05,
+      "loss": 0.4511,
+      "step": 120
+    },
+    {
+      "epoch": 0.10517799352750809,
+      "grad_norm": 0.6918266955210028,
+      "learning_rate": 1.738544474393531e-05,
+      "loss": 0.4259,
+      "step": 130
+    },
+    {
+      "epoch": 0.11326860841423948,
+      "grad_norm": 0.48391302376719975,
+      "learning_rate": 1.8733153638814018e-05,
+      "loss": 0.4271,
+      "step": 140
+    },
+    {
+      "epoch": 0.12135922330097088,
+      "grad_norm": 0.6600203232524687,
+      "learning_rate": 2.0080862533692725e-05,
+      "loss": 0.415,
+      "step": 150
+    },
+    {
+      "epoch": 0.12944983818770225,
+      "grad_norm": 0.6491506270021533,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.4047,
+      "step": 160
+    },
+    {
+      "epoch": 0.13754045307443366,
+      "grad_norm": 1.108358955768956,
+      "learning_rate": 2.2776280323450135e-05,
+      "loss": 0.3977,
+      "step": 170
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 1.0220828997842137,
+      "learning_rate": 2.4123989218328842e-05,
+      "loss": 0.3972,
+      "step": 180
+    },
+    {
+      "epoch": 0.15372168284789645,
+      "grad_norm": 0.7866428488585541,
+      "learning_rate": 2.547169811320755e-05,
+      "loss": 0.3886,
+      "step": 190
+    },
+    {
+      "epoch": 0.16181229773462782,
+      "grad_norm": 0.7956757749575036,
+      "learning_rate": 2.6819407008086256e-05,
+      "loss": 0.3881,
+      "step": 200
+    },
+    {
+      "epoch": 0.16990291262135923,
+      "grad_norm": 0.7154298194551921,
+      "learning_rate": 2.8167115902964963e-05,
+      "loss": 0.3915,
+      "step": 210
+    },
+    {
+      "epoch": 0.1779935275080906,
+      "grad_norm": 0.7172238259413939,
+      "learning_rate": 2.9514824797843667e-05,
+      "loss": 0.384,
+      "step": 220
+    },
+    {
+      "epoch": 0.18608414239482202,
+      "grad_norm": 0.7995655410192202,
+      "learning_rate": 3.086253369272237e-05,
+      "loss": 0.3748,
+      "step": 230
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 13.800265685156136,
+      "learning_rate": 3.221024258760108e-05,
+      "loss": 0.3732,
+      "step": 240
+    },
+    {
+      "epoch": 0.2022653721682848,
+      "grad_norm": 0.9742105484202757,
+      "learning_rate": 3.355795148247979e-05,
+      "loss": 0.3778,
+      "step": 250
+    },
+    {
+      "epoch": 0.21035598705501618,
+      "grad_norm": 0.7304881990540768,
+      "learning_rate": 3.490566037735849e-05,
+      "loss": 0.3779,
+      "step": 260
+    },
+    {
+      "epoch": 0.21844660194174756,
+      "grad_norm": 0.6020503863674406,
+      "learning_rate": 3.62533692722372e-05,
+      "loss": 0.3659,
+      "step": 270
+    },
+    {
+      "epoch": 0.22653721682847897,
+      "grad_norm": 0.6249460279802084,
+      "learning_rate": 3.76010781671159e-05,
+      "loss": 0.3678,
+      "step": 280
+    },
+    {
+      "epoch": 0.23462783171521034,
+      "grad_norm": 0.7546333918045006,
+      "learning_rate": 3.894878706199461e-05,
+      "loss": 0.3608,
+      "step": 290
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 1.0594379228439452,
+      "learning_rate": 4.0296495956873316e-05,
+      "loss": 0.3623,
+      "step": 300
+    },
+    {
+      "epoch": 0.25080906148867316,
+      "grad_norm": 0.8968179435463678,
+      "learning_rate": 4.164420485175202e-05,
+      "loss": 0.361,
+      "step": 310
+    },
+    {
+      "epoch": 0.2588996763754045,
+      "grad_norm": 0.6785334732318974,
+      "learning_rate": 4.299191374663073e-05,
+      "loss": 0.3587,
+      "step": 320
+    },
+    {
+      "epoch": 0.2669902912621359,
+      "grad_norm": 0.7864105607470325,
+      "learning_rate": 4.433962264150944e-05,
+      "loss": 0.3641,
+      "step": 330
+    },
+    {
+      "epoch": 0.2750809061488673,
+      "grad_norm": 0.7520885489880877,
+      "learning_rate": 4.5687331536388144e-05,
+      "loss": 0.3571,
+      "step": 340
+    },
+    {
+      "epoch": 0.28317152103559873,
+      "grad_norm": 0.7154271960483066,
+      "learning_rate": 4.703504043126685e-05,
+      "loss": 0.3564,
+      "step": 350
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.5540540831665184,
+      "learning_rate": 4.838274932614555e-05,
+      "loss": 0.3499,
+      "step": 360
+    },
+    {
+      "epoch": 0.2993527508090615,
+      "grad_norm": 0.8972157863979777,
+      "learning_rate": 4.973045822102426e-05,
+      "loss": 0.3546,
+      "step": 370
+    },
+    {
+      "epoch": 0.3074433656957929,
+      "grad_norm": 0.5977035180668759,
+      "learning_rate": 4.9999290952604396e-05,
+      "loss": 0.358,
+      "step": 380
+    },
+    {
+      "epoch": 0.3155339805825243,
+      "grad_norm": 0.6794119464101197,
+      "learning_rate": 4.9996410516491115e-05,
+      "loss": 0.3442,
+      "step": 390
+    },
+    {
+      "epoch": 0.32362459546925565,
+      "grad_norm": 0.6445475083450791,
+      "learning_rate": 4.9991314631296585e-05,
+      "loss": 0.3504,
+      "step": 400
+    },
+    {
+      "epoch": 0.33171521035598706,
+      "grad_norm": 0.6254469656785541,
+      "learning_rate": 4.9984003748672604e-05,
+      "loss": 0.3451,
+      "step": 410
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 0.6124809359745962,
+      "learning_rate": 4.997447851658774e-05,
+      "loss": 0.3373,
+      "step": 420
+    },
+    {
+      "epoch": 0.3478964401294498,
+      "grad_norm": 0.8314626543045377,
+      "learning_rate": 4.9962739779269887e-05,
+      "loss": 0.3427,
+      "step": 430
+    },
+    {
+      "epoch": 0.3559870550161812,
+      "grad_norm": 0.7096404483846562,
+      "learning_rate": 4.9948788577131414e-05,
+      "loss": 0.3402,
+      "step": 440
+    },
+    {
+      "epoch": 0.3640776699029126,
+      "grad_norm": 0.5010872233529703,
+      "learning_rate": 4.993262614667696e-05,
+      "loss": 0.3404,
+      "step": 450
+    },
+    {
+      "epoch": 0.37216828478964403,
+      "grad_norm": 0.7096873040187438,
+      "learning_rate": 4.9914253920393884e-05,
+      "loss": 0.3374,
+      "step": 460
+    },
+    {
+      "epoch": 0.3802588996763754,
+      "grad_norm": 0.6463889915075135,
+      "learning_rate": 4.9893673526625265e-05,
+      "loss": 0.3374,
+      "step": 470
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.698463838389325,
+      "learning_rate": 4.987088678942555e-05,
+      "loss": 0.338,
+      "step": 480
+    },
+    {
+      "epoch": 0.3964401294498382,
+      "grad_norm": 0.6798285617884782,
+      "learning_rate": 4.984589572839897e-05,
+      "loss": 0.3335,
+      "step": 490
+    },
+    {
+      "epoch": 0.4045307443365696,
+      "grad_norm": 0.4599792513564179,
+      "learning_rate": 4.9818702558520485e-05,
+      "loss": 0.3299,
+      "step": 500
+    },
+    {
+      "epoch": 0.41262135922330095,
+      "grad_norm": 0.49728635683969147,
+      "learning_rate": 4.978930968993946e-05,
+      "loss": 0.3388,
+      "step": 510
+    },
+    {
+      "epoch": 0.42071197411003236,
+      "grad_norm": 0.4217113735156639,
+      "learning_rate": 4.9757719727766085e-05,
+      "loss": 0.3319,
+      "step": 520
+    },
+    {
+      "epoch": 0.42880258899676377,
+      "grad_norm": 0.48369122768285383,
+      "learning_rate": 4.972393547184046e-05,
+      "loss": 0.33,
+      "step": 530
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.43182058716411714,
+      "learning_rate": 4.968795991648446e-05,
+      "loss": 0.3252,
+      "step": 540
+    },
+    {
+      "epoch": 0.4449838187702265,
+      "grad_norm": 0.5876386529955178,
+      "learning_rate": 4.9649796250236344e-05,
+      "loss": 0.324,
+      "step": 550
+    },
+    {
+      "epoch": 0.45307443365695793,
+      "grad_norm": 0.5486057928498816,
+      "learning_rate": 4.960944785556814e-05,
+      "loss": 0.3254,
+      "step": 560
+    },
+    {
+      "epoch": 0.46116504854368934,
+      "grad_norm": 0.4867958049194759,
+      "learning_rate": 4.956691830858585e-05,
+      "loss": 0.3216,
+      "step": 570
+    },
+    {
+      "epoch": 0.4692556634304207,
+      "grad_norm": 0.4571043410958847,
+      "learning_rate": 4.952221137871252e-05,
+      "loss": 0.3206,
+      "step": 580
+    },
+    {
+      "epoch": 0.4773462783171521,
+      "grad_norm": 0.5063667000215466,
+      "learning_rate": 4.947533102835413e-05,
+      "loss": 0.322,
+      "step": 590
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.4522556669690265,
+      "learning_rate": 4.942628141254843e-05,
+      "loss": 0.3257,
+      "step": 600
+    },
+    {
+      "epoch": 0.4935275080906149,
+      "grad_norm": 0.4321047527271199,
+      "learning_rate": 4.937506687859666e-05,
+      "loss": 0.3197,
+      "step": 610
+    },
+    {
+      "epoch": 0.5016181229773463,
+      "grad_norm": 0.393401694292632,
+      "learning_rate": 4.932169196567824e-05,
+      "loss": 0.3209,
+      "step": 620
+    },
+    {
+      "epoch": 0.5097087378640777,
+      "grad_norm": 0.39900307971080384,
+      "learning_rate": 4.9266161404448454e-05,
+      "loss": 0.3102,
+      "step": 630
+    },
+    {
+      "epoch": 0.517799352750809,
+      "grad_norm": 0.4716849647042864,
+      "learning_rate": 4.920848011661919e-05,
+      "loss": 0.3195,
+      "step": 640
+    },
+    {
+      "epoch": 0.5258899676375405,
+      "grad_norm": 0.460323217460504,
+      "learning_rate": 4.914865321452274e-05,
+      "loss": 0.3151,
+      "step": 650
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 0.45925505988077353,
+      "learning_rate": 4.908668600065862e-05,
+      "loss": 0.3151,
+      "step": 660
+    },
+    {
+      "epoch": 0.5420711974110033,
+      "grad_norm": 0.5830461135145422,
+      "learning_rate": 4.90225839672237e-05,
+      "loss": 0.3216,
+      "step": 670
+    },
+    {
+      "epoch": 0.5501618122977346,
+      "grad_norm": 0.4701717654524378,
+      "learning_rate": 4.8956352795625325e-05,
+      "loss": 0.3168,
+      "step": 680
+    },
+    {
+      "epoch": 0.558252427184466,
+      "grad_norm": 0.337481033517792,
+      "learning_rate": 4.8887998355977886e-05,
+      "loss": 0.313,
+      "step": 690
+    },
+    {
+      "epoch": 0.5663430420711975,
+      "grad_norm": 0.6269962483834282,
+      "learning_rate": 4.881752670658244e-05,
+      "loss": 0.3132,
+      "step": 700
+    },
+    {
+      "epoch": 0.5744336569579288,
+      "grad_norm": 0.4662171259348263,
+      "learning_rate": 4.87449440933898e-05,
+      "loss": 0.314,
+      "step": 710
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.5290025725406212,
+      "learning_rate": 4.867025694944698e-05,
+      "loss": 0.3109,
+      "step": 720
+    },
+    {
+      "epoch": 0.5906148867313916,
+      "grad_norm": 0.43779075196037404,
+      "learning_rate": 4.859347189432699e-05,
+      "loss": 0.3188,
+      "step": 730
+    },
+    {
+      "epoch": 0.598705501618123,
+      "grad_norm": 0.5368922851512631,
+      "learning_rate": 4.8514595733542144e-05,
+      "loss": 0.3053,
+      "step": 740
+    },
+    {
+      "epoch": 0.6067961165048543,
+      "grad_norm": 0.3739446436447005,
+      "learning_rate": 4.8433635457940915e-05,
+      "loss": 0.3095,
+      "step": 750
+    },
+    {
+      "epoch": 0.6148867313915858,
+      "grad_norm": 0.36805536772795056,
+      "learning_rate": 4.8350598243088283e-05,
+      "loss": 0.3081,
+      "step": 760
+    },
+    {
+      "epoch": 0.6229773462783171,
+      "grad_norm": 0.6081454922011427,
+      "learning_rate": 4.8265491448629804e-05,
+      "loss": 0.3096,
+      "step": 770
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.606989309832587,
+      "learning_rate": 4.817832261763928e-05,
+      "loss": 0.3064,
+      "step": 780
+    },
+    {
+      "epoch": 0.63915857605178,
+      "grad_norm": 0.374478457061797,
+      "learning_rate": 4.8089099475950257e-05,
+      "loss": 0.3075,
+      "step": 790
+    },
+    {
+      "epoch": 0.6472491909385113,
+      "grad_norm": 0.4495226360651449,
+      "learning_rate": 4.7997829931471225e-05,
+      "loss": 0.3055,
+      "step": 800
+    },
+    {
+      "epoch": 0.6553398058252428,
+      "grad_norm": 0.48869687342847196,
+      "learning_rate": 4.7904522073484786e-05,
+      "loss": 0.3056,
+      "step": 810
+    },
+    {
+      "epoch": 0.6634304207119741,
+      "grad_norm": 0.4244338878096003,
+      "learning_rate": 4.780918417193065e-05,
+      "loss": 0.3068,
+      "step": 820
+    },
+    {
+      "epoch": 0.6715210355987055,
+      "grad_norm": 0.8568600210635637,
+      "learning_rate": 4.7711824676672726e-05,
+      "loss": 0.311,
+      "step": 830
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.4818809073072544,
+      "learning_rate": 4.76124522167501e-05,
+      "loss": 0.3139,
+      "step": 840
+    },
+    {
+      "epoch": 0.6877022653721683,
+      "grad_norm": 0.5673965426592162,
+      "learning_rate": 4.751107559961238e-05,
+      "loss": 0.3085,
+      "step": 850
+    },
+    {
+      "epoch": 0.6957928802588996,
+      "grad_norm": 0.42296772741332755,
+      "learning_rate": 4.740770381033894e-05,
+      "loss": 0.3129,
+      "step": 860
+    },
+    {
+      "epoch": 0.7038834951456311,
+      "grad_norm": 0.4358296754132214,
+      "learning_rate": 4.730234601084268e-05,
+      "loss": 0.3058,
+      "step": 870
+    },
+    {
+      "epoch": 0.7119741100323624,
+      "grad_norm": 0.4373965191621123,
+      "learning_rate": 4.719501153905793e-05,
+      "loss": 0.3025,
+      "step": 880
+    },
+    {
+      "epoch": 0.7200647249190939,
+      "grad_norm": 0.3787352736875979,
+      "learning_rate": 4.7085709908112866e-05,
+      "loss": 0.3034,
+      "step": 890
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.469843837143792,
+      "learning_rate": 4.6974450805486305e-05,
+      "loss": 0.303,
+      "step": 900
+    },
+    {
+      "epoch": 0.7362459546925566,
+      "grad_norm": 0.4394456386334091,
+      "learning_rate": 4.686124409214917e-05,
+      "loss": 0.2975,
+      "step": 910
+    },
+    {
+      "epoch": 0.7443365695792881,
+      "grad_norm": 0.37340718943669743,
+      "learning_rate": 4.674609980169042e-05,
+      "loss": 0.3002,
+      "step": 920
+    },
+    {
+      "epoch": 0.7524271844660194,
+      "grad_norm": 0.47566576341803307,
+      "learning_rate": 4.662902813942784e-05,
+      "loss": 0.3051,
+      "step": 930
+    },
+    {
+      "epoch": 0.7605177993527508,
+      "grad_norm": 0.4410622981522137,
+      "learning_rate": 4.651003948150349e-05,
+      "loss": 0.2962,
+      "step": 940
+    },
+    {
+      "epoch": 0.7686084142394822,
+      "grad_norm": 0.41727057267800893,
+      "learning_rate": 4.638914437396408e-05,
+      "loss": 0.2959,
+      "step": 950
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.3766577031650596,
+      "learning_rate": 4.626635353182626e-05,
+      "loss": 0.2951,
+      "step": 960
+    },
+    {
+      "epoch": 0.7847896440129449,
+      "grad_norm": 0.36286446424405566,
+      "learning_rate": 4.614167783812694e-05,
+      "loss": 0.294,
+      "step": 970
+    },
+    {
+      "epoch": 0.7928802588996764,
+      "grad_norm": 0.4051751512540439,
+      "learning_rate": 4.601512834295874e-05,
+      "loss": 0.2944,
+      "step": 980
+    },
+    {
+      "epoch": 0.8009708737864077,
+      "grad_norm": 0.37027627803204927,
+      "learning_rate": 4.588671626249057e-05,
+      "loss": 0.2952,
+      "step": 990
+    },
+    {
+      "epoch": 0.8090614886731392,
+      "grad_norm": 0.464872048858869,
+      "learning_rate": 4.5756452977973585e-05,
+      "loss": 0.292,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8171521035598706,
+      "grad_norm": 0.4408350560704109,
+      "learning_rate": 4.56243500347324e-05,
+      "loss": 0.3009,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.49398039939609895,
+      "learning_rate": 4.549041914114188e-05,
+      "loss": 0.2969,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.5592891850787267,
+      "learning_rate": 4.535467216758936e-05,
+      "loss": 0.2981,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8414239482200647,
+      "grad_norm": 0.4230495195351826,
+      "learning_rate": 4.5217121145422616e-05,
+      "loss": 0.2883,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8495145631067961,
+      "grad_norm": 0.33133186454457314,
+      "learning_rate": 4.5077778265883477e-05,
+      "loss": 0.2927,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8576051779935275,
+      "grad_norm": 0.3936509630212508,
+      "learning_rate": 4.4936655879027336e-05,
+      "loss": 0.2948,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8656957928802589,
+      "grad_norm": 0.4500916019115437,
+      "learning_rate": 4.479376649262855e-05,
+      "loss": 0.2943,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.3668748277040676,
+      "learning_rate": 4.464912277107185e-05,
+      "loss": 0.2928,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8818770226537217,
+      "grad_norm": 0.320494366403153,
+      "learning_rate": 4.450273753422992e-05,
+      "loss": 0.2909,
+      "step": 1090
+    },
+    {
+      "epoch": 0.889967637540453,
+      "grad_norm": 0.37987923612254676,
+      "learning_rate": 4.435462375632711e-05,
+      "loss": 0.2964,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8980582524271845,
+      "grad_norm": 0.39981221876873035,
+      "learning_rate": 4.420479456478957e-05,
+      "loss": 0.2994,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9061488673139159,
+      "grad_norm": 0.3091126654651453,
+      "learning_rate": 4.405326323908178e-05,
+      "loss": 0.2851,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9142394822006472,
+      "grad_norm": 0.47407567538309503,
+      "learning_rate": 4.390004320952947e-05,
+      "loss": 0.2914,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.4473436088105401,
+      "learning_rate": 4.374514805612942e-05,
+      "loss": 0.2958,
+      "step": 1140
+    },
+    {
+      "epoch": 0.93042071197411,
+      "grad_norm": 0.3335572127939302,
+      "learning_rate": 4.358859150734576e-05,
+      "loss": 0.2915,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9385113268608414,
+      "grad_norm": 0.4410700993823537,
+      "learning_rate": 4.343038743889324e-05,
+      "loss": 0.2936,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9466019417475728,
+      "grad_norm": 0.4566532222343424,
+      "learning_rate": 4.3270549872507415e-05,
+      "loss": 0.2921,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9546925566343042,
+      "grad_norm": 0.34831612960636166,
+      "learning_rate": 4.3109092974701895e-05,
+      "loss": 0.2961,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9627831715210357,
+      "grad_norm": 0.33970324826345877,
+      "learning_rate": 4.2946031055512733e-05,
+      "loss": 0.2887,
+      "step": 1190
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.46192580192198573,
+      "learning_rate": 4.2781378567230145e-05,
+      "loss": 0.29,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9789644012944984,
+      "grad_norm": 0.4615119904861507,
+      "learning_rate": 4.2615150103117576e-05,
+      "loss": 0.2886,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9870550161812298,
+      "grad_norm": 0.2981731488057052,
+      "learning_rate": 4.24473603961183e-05,
+      "loss": 0.2908,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9951456310679612,
+      "grad_norm": 0.3632890670918485,
+      "learning_rate": 4.227802431754961e-05,
+      "loss": 0.2924,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0032362459546926,
+      "grad_norm": 0.4150403194043393,
+      "learning_rate": 4.210715687578481e-05,
+      "loss": 0.2727,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0113268608414239,
+      "grad_norm": 0.34869821901507675,
+      "learning_rate": 4.193477321492293e-05,
+      "loss": 0.2453,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.3601010037088075,
+      "learning_rate": 4.176088861344657e-05,
+      "loss": 0.2414,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0275080906148868,
+      "grad_norm": 0.32325115826296424,
+      "learning_rate": 4.158551848286773e-05,
+      "loss": 0.2425,
+      "step": 1270
+    },
+    {
+      "epoch": 1.035598705501618,
+      "grad_norm": 0.38691862676257593,
+      "learning_rate": 4.140867836636189e-05,
+      "loss": 0.247,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0436893203883495,
+      "grad_norm": 0.34882276619380004,
+      "learning_rate": 4.1230383937390374e-05,
+      "loss": 0.239,
+      "step": 1290
+    },
+    {
+      "epoch": 1.051779935275081,
+      "grad_norm": 0.3951880114595209,
+      "learning_rate": 4.1050650998311215e-05,
+      "loss": 0.2375,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0598705501618122,
+      "grad_norm": 0.3736593550545292,
+      "learning_rate": 4.086949547897862e-05,
+      "loss": 0.2479,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.34774896387912796,
+      "learning_rate": 4.068693343533103e-05,
+      "loss": 0.2372,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0760517799352751,
+      "grad_norm": 0.3330973525666551,
+      "learning_rate": 4.050298104796812e-05,
+      "loss": 0.2413,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0841423948220066,
+      "grad_norm": 0.27924340181820345,
+      "learning_rate": 4.0317654620716704e-05,
+      "loss": 0.2454,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0922330097087378,
+      "grad_norm": 0.3807786754448861,
+      "learning_rate": 4.013097057918566e-05,
+      "loss": 0.2457,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1003236245954693,
+      "grad_norm": 0.34402611290470836,
+      "learning_rate": 3.9942945469310175e-05,
+      "loss": 0.2406,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1084142394822007,
+      "grad_norm": 0.3087094208360471,
+      "learning_rate": 3.9753595955885264e-05,
+      "loss": 0.2403,
+      "step": 1370
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.334687554837771,
+      "learning_rate": 3.9562938821088705e-05,
+      "loss": 0.2443,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1245954692556634,
+      "grad_norm": 0.34101610586326964,
+      "learning_rate": 3.9370990962993695e-05,
+      "loss": 0.2455,
+      "step": 1390
+    },
+    {
+      "epoch": 1.132686084142395,
+      "grad_norm": 0.36304196735659044,
+      "learning_rate": 3.9177769394071086e-05,
+      "loss": 0.2423,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1407766990291262,
+      "grad_norm": 0.32455755430896477,
+      "learning_rate": 3.898329123968163e-05,
+      "loss": 0.2424,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1488673139158576,
+      "grad_norm": 0.29393617658060145,
+      "learning_rate": 3.87875737365581e-05,
+      "loss": 0.2383,
+      "step": 1420
+    },
+    {
+      "epoch": 1.156957928802589,
+      "grad_norm": 0.40951987232018516,
+      "learning_rate": 3.8590634231277626e-05,
+      "loss": 0.2605,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.3441977909919604,
+      "learning_rate": 3.8392490178724184e-05,
+      "loss": 0.2455,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1731391585760518,
+      "grad_norm": 0.3112960711261136,
+      "learning_rate": 3.8193159140541645e-05,
+      "loss": 0.2419,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1812297734627832,
+      "grad_norm": 0.3032973624770194,
+      "learning_rate": 3.7992658783577215e-05,
+      "loss": 0.2422,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1893203883495145,
+      "grad_norm": 0.29109047481785005,
+      "learning_rate": 3.779100687831563e-05,
+      "loss": 0.2462,
+      "step": 1470
+    },
+    {
+      "epoch": 1.197411003236246,
+      "grad_norm": 0.2948022665787199,
+      "learning_rate": 3.758822129730415e-05,
+      "loss": 0.245,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2055016181229774,
+      "grad_norm": 0.31258495126302516,
+      "learning_rate": 3.738432001356851e-05,
+      "loss": 0.2455,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.27328969134267245,
+      "learning_rate": 3.7179321099019916e-05,
+      "loss": 0.2369,
+      "step": 1500
+    },
+    {
+      "epoch": 1.22168284789644,
+      "grad_norm": 0.2847184801501813,
+      "learning_rate": 3.6973242722853365e-05,
+      "loss": 0.2402,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2297734627831716,
+      "grad_norm": 0.29812191017952583,
+      "learning_rate": 3.6766103149937295e-05,
+      "loss": 0.2427,
+      "step": 1520
+    },
+    {
+      "epoch": 1.237864077669903,
+      "grad_norm": 0.3017615376479855,
+      "learning_rate": 3.655792073919471e-05,
+      "loss": 0.2363,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2459546925566343,
+      "grad_norm": 0.27567364203177047,
+      "learning_rate": 3.634871394197607e-05,
+      "loss": 0.2388,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2540453074433657,
+      "grad_norm": 0.2685513406212581,
+      "learning_rate": 3.6138501300423934e-05,
+      "loss": 0.2378,
+      "step": 1550
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.25546275734682755,
+      "learning_rate": 3.592730144582948e-05,
+      "loss": 0.2341,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2702265372168284,
+      "grad_norm": 0.2841532225060496,
+      "learning_rate": 3.571513309698131e-05,
+      "loss": 0.2366,
+      "step": 1570
+    },
+    {
+      "epoch": 1.27831715210356,
+      "grad_norm": 0.30541995269870376,
+      "learning_rate": 3.5502015058506335e-05,
+      "loss": 0.2375,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2864077669902914,
+      "grad_norm": 0.34844869264740597,
+      "learning_rate": 3.528796621920307e-05,
+      "loss": 0.239,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2944983818770226,
+      "grad_norm": 0.30708349077357805,
+      "learning_rate": 3.50730055503676e-05,
+      "loss": 0.2356,
+      "step": 1600
+    },
+    {
+      "epoch": 1.302588996763754,
+      "grad_norm": 0.3175696506236882,
+      "learning_rate": 3.485715210411204e-05,
+      "loss": 0.2358,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.29197347345983926,
+      "learning_rate": 3.4640425011676034e-05,
+      "loss": 0.2408,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3187702265372168,
+      "grad_norm": 0.28169129894596506,
+      "learning_rate": 3.442284348173106e-05,
+      "loss": 0.2395,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3268608414239482,
+      "grad_norm": 0.3992186155056125,
+      "learning_rate": 3.420442679867796e-05,
+      "loss": 0.2391,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3349514563106797,
+      "grad_norm": 0.35636813012626095,
+      "learning_rate": 3.398519432093782e-05,
+      "loss": 0.2374,
+      "step": 1650
+    },
+    {
+      "epoch": 1.343042071197411,
+      "grad_norm": 0.27742962322683806,
+      "learning_rate": 3.376516547923614e-05,
+      "loss": 0.2336,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3511326860841424,
+      "grad_norm": 0.28660763778902115,
+      "learning_rate": 3.3544359774880714e-05,
+      "loss": 0.2371,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.40153445123463827,
+      "learning_rate": 3.3322796778033204e-05,
+      "loss": 0.2376,
+      "step": 1680
+    },
+    {
+      "epoch": 1.367313915857605,
+      "grad_norm": 0.3035749815074104,
+      "learning_rate": 3.3100496125974624e-05,
+      "loss": 0.2377,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3754045307443366,
+      "grad_norm": 0.342174636479568,
+      "learning_rate": 3.2877477521364895e-05,
+      "loss": 0.2347,
+      "step": 1700
+    },
+    {
+      "epoch": 1.383495145631068,
+      "grad_norm": 0.3199103842497454,
+      "learning_rate": 3.2653760730496555e-05,
+      "loss": 0.2297,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3915857605177995,
+      "grad_norm": 0.27362978201436644,
+      "learning_rate": 3.242936558154285e-05,
+      "loss": 0.2356,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3996763754045307,
+      "grad_norm": 0.27057113114600895,
+      "learning_rate": 3.2204311962800426e-05,
+      "loss": 0.2322,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.26643871030324334,
+      "learning_rate": 3.197861982092651e-05,
+      "loss": 0.2384,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4158576051779934,
+      "grad_norm": 0.31497718220761595,
+      "learning_rate": 3.175230915917108e-05,
+      "loss": 0.2427,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4239482200647249,
+      "grad_norm": 0.3267383691103009,
+      "learning_rate": 3.152540003560398e-05,
+      "loss": 0.2417,
+      "step": 1760
+    },
+    {
+      "epoch": 1.4320388349514563,
+      "grad_norm": 0.29663197558136717,
+      "learning_rate": 3.129791256133712e-05,
+      "loss": 0.2288,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4401294498381878,
+      "grad_norm": 0.32445565003875587,
+      "learning_rate": 3.106986689874204e-05,
+      "loss": 0.2347,
+      "step": 1780
+    },
+    {
+      "epoch": 1.448220064724919,
+      "grad_norm": 0.29535887592241783,
+      "learning_rate": 3.0841283259662875e-05,
+      "loss": 0.2385,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.2733621998571008,
+      "learning_rate": 3.0612181903625014e-05,
+      "loss": 0.2359,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4644012944983817,
+      "grad_norm": 0.2698891118245767,
+      "learning_rate": 3.0382583136039444e-05,
+      "loss": 0.2339,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4724919093851132,
+      "grad_norm": 0.30735748022487547,
+      "learning_rate": 3.015250730640308e-05,
+      "loss": 0.237,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4805825242718447,
+      "grad_norm": 0.35024703112347744,
+      "learning_rate": 2.9921974806495178e-05,
+      "loss": 0.2301,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4886731391585761,
+      "grad_norm": 0.27111655738392937,
+      "learning_rate": 2.969100606856998e-05,
+      "loss": 0.2339,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4967637540453074,
+      "grad_norm": 0.2787522117740128,
+      "learning_rate": 2.9459621563545825e-05,
+      "loss": 0.2385,
+      "step": 1850
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.2969133211792456,
+      "learning_rate": 2.9227841799190775e-05,
+      "loss": 0.2305,
+      "step": 1860
+    },
+    {
+      "epoch": 1.51294498381877,
+      "grad_norm": 0.2613286640167969,
+      "learning_rate": 2.8995687318304975e-05,
+      "loss": 0.2328,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5210355987055015,
+      "grad_norm": 0.27216163489187184,
+      "learning_rate": 2.8763178696899995e-05,
+      "loss": 0.2373,
+      "step": 1880
+    },
+    {
+      "epoch": 1.529126213592233,
+      "grad_norm": 0.2964461145886047,
+      "learning_rate": 2.853033654237507e-05,
+      "loss": 0.2289,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5372168284789645,
+      "grad_norm": 0.32949202305010017,
+      "learning_rate": 2.8297181491690756e-05,
+      "loss": 0.23,
+      "step": 1900
+    },
+    {
+      "epoch": 1.545307443365696,
+      "grad_norm": 0.3071880700004742,
+      "learning_rate": 2.8063734209539773e-05,
+      "loss": 0.2367,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 0.2529468819780906,
+      "learning_rate": 2.783001538651554e-05,
+      "loss": 0.2292,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5614886731391586,
+      "grad_norm": 0.2773840151703638,
+      "learning_rate": 2.7596045737278336e-05,
+      "loss": 0.2398,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5695792880258899,
+      "grad_norm": 0.295185746563609,
+      "learning_rate": 2.7361845998719315e-05,
+      "loss": 0.2325,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5776699029126213,
+      "grad_norm": 0.25593073129115584,
+      "learning_rate": 2.7127436928122612e-05,
+      "loss": 0.2367,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5857605177993528,
+      "grad_norm": 0.3516341035154705,
+      "learning_rate": 2.6892839301325623e-05,
+      "loss": 0.239,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5938511326860842,
+      "grad_norm": 0.27153437539635045,
+      "learning_rate": 2.6658073910877603e-05,
+      "loss": 0.2288,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.26490849237503283,
+      "learning_rate": 2.6423161564196803e-05,
+      "loss": 0.231,
+      "step": 1980
+    },
+    {
+      "epoch": 1.610032362459547,
+      "grad_norm": 0.25415769405974487,
+      "learning_rate": 2.6188123081726306e-05,
+      "loss": 0.2341,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6181229773462782,
+      "grad_norm": 0.2566774871023634,
+      "learning_rate": 2.5952979295088714e-05,
+      "loss": 0.2303,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6262135922330097,
+      "grad_norm": 0.28720215998836357,
+      "learning_rate": 2.57177510452398e-05,
+      "loss": 0.2297,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6343042071197411,
+      "grad_norm": 0.26535154825569424,
+      "learning_rate": 2.5482459180621377e-05,
+      "loss": 0.2336,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6423948220064726,
+      "grad_norm": 0.2348030131858575,
+      "learning_rate": 2.524712455531347e-05,
+      "loss": 0.2283,
+      "step": 2030
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.2981340903697484,
+      "learning_rate": 2.501176802718599e-05,
+      "loss": 0.2367,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6585760517799353,
+      "grad_norm": 0.2508609786813777,
+      "learning_rate": 2.4776410456050165e-05,
+      "loss": 0.232,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.2760175375931361,
+      "learning_rate": 2.4541072701809624e-05,
+      "loss": 0.2348,
+      "step": 2060
+    },
+    {
+      "epoch": 1.674757281553398,
+      "grad_norm": 0.2681609470474934,
+      "learning_rate": 2.4305775622611627e-05,
+      "loss": 0.2285,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6828478964401294,
+      "grad_norm": 0.3323878551187027,
+      "learning_rate": 2.4070540072998372e-05,
+      "loss": 0.2272,
+      "step": 2080
+    },
+    {
+      "epoch": 1.690938511326861,
+      "grad_norm": 0.24794381758629244,
+      "learning_rate": 2.3835386902058637e-05,
+      "loss": 0.2292,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.26234035679605566,
+      "learning_rate": 2.360033695157995e-05,
+      "loss": 0.2337,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7071197411003236,
+      "grad_norm": 0.3095409200594481,
+      "learning_rate": 2.3365411054201315e-05,
+      "loss": 0.2265,
+      "step": 2110
+    },
+    {
+      "epoch": 1.715210355987055,
+      "grad_norm": 0.24814239367844307,
+      "learning_rate": 2.3130630031566818e-05,
+      "loss": 0.2269,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7233009708737863,
+      "grad_norm": 0.2503027719931105,
+      "learning_rate": 2.2896014692480226e-05,
+      "loss": 0.231,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7313915857605178,
+      "grad_norm": 0.24658598678553392,
+      "learning_rate": 2.266158583106063e-05,
+      "loss": 0.228,
+      "step": 2140
+    },
+    {
+      "epoch": 1.7394822006472492,
+      "grad_norm": 0.23999491418212948,
+      "learning_rate": 2.2427364224899502e-05,
+      "loss": 0.2289,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.24294562795428556,
+      "learning_rate": 2.2193370633219115e-05,
+      "loss": 0.2295,
+      "step": 2160
+    },
+    {
+      "epoch": 1.755663430420712,
+      "grad_norm": 0.27195306057038027,
+      "learning_rate": 2.1959625795032664e-05,
+      "loss": 0.2312,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7637540453074434,
+      "grad_norm": 0.31612469238407165,
+      "learning_rate": 2.1726150427306182e-05,
+      "loss": 0.2286,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7718446601941746,
+      "grad_norm": 0.29752263873224494,
+      "learning_rate": 2.1492965223122305e-05,
+      "loss": 0.2267,
+      "step": 2190
+    },
+    {
+      "epoch": 1.779935275080906,
+      "grad_norm": 0.2603667246009949,
+      "learning_rate": 2.126009084984629e-05,
+      "loss": 0.2251,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7880258899676376,
+      "grad_norm": 0.26289071039247,
+      "learning_rate": 2.102754794729426e-05,
+      "loss": 0.2246,
+      "step": 2210
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.263903471162293,
+      "learning_rate": 2.079535712590382e-05,
+      "loss": 0.2282,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8042071197411005,
+      "grad_norm": 0.27035700806263924,
+      "learning_rate": 2.056353896490742e-05,
+      "loss": 0.2231,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8122977346278317,
+      "grad_norm": 0.2654192689718002,
+      "learning_rate": 2.0332114010508334e-05,
+      "loss": 0.2268,
+      "step": 2240
+    },
+    {
+      "epoch": 1.820388349514563,
+      "grad_norm": 0.23959916600696418,
+      "learning_rate": 2.010110277405966e-05,
+      "loss": 0.2274,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8284789644012944,
+      "grad_norm": 0.24872872826123787,
+      "learning_rate": 1.9870525730246424e-05,
+      "loss": 0.2293,
+      "step": 2260
+    },
+    {
+      "epoch": 1.8365695792880259,
+      "grad_norm": 0.2645954361424342,
+      "learning_rate": 1.9640403315270824e-05,
+      "loss": 0.2286,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.24420673184598343,
+      "learning_rate": 1.9410755925041006e-05,
+      "loss": 0.2257,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8527508090614888,
+      "grad_norm": 0.24705849748195569,
+      "learning_rate": 1.918160391336335e-05,
+      "loss": 0.2259,
+      "step": 2290
+    },
+    {
+      "epoch": 1.86084142394822,
+      "grad_norm": 0.2951271763878442,
+      "learning_rate": 1.8952967590138472e-05,
+      "loss": 0.2248,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8689320388349513,
+      "grad_norm": 0.2520418480110052,
+      "learning_rate": 1.8724867219561203e-05,
+      "loss": 0.2287,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8770226537216828,
+      "grad_norm": 0.294901863474176,
+      "learning_rate": 1.8497323018324476e-05,
+      "loss": 0.2252,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8851132686084142,
+      "grad_norm": 0.2620675111989084,
+      "learning_rate": 1.8270355153827598e-05,
+      "loss": 0.2243,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.21729342901905718,
+      "learning_rate": 1.804398374238872e-05,
+      "loss": 0.2231,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9012944983818771,
+      "grad_norm": 0.26438351131145754,
+      "learning_rate": 1.781822884746196e-05,
+      "loss": 0.2203,
+      "step": 2350
+    },
+    {
+      "epoch": 1.9093851132686084,
+      "grad_norm": 0.22358504292564055,
+      "learning_rate": 1.7593110477859153e-05,
+      "loss": 0.223,
+      "step": 2360
+    },
+    {
+      "epoch": 1.9174757281553398,
+      "grad_norm": 0.24990353950246044,
+      "learning_rate": 1.736864858597645e-05,
+      "loss": 0.2233,
+      "step": 2370
+    },
+    {
+      "epoch": 1.925566343042071,
+      "grad_norm": 0.24537211238465853,
+      "learning_rate": 1.7144863066025955e-05,
+      "loss": 0.2193,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9336569579288025,
+      "grad_norm": 0.2394304662544921,
+      "learning_rate": 1.692177375227242e-05,
+      "loss": 0.2251,
+      "step": 2390
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.2785110439501238,
+      "learning_rate": 1.669940041727538e-05,
+      "loss": 0.218,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9498381877022655,
+      "grad_norm": 0.23030205395765063,
+      "learning_rate": 1.6477762770136707e-05,
+      "loss": 0.225,
+      "step": 2410
+    },
+    {
+      "epoch": 1.9579288025889967,
+      "grad_norm": 0.24458486846119534,
+      "learning_rate": 1.625688045475371e-05,
+      "loss": 0.2208,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9660194174757282,
+      "grad_norm": 0.223552780447482,
+      "learning_rate": 1.603677304807815e-05,
+      "loss": 0.2223,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9741100323624594,
+      "grad_norm": 0.23262918496301524,
+      "learning_rate": 1.5817460058381088e-05,
+      "loss": 0.2274,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9822006472491909,
+      "grad_norm": 0.2210251232882193,
+      "learning_rate": 1.5598960923523842e-05,
+      "loss": 0.2248,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.2543751678692885,
+      "learning_rate": 1.5381295009235262e-05,
+      "loss": 0.2277,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9983818770226538,
+      "grad_norm": 0.2767394174239906,
+      "learning_rate": 1.5164481607395238e-05,
+      "loss": 0.2243,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0064724919093853,
+      "grad_norm": 0.2831027468460383,
+      "learning_rate": 1.4948539934324923e-05,
+      "loss": 0.1818,
+      "step": 2480
+    },
+    {
+      "epoch": 2.0145631067961167,
+      "grad_norm": 0.23514037047255174,
+      "learning_rate": 1.4733489129083534e-05,
+      "loss": 0.1718,
+      "step": 2490
+    },
+    {
+      "epoch": 2.0226537216828477,
+      "grad_norm": 0.23543211368476746,
+      "learning_rate": 1.4519348251772058e-05,
+      "loss": 0.1696,
+      "step": 2500
+    },
+    {
+      "epoch": 2.030744336569579,
+      "grad_norm": 0.24256191701199467,
+      "learning_rate": 1.4306136281843962e-05,
+      "loss": 0.1691,
+      "step": 2510
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.22234069623742508,
+      "learning_rate": 1.4093872116422979e-05,
+      "loss": 0.169,
+      "step": 2520
+    },
+    {
+      "epoch": 2.046925566343042,
+      "grad_norm": 0.23132835680535235,
+      "learning_rate": 1.3882574568628315e-05,
+      "loss": 0.168,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0550161812297736,
+      "grad_norm": 0.23134456189741553,
+      "learning_rate": 1.3672262365907163e-05,
+      "loss": 0.1684,
+      "step": 2540
+    },
+    {
+      "epoch": 2.063106796116505,
+      "grad_norm": 0.22760921973646828,
+      "learning_rate": 1.3462954148374899e-05,
+      "loss": 0.1661,
+      "step": 2550
+    },
+    {
+      "epoch": 2.071197411003236,
+      "grad_norm": 0.21228629003960797,
+      "learning_rate": 1.3254668467163029e-05,
+      "loss": 0.1696,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0792880258899675,
+      "grad_norm": 0.2182291343034792,
+      "learning_rate": 1.3047423782774937e-05,
+      "loss": 0.1649,
+      "step": 2570
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.21718735716811796,
+      "learning_rate": 1.2841238463449743e-05,
+      "loss": 0.1719,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0954692556634305,
+      "grad_norm": 0.22472401035201942,
+      "learning_rate": 1.2636130783534319e-05,
+      "loss": 0.1703,
+      "step": 2590
+    },
+    {
+      "epoch": 2.103559870550162,
+      "grad_norm": 0.21069125465698005,
+      "learning_rate": 1.2432118921863604e-05,
+      "loss": 0.168,
+      "step": 2600
+    },
+    {
+      "epoch": 2.1116504854368934,
+      "grad_norm": 0.21446866638864276,
+      "learning_rate": 1.2229220960149431e-05,
+      "loss": 0.1695,
+      "step": 2610
+    },
+    {
+      "epoch": 2.1197411003236244,
+      "grad_norm": 0.21738468937149366,
+      "learning_rate": 1.2027454881377889e-05,
+      "loss": 0.1675,
+      "step": 2620
+    },
+    {
+      "epoch": 2.127831715210356,
+      "grad_norm": 0.22154845083846222,
+      "learning_rate": 1.1826838568215526e-05,
+      "loss": 0.1685,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.20713221800601425,
+      "learning_rate": 1.1627389801424351e-05,
+      "loss": 0.1706,
+      "step": 2640
+    },
+    {
+      "epoch": 2.144012944983819,
+      "grad_norm": 0.20802282917927592,
+      "learning_rate": 1.1429126258285946e-05,
+      "loss": 0.1661,
+      "step": 2650
+    },
+    {
+      "epoch": 2.1521035598705502,
+      "grad_norm": 0.20646189593102915,
+      "learning_rate": 1.1232065511034696e-05,
+      "loss": 0.1663,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1601941747572817,
+      "grad_norm": 0.21341477579858295,
+      "learning_rate": 1.1036225025300357e-05,
+      "loss": 0.1687,
+      "step": 2670
+    },
+    {
+      "epoch": 2.168284789644013,
+      "grad_norm": 0.21356709357188863,
+      "learning_rate": 1.0841622158560085e-05,
+      "loss": 0.1668,
+      "step": 2680
+    },
+    {
+      "epoch": 2.176375404530744,
+      "grad_norm": 0.20685337347705235,
+      "learning_rate": 1.0648274158599994e-05,
+      "loss": 0.1698,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.22156407557595678,
+      "learning_rate": 1.0456198161986489e-05,
+      "loss": 0.1753,
+      "step": 2700
+    },
+    {
+      "epoch": 2.192556634304207,
+      "grad_norm": 0.2035462284179136,
+      "learning_rate": 1.0265411192547462e-05,
+      "loss": 0.1718,
+      "step": 2710
+    },
+    {
+      "epoch": 2.2006472491909386,
+      "grad_norm": 0.20189924705262216,
+      "learning_rate": 1.0075930159863416e-05,
+      "loss": 0.1661,
+      "step": 2720
+    },
+    {
+      "epoch": 2.20873786407767,
+      "grad_norm": 0.20956549431298424,
+      "learning_rate": 9.887771857768796e-06,
+      "loss": 0.1683,
+      "step": 2730
+    },
+    {
+      "epoch": 2.2168284789644015,
+      "grad_norm": 0.2140205114087363,
+      "learning_rate": 9.700952962863513e-06,
+      "loss": 0.1701,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2249190938511325,
+      "grad_norm": 0.20398172720239158,
+      "learning_rate": 9.515490033034893e-06,
+      "loss": 0.1667,
+      "step": 2750
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.20165594228560696,
+      "learning_rate": 9.331399505990168e-06,
+      "loss": 0.1655,
+      "step": 2760
+    },
+    {
+      "epoch": 2.2411003236245954,
+      "grad_norm": 0.20182970094436348,
+      "learning_rate": 9.148697697799533e-06,
+      "loss": 0.1694,
+      "step": 2770
+    },
+    {
+      "epoch": 2.249190938511327,
+      "grad_norm": 0.2291167648037075,
+      "learning_rate": 8.967400801450105e-06,
+      "loss": 0.1685,
+      "step": 2780
+    },
+    {
+      "epoch": 2.2572815533980584,
+      "grad_norm": 0.20375204743694134,
+      "learning_rate": 8.787524885410678e-06,
+      "loss": 0.1638,
+      "step": 2790
+    },
+    {
+      "epoch": 2.26537216828479,
+      "grad_norm": 0.214007858233087,
+      "learning_rate": 8.60908589220758e-06,
+      "loss": 0.1676,
+      "step": 2800
+    },
+    {
+      "epoch": 2.273462783171521,
+      "grad_norm": 0.2025950468246571,
+      "learning_rate": 8.432099637011693e-06,
+      "loss": 0.1657,
+      "step": 2810
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.21480260045098348,
+      "learning_rate": 8.256581806236704e-06,
+      "loss": 0.1649,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2896440129449838,
+      "grad_norm": 0.2052262400110326,
+      "learning_rate": 8.082547956148873e-06,
+      "loss": 0.1663,
+      "step": 2830
+    },
+    {
+      "epoch": 2.2977346278317152,
+      "grad_norm": 0.20019987270861134,
+      "learning_rate": 7.91001351148819e-06,
+      "loss": 0.1653,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3058252427184467,
+      "grad_norm": 0.207551520801799,
+      "learning_rate": 7.738993764101324e-06,
+      "loss": 0.1677,
+      "step": 2850
+    },
+    {
+      "epoch": 2.313915857605178,
+      "grad_norm": 0.19280865204847908,
+      "learning_rate": 7.569503871586292e-06,
+      "loss": 0.1674,
+      "step": 2860
+    },
+    {
+      "epoch": 2.3220064724919096,
+      "grad_norm": 0.20122645680053192,
+      "learning_rate": 7.401558855949004e-06,
+      "loss": 0.1678,
+      "step": 2870
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.1945489837836613,
+      "learning_rate": 7.235173602271875e-06,
+      "loss": 0.162,
+      "step": 2880
+    },
+    {
+      "epoch": 2.338187702265372,
+      "grad_norm": 0.2051511298220981,
+      "learning_rate": 7.070362857394538e-06,
+      "loss": 0.1712,
+      "step": 2890
+    },
+    {
+      "epoch": 2.3462783171521036,
+      "grad_norm": 0.2038615765888219,
+      "learning_rate": 6.907141228606831e-06,
+      "loss": 0.1664,
+      "step": 2900
+    },
+    {
+      "epoch": 2.354368932038835,
+      "grad_norm": 0.19824356024446557,
+      "learning_rate": 6.745523182354147e-06,
+      "loss": 0.1634,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3624595469255665,
+      "grad_norm": 0.1897897800458373,
+      "learning_rate": 6.585523042955233e-06,
+      "loss": 0.1633,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3705501618122975,
+      "grad_norm": 0.19361276633294466,
+      "learning_rate": 6.427154991332665e-06,
+      "loss": 0.1665,
+      "step": 2930
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.19269698264895235,
+      "learning_rate": 6.2704330637559315e-06,
+      "loss": 0.1632,
+      "step": 2940
+    },
+    {
+      "epoch": 2.3867313915857604,
+      "grad_norm": 0.19487091156668288,
+      "learning_rate": 6.115371150597413e-06,
+      "loss": 0.1657,
+      "step": 2950
+    },
+    {
+      "epoch": 2.394822006472492,
+      "grad_norm": 0.17726783434092147,
+      "learning_rate": 5.961982995101301e-06,
+      "loss": 0.1652,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4029126213592233,
+      "grad_norm": 0.19174116985719267,
+      "learning_rate": 5.810282192165442e-06,
+      "loss": 0.1674,
+      "step": 2970
+    },
+    {
+      "epoch": 2.411003236245955,
+      "grad_norm": 0.20196362222659836,
+      "learning_rate": 5.660282187136507e-06,
+      "loss": 0.1648,
+      "step": 2980
+    },
+    {
+      "epoch": 2.4190938511326863,
+      "grad_norm": 0.2037293127188244,
+      "learning_rate": 5.511996274618253e-06,
+      "loss": 0.1673,
+      "step": 2990
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.2075997573199353,
+      "learning_rate": 5.365437597293238e-06,
+      "loss": 0.1681,
+      "step": 3000
+    },
+    {
+      "epoch": 2.4352750809061487,
+      "grad_norm": 0.19485346480204271,
+      "learning_rate": 5.220619144757996e-06,
+      "loss": 0.168,
+      "step": 3010
+    },
+    {
+      "epoch": 2.44336569579288,
+      "grad_norm": 0.20032468931249656,
+      "learning_rate": 5.077553752371708e-06,
+      "loss": 0.1695,
+      "step": 3020
+    },
+    {
+      "epoch": 2.4514563106796117,
+      "grad_norm": 0.18780814063723766,
+      "learning_rate": 4.936254100118656e-06,
+      "loss": 0.1626,
+      "step": 3030
+    },
+    {
+      "epoch": 2.459546925566343,
+      "grad_norm": 0.20597954746449326,
+      "learning_rate": 4.796732711484342e-06,
+      "loss": 0.1688,
+      "step": 3040
+    },
+    {
+      "epoch": 2.4676375404530746,
+      "grad_norm": 0.1884986064298571,
+      "learning_rate": 4.659001952345538e-06,
+      "loss": 0.1618,
+      "step": 3050
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.19330264761303187,
+      "learning_rate": 4.523074029874291e-06,
+      "loss": 0.1611,
+      "step": 3060
+    },
+    {
+      "epoch": 2.483818770226537,
+      "grad_norm": 0.1899995190829889,
+      "learning_rate": 4.388960991455998e-06,
+      "loss": 0.1642,
+      "step": 3070
+    },
+    {
+      "epoch": 2.4919093851132685,
+      "grad_norm": 0.20352479818055758,
+      "learning_rate": 4.256674723621621e-06,
+      "loss": 0.165,
+      "step": 3080
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.19705480301688083,
+      "learning_rate": 4.126226950994211e-06,
+      "loss": 0.1633,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5080906148867315,
+      "grad_norm": 0.19973033430929824,
+      "learning_rate": 3.997629235249692e-06,
+      "loss": 0.1645,
+      "step": 3100
+    },
+    {
+      "epoch": 2.516181229773463,
+      "grad_norm": 0.19227297342228458,
+      "learning_rate": 3.870892974092197e-06,
+      "loss": 0.1657,
+      "step": 3110
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.19226389494086693,
+      "learning_rate": 3.7460294002438444e-06,
+      "loss": 0.1649,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5323624595469254,
+      "grad_norm": 0.23947157434034713,
+      "learning_rate": 3.6230495804491864e-06,
+      "loss": 0.1638,
+      "step": 3130
+    },
+    {
+      "epoch": 2.540453074433657,
+      "grad_norm": 0.18606261818175104,
+      "learning_rate": 3.5019644144943576e-06,
+      "loss": 0.1657,
+      "step": 3140
+    },
+    {
+      "epoch": 2.5485436893203883,
+      "grad_norm": 0.18558275603040017,
+      "learning_rate": 3.382784634241015e-06,
+      "loss": 0.1626,
+      "step": 3150
+    },
+    {
+      "epoch": 2.55663430420712,
+      "grad_norm": 0.1838874793051254,
+      "learning_rate": 3.2655208026751816e-06,
+      "loss": 0.1645,
+      "step": 3160
+    },
+    {
+      "epoch": 2.5647249190938513,
+      "grad_norm": 0.18161462708162493,
+      "learning_rate": 3.150183312971014e-06,
+      "loss": 0.1625,
+      "step": 3170
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.18098413098000662,
+      "learning_rate": 3.036782387569659e-06,
+      "loss": 0.1623,
+      "step": 3180
+    },
+    {
+      "epoch": 2.5809061488673137,
+      "grad_norm": 0.19242866333812558,
+      "learning_rate": 2.9253280772732595e-06,
+      "loss": 0.1627,
+      "step": 3190
+    },
+    {
+      "epoch": 2.588996763754045,
+      "grad_norm": 0.17889937009144094,
+      "learning_rate": 2.8158302603540965e-06,
+      "loss": 0.1609,
+      "step": 3200
+    },
+    {
+      "epoch": 2.5970873786407767,
+      "grad_norm": 0.1879805259242369,
+      "learning_rate": 2.708298641679105e-06,
+      "loss": 0.1642,
+      "step": 3210
+    },
+    {
+      "epoch": 2.605177993527508,
+      "grad_norm": 0.18048882079511047,
+      "learning_rate": 2.6027427518497153e-06,
+      "loss": 0.1634,
+      "step": 3220
+    },
+    {
+      "epoch": 2.6132686084142396,
+      "grad_norm": 0.18903574393083356,
+      "learning_rate": 2.49917194635714e-06,
+      "loss": 0.1607,
+      "step": 3230
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.18631019977805036,
+      "learning_rate": 2.397595404753225e-06,
+      "loss": 0.1589,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6294498381877025,
+      "grad_norm": 0.18583873327326217,
+      "learning_rate": 2.2980221298367995e-06,
+      "loss": 0.1679,
+      "step": 3250
+    },
+    {
+      "epoch": 2.6375404530744335,
+      "grad_norm": 0.1872827952915852,
+      "learning_rate": 2.2004609468558175e-06,
+      "loss": 0.1648,
+      "step": 3260
+    },
+    {
+      "epoch": 2.645631067961165,
+      "grad_norm": 0.18646198991723772,
+      "learning_rate": 2.1049205027251216e-06,
+      "loss": 0.1648,
+      "step": 3270
+    },
+    {
+      "epoch": 2.6537216828478964,
+      "grad_norm": 0.18510117286188082,
+      "learning_rate": 2.0114092652600806e-06,
+      "loss": 0.1669,
+      "step": 3280
+    },
+    {
+      "epoch": 2.661812297734628,
+      "grad_norm": 0.1813478635510933,
+      "learning_rate": 1.919935522426081e-06,
+      "loss": 0.1601,
+      "step": 3290
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.18995739003728715,
+      "learning_rate": 1.8305073816039492e-06,
+      "loss": 0.165,
+      "step": 3300
+    },
+    {
+      "epoch": 2.6779935275080904,
+      "grad_norm": 0.18850255636775962,
+      "learning_rate": 1.7431327688714139e-06,
+      "loss": 0.1686,
+      "step": 3310
+    },
+    {
+      "epoch": 2.686084142394822,
+      "grad_norm": 0.19405450386616901,
+      "learning_rate": 1.6578194283005804e-06,
+      "loss": 0.1586,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6941747572815533,
+      "grad_norm": 0.18168641096398497,
+      "learning_rate": 1.5745749212715794e-06,
+      "loss": 0.1669,
+      "step": 3330
+    },
+    {
+      "epoch": 2.7022653721682848,
+      "grad_norm": 0.18390467428127966,
+      "learning_rate": 1.4934066258024182e-06,
+      "loss": 0.1681,
+      "step": 3340
+    },
+    {
+      "epoch": 2.7103559870550162,
+      "grad_norm": 0.18623903954773796,
+      "learning_rate": 1.4143217358950217e-06,
+      "loss": 0.1648,
+      "step": 3350
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.18708444577479785,
+      "learning_rate": 1.3373272608976668e-06,
+      "loss": 0.1635,
+      "step": 3360
+    },
+    {
+      "epoch": 2.726537216828479,
+      "grad_norm": 0.1850383082145996,
+      "learning_rate": 1.2624300248836928e-06,
+      "loss": 0.1607,
+      "step": 3370
+    },
+    {
+      "epoch": 2.73462783171521,
+      "grad_norm": 0.19358386622393567,
+      "learning_rate": 1.1896366660467173e-06,
+      "loss": 0.1645,
+      "step": 3380
+    },
+    {
+      "epoch": 2.7427184466019416,
+      "grad_norm": 0.17960690916881664,
+      "learning_rate": 1.1189536361122799e-06,
+      "loss": 0.1628,
+      "step": 3390
+    },
+    {
+      "epoch": 2.750809061488673,
+      "grad_norm": 0.20127134763525154,
+      "learning_rate": 1.0503871997660036e-06,
+      "loss": 0.168,
+      "step": 3400
+    },
+    {
+      "epoch": 2.7588996763754046,
+      "grad_norm": 0.19658936043581707,
+      "learning_rate": 9.83943434098372e-07,
+      "loss": 0.1633,
+      "step": 3410
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.18650679653778446,
+      "learning_rate": 9.196282280661023e-07,
+      "loss": 0.1673,
+      "step": 3420
+    },
+    {
+      "epoch": 2.775080906148867,
+      "grad_norm": 0.1733323921140541,
+      "learning_rate": 8.574472819702029e-07,
+      "loss": 0.1652,
+      "step": 3430
+    },
+    {
+      "epoch": 2.783171521035599,
+      "grad_norm": 0.17972872694953243,
+      "learning_rate": 7.974061069507571e-07,
+      "loss": 0.1636,
+      "step": 3440
+    },
+    {
+      "epoch": 2.79126213592233,
+      "grad_norm": 0.1777461753836034,
+      "learning_rate": 7.395100244984604e-07,
+      "loss": 0.1634,
+      "step": 3450
+    },
+    {
+      "epoch": 2.7993527508090614,
+      "grad_norm": 0.18989630729322288,
+      "learning_rate": 6.837641659829807e-07,
+      "loss": 0.1661,
+      "step": 3460
+    },
+    {
+      "epoch": 2.807443365695793,
+      "grad_norm": 0.17625775327505652,
+      "learning_rate": 6.301734721981533e-07,
+      "loss": 0.1643,
+      "step": 3470
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.1862820489099266,
+      "learning_rate": 5.787426929240808e-07,
+      "loss": 0.1643,
+      "step": 3480
+    },
+    {
+      "epoch": 2.823624595469256,
+      "grad_norm": 0.1790622651864573,
+      "learning_rate": 5.294763865061558e-07,
+      "loss": 0.1626,
+      "step": 3490
+    },
+    {
+      "epoch": 2.831715210355987,
+      "grad_norm": 0.173713637483726,
+      "learning_rate": 4.823789194510514e-07,
+      "loss": 0.1639,
+      "step": 3500
+    },
+    {
+      "epoch": 2.8398058252427183,
+      "grad_norm": 0.1866652309901256,
+      "learning_rate": 4.3745446603971064e-07,
+      "loss": 0.1602,
+      "step": 3510
+    },
+    {
+      "epoch": 2.8478964401294498,
+      "grad_norm": 0.17910113776598907,
+      "learning_rate": 3.947070079573872e-07,
+      "loss": 0.1642,
+      "step": 3520
+    },
+    {
+      "epoch": 2.855987055016181,
+      "grad_norm": 0.18065751924055337,
+      "learning_rate": 3.541403339407279e-07,
+      "loss": 0.1656,
+      "step": 3530
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.18007167117331419,
+      "learning_rate": 3.1575803944199624e-07,
+      "loss": 0.1651,
+      "step": 3540
+    },
+    {
+      "epoch": 2.872168284789644,
+      "grad_norm": 0.18247762658055755,
+      "learning_rate": 2.7956352631038906e-07,
+      "loss": 0.1623,
+      "step": 3550
+    },
+    {
+      "epoch": 2.8802588996763756,
+      "grad_norm": 0.17849381652959184,
+      "learning_rate": 2.4556000249054133e-07,
+      "loss": 0.1653,
+      "step": 3560
+    },
+    {
+      "epoch": 2.8883495145631066,
+      "grad_norm": 0.1846373828579316,
+      "learning_rate": 2.1375048173818412e-07,
+      "loss": 0.1691,
+      "step": 3570
+    },
+    {
+      "epoch": 2.896440129449838,
+      "grad_norm": 0.17151731227871805,
+      "learning_rate": 1.8413778335305e-07,
+      "loss": 0.1606,
+      "step": 3580
+    },
+    {
+      "epoch": 2.9045307443365695,
+      "grad_norm": 0.22901848451176904,
+      "learning_rate": 1.567245319290006e-07,
+      "loss": 0.1607,
+      "step": 3590
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.1841472999781577,
+      "learning_rate": 1.315131571213879e-07,
+      "loss": 0.1659,
+      "step": 3600
+    },
+    {
+      "epoch": 2.9207119741100325,
+      "grad_norm": 0.17620735263373793,
+      "learning_rate": 1.0850589343172624e-07,
+      "loss": 0.163,
+      "step": 3610
+    },
+    {
+      "epoch": 2.9288025889967635,
+      "grad_norm": 0.1732366383013906,
+      "learning_rate": 8.770478000964532e-08,
+      "loss": 0.1603,
+      "step": 3620
+    },
+    {
+      "epoch": 2.9368932038834954,
+      "grad_norm": 0.18106989776343552,
+      "learning_rate": 6.911166047215145e-08,
+      "loss": 0.1639,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9449838187702264,
+      "grad_norm": 0.18480994602766684,
+      "learning_rate": 5.272818274023872e-08,
+      "loss": 0.163,
+      "step": 3640
+    },
+    {
+      "epoch": 2.953074433656958,
+      "grad_norm": 0.17298187343963853,
+      "learning_rate": 3.855579889282257e-08,
+      "loss": 0.1607,
+      "step": 3650
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.1826761908488716,
+      "learning_rate": 2.6595765038045507e-08,
+      "loss": 0.1598,
+      "step": 3660
+    },
+    {
+      "epoch": 2.969255663430421,
+      "grad_norm": 0.18104206371371273,
+      "learning_rate": 1.6849141201946693e-08,
+      "loss": 0.1656,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9773462783171523,
+      "grad_norm": 0.17420913687043998,
+      "learning_rate": 9.316791234506572e-09,
+      "loss": 0.1589,
+      "step": 3680
+    },
+    {
+      "epoch": 2.9854368932038833,
+      "grad_norm": 0.177305619544985,
+      "learning_rate": 3.999382733096968e-09,
+      "loss": 0.1633,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9935275080906147,
+      "grad_norm": 0.17610657137080274,
+      "learning_rate": 8.973869832895609e-10,
+      "loss": 0.1623,
+      "step": 3700
+    },
+    {
+      "epoch": 3.0,
+      "step": 3708,
+      "total_flos": 6.364819972765516e+18,
+      "train_loss": 0.2649726264235014,
+      "train_runtime": 208629.7363,
+      "train_samples_per_second": 0.569,
+      "train_steps_per_second": 0.018
+    }
+  ],
+  "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": 6.364819972765516e+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:cea2cc34d068b43f5f15fcf4b3b00ad0f535091c89422c3523e5bd862f073f10
+size 7736