Upload folder using huggingface_hub

3b-w-cot/README.md

@@ -0,0 +1,165 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen2.5-3B-Instruct
+tags:
+- generated_from_trainer
+datasets:
+- train.jsonl
+model-index:
+- name: outputs/out
+  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. -->
+
+[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
+<details><summary>See axolotl config</summary>
+
+axolotl version: `0.7.0`
+```yaml
+base_model: Qwen/Qwen2.5-3B-Instruct
+model_type: AutoModelForCausalLM
+tokenizer_type: AutoTokenizer
+trust_remote_code: false
+
+load_in_8bit: false
+load_in_4bit: false
+strict: false
+
+output_dir: ./outputs/out
+chat_template: qwen_25
+datasets:
+  - path: train.jsonl
+    type: chat_template
+    field_messages: messages
+    message_field_role: role
+    message_field_content: content
+    roles:
+      system:
+        - system
+      user:
+        - user
+      assistant:
+        - assistant
+
+dataset_prepared_path: last_run_prepared
+val_set_size: 0.005
+output_dir: ./outputs/out
+eval_sample_packing: False
+
+sequence_len: 8192
+sample_packing: False
+pad_to_sequence_len: False
+
+wandb_project: mergedbench
+wandb_entity:
+wandb_watch:
+wandb_name:
+wandb_log_model:
+
+plugins:
+  - axolotl.integrations.liger.LigerPlugin
+liger_rope: true
+liger_rms_norm: true
+liger_swiglu: true
+liger_fused_linear_cross_entropy: true
+
+gradient_accumulation_steps: 4
+micro_batch_size: 8
+eval_batch_size: 4
+num_epochs: 3
+optimizer: paged_adamw_8bit
+lr_scheduler: cosine
+learning_rate: 2e-5
+
+train_on_inputs: false
+group_by_length: false
+bf16: auto
+fp16:
+tf32: false
+
+gradient_checkpointing: true
+gradient_checkpointing_kwargs:
+  use_reentrant: false
+early_stopping_patience:
+resume_from_checkpoint:
+logging_steps: 1
+xformers_attention:
+flash_attention: true
+
+warmup_steps: 30
+evals_per_epoch: 3
+eval_max_new_tokens: 128
+eval_table_size:
+saves_per_epoch: 1
+debug:
+deepspeed: deepspeed_configs/zero1.json
+weight_decay: 0.01
+fsdp:
+fsdp_config:
+special_tokens:
+```
+
+</details><br>
+
+# outputs/out
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) on the train.jsonl dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.2847
+
+## 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: 2e-05
+- train_batch_size: 8
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 64
+- total_eval_batch_size: 8
+- optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 30
+- num_epochs: 3.0
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 1.4163        | 0.0040 | 1    | 1.4218          |
+| 0.3799        | 0.3323 | 83   | 0.3376          |
+| 0.3263        | 0.6647 | 166  | 0.3207          |
+| 0.3213        | 0.9970 | 249  | 0.3041          |
+| 0.2369        | 1.3283 | 332  | 0.3128          |
+| 0.2436        | 1.6607 | 415  | 0.3041          |
+| 0.2159        | 1.9930 | 498  | 0.2962          |
+| 0.1832        | 2.3243 | 581  | 0.2914          |
+| 0.1941        | 2.6567 | 664  | 0.2865          |
+| 0.185         | 2.9890 | 747  | 0.2847          |
+
+
+### Framework versions
+
+- Transformers 4.48.3
+- Pytorch 2.5.1+cu121
+- Datasets 3.2.0
+- Tokenizers 0.21.0

3b-w-cot/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
+}

3b-w-cot/checkpoint-249/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
+}

3b-w-cot/checkpoint-249/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-3B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.3",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

3b-w-cot/checkpoint-249/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.48.3"
+}

3b-w-cot/checkpoint-249/latest

@@ -0,0 +1 @@
+global_step249

3b-w-cot/checkpoint-249/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d98d3d655ac51584b495c5652e2bb2cb14f1632265a8d91ec83deece94fc4242
+size 4957560304

3b-w-cot/checkpoint-249/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9e5b0033ab9498c19469e02588eda7896057da2526c4590819382584ce0c317
+size 1836696752

3b-w-cot/checkpoint-249/model.safetensors.index.json

@@ -0,0 +1,442 @@
+{
+  "metadata": {
+    "total_size": 6794207232
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

3b-w-cot/checkpoint-249/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7e52325e9d729519836af640f8f754a93ee06730fb2953b5309434b53b17562
+size 14512

3b-w-cot/checkpoint-249/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a93593cf0342eb47876986e1063102e1546354426a2324c46ddcf1cbecae803
+size 14512

3b-w-cot/checkpoint-249/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe57fe36c47ba8cc85686021517d2af1000494c57709a51ad19a90ac2cb505a7
+size 1064

3b-w-cot/checkpoint-249/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": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

3b-w-cot/checkpoint-249/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

3b-w-cot/checkpoint-249/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,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

3b-w-cot/checkpoint-249/trainer_state.json

@@ -0,0 +1,1808 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.996996996996997,
+  "eval_steps": 83,
+  "global_step": 249,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004004004004004004,
+      "grad_norm": 6.739165782928467,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.4163,
+      "step": 1
+    },
+    {
+      "epoch": 0.004004004004004004,
+      "eval_loss": 1.4217944145202637,
+      "eval_runtime": 5.678,
+      "eval_samples_per_second": 14.266,
+      "eval_steps_per_second": 1.937,
+      "step": 1
+    },
+    {
+      "epoch": 0.008008008008008008,
+      "grad_norm": 6.897192001342773,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.3767,
+      "step": 2
+    },
+    {
+      "epoch": 0.012012012012012012,
+      "grad_norm": 6.74431037902832,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.3848,
+      "step": 3
+    },
+    {
+      "epoch": 0.016016016016016016,
+      "grad_norm": 6.999237537384033,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.4294,
+      "step": 4
+    },
+    {
+      "epoch": 0.02002002002002002,
+      "grad_norm": 4.510103702545166,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.2646,
+      "step": 5
+    },
+    {
+      "epoch": 0.024024024024024024,
+      "grad_norm": 4.533900737762451,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.2563,
+      "step": 6
+    },
+    {
+      "epoch": 0.028028028028028028,
+      "grad_norm": 3.565216541290283,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.1061,
+      "step": 7
+    },
+    {
+      "epoch": 0.03203203203203203,
+      "grad_norm": 3.1937592029571533,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.0445,
+      "step": 8
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 3.0942018032073975,
+      "learning_rate": 6e-06,
+      "loss": 0.8555,
+      "step": 9
+    },
+    {
+      "epoch": 0.04004004004004004,
+      "grad_norm": 4.628591060638428,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8539,
+      "step": 10
+    },
+    {
+      "epoch": 0.044044044044044044,
+      "grad_norm": 5.402413845062256,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.7684,
+      "step": 11
+    },
+    {
+      "epoch": 0.04804804804804805,
+      "grad_norm": 1.96303391456604,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6347,
+      "step": 12
+    },
+    {
+      "epoch": 0.05205205205205205,
+      "grad_norm": 1.33661687374115,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.6692,
+      "step": 13
+    },
+    {
+      "epoch": 0.056056056056056056,
+      "grad_norm": 0.9704915285110474,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.6385,
+      "step": 14
+    },
+    {
+      "epoch": 0.06006006006006006,
+      "grad_norm": 0.7543495893478394,
+      "learning_rate": 1e-05,
+      "loss": 0.5901,
+      "step": 15
+    },
+    {
+      "epoch": 0.06406406406406406,
+      "grad_norm": 1.3593250513076782,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.5645,
+      "step": 16
+    },
+    {
+      "epoch": 0.06806806806806807,
+      "grad_norm": 1.147750735282898,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 0.5508,
+      "step": 17
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 0.8024477958679199,
+      "learning_rate": 1.2e-05,
+      "loss": 0.5282,
+      "step": 18
+    },
+    {
+      "epoch": 0.07607607607607608,
+      "grad_norm": 0.7931386232376099,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 0.4929,
+      "step": 19
+    },
+    {
+      "epoch": 0.08008008008008008,
+      "grad_norm": 0.6702373623847961,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4868,
+      "step": 20
+    },
+    {
+      "epoch": 0.08408408408408409,
+      "grad_norm": 0.747148871421814,
+      "learning_rate": 1.4e-05,
+      "loss": 0.4857,
+      "step": 21
+    },
+    {
+      "epoch": 0.08808808808808809,
+      "grad_norm": 0.5935061573982239,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 0.4908,
+      "step": 22
+    },
+    {
+      "epoch": 0.0920920920920921,
+      "grad_norm": 0.5696635842323303,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.4437,
+      "step": 23
+    },
+    {
+      "epoch": 0.0960960960960961,
+      "grad_norm": 0.5861401557922363,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.4442,
+      "step": 24
+    },
+    {
+      "epoch": 0.1001001001001001,
+      "grad_norm": 0.5833226442337036,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.4585,
+      "step": 25
+    },
+    {
+      "epoch": 0.1041041041041041,
+      "grad_norm": 0.48214447498321533,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.3994,
+      "step": 26
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 0.49264758825302124,
+      "learning_rate": 1.8e-05,
+      "loss": 0.4609,
+      "step": 27
+    },
+    {
+      "epoch": 0.11211211211211211,
+      "grad_norm": 0.5025641322135925,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.4816,
+      "step": 28
+    },
+    {
+      "epoch": 0.11611611611611612,
+      "grad_norm": 0.46342933177948,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 0.4435,
+      "step": 29
+    },
+    {
+      "epoch": 0.12012012012012012,
+      "grad_norm": 0.46523571014404297,
+      "learning_rate": 2e-05,
+      "loss": 0.4208,
+      "step": 30
+    },
+    {
+      "epoch": 0.12412412412412413,
+      "grad_norm": 0.4298263192176819,
+      "learning_rate": 1.9999904008949705e-05,
+      "loss": 0.4137,
+      "step": 31
+    },
+    {
+      "epoch": 0.12812812812812813,
+      "grad_norm": 0.43352752923965454,
+      "learning_rate": 1.999961603764167e-05,
+      "loss": 0.4096,
+      "step": 32
+    },
+    {
+      "epoch": 0.13213213213213212,
+      "grad_norm": 0.4369907081127167,
+      "learning_rate": 1.9999136091604433e-05,
+      "loss": 0.4128,
+      "step": 33
+    },
+    {
+      "epoch": 0.13613613613613615,
+      "grad_norm": 0.4188078045845032,
+      "learning_rate": 1.99984641800521e-05,
+      "loss": 0.4171,
+      "step": 34
+    },
+    {
+      "epoch": 0.14014014014014015,
+      "grad_norm": 0.4548235237598419,
+      "learning_rate": 1.9997600315884166e-05,
+      "loss": 0.4236,
+      "step": 35
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 0.4751495122909546,
+      "learning_rate": 1.999654451568528e-05,
+      "loss": 0.438,
+      "step": 36
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 0.4333605170249939,
+      "learning_rate": 1.9995296799724914e-05,
+      "loss": 0.4208,
+      "step": 37
+    },
+    {
+      "epoch": 0.15215215215215216,
+      "grad_norm": 0.48084449768066406,
+      "learning_rate": 1.999385719195698e-05,
+      "loss": 0.3969,
+      "step": 38
+    },
+    {
+      "epoch": 0.15615615615615616,
+      "grad_norm": 0.42627257108688354,
+      "learning_rate": 1.9992225720019377e-05,
+      "loss": 0.4503,
+      "step": 39
+    },
+    {
+      "epoch": 0.16016016016016016,
+      "grad_norm": 0.41883718967437744,
+      "learning_rate": 1.9990402415233436e-05,
+      "loss": 0.4322,
+      "step": 40
+    },
+    {
+      "epoch": 0.16416416416416416,
+      "grad_norm": 0.39587706327438354,
+      "learning_rate": 1.998838731260335e-05,
+      "loss": 0.3841,
+      "step": 41
+    },
+    {
+      "epoch": 0.16816816816816818,
+      "grad_norm": 0.40334609150886536,
+      "learning_rate": 1.9986180450815485e-05,
+      "loss": 0.3737,
+      "step": 42
+    },
+    {
+      "epoch": 0.17217217217217218,
+      "grad_norm": 0.3715899884700775,
+      "learning_rate": 1.9983781872237634e-05,
+      "loss": 0.387,
+      "step": 43
+    },
+    {
+      "epoch": 0.17617617617617617,
+      "grad_norm": 0.38875311613082886,
+      "learning_rate": 1.9981191622918217e-05,
+      "loss": 0.3695,
+      "step": 44
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 0.39087679982185364,
+      "learning_rate": 1.997840975258538e-05,
+      "loss": 0.3934,
+      "step": 45
+    },
+    {
+      "epoch": 0.1841841841841842,
+      "grad_norm": 0.41504260897636414,
+      "learning_rate": 1.9975436314646052e-05,
+      "loss": 0.3769,
+      "step": 46
+    },
+    {
+      "epoch": 0.1881881881881882,
+      "grad_norm": 0.36994805932044983,
+      "learning_rate": 1.9972271366184922e-05,
+      "loss": 0.3807,
+      "step": 47
+    },
+    {
+      "epoch": 0.1921921921921922,
+      "grad_norm": 0.39996904134750366,
+      "learning_rate": 1.996891496796334e-05,
+      "loss": 0.3903,
+      "step": 48
+    },
+    {
+      "epoch": 0.1961961961961962,
+      "grad_norm": 0.3904661536216736,
+      "learning_rate": 1.9965367184418138e-05,
+      "loss": 0.3785,
+      "step": 49
+    },
+    {
+      "epoch": 0.2002002002002002,
+      "grad_norm": 0.39596742391586304,
+      "learning_rate": 1.9961628083660406e-05,
+      "loss": 0.3806,
+      "step": 50
+    },
+    {
+      "epoch": 0.2042042042042042,
+      "grad_norm": 0.39424648880958557,
+      "learning_rate": 1.9957697737474198e-05,
+      "loss": 0.3736,
+      "step": 51
+    },
+    {
+      "epoch": 0.2082082082082082,
+      "grad_norm": 0.41719183325767517,
+      "learning_rate": 1.9953576221315116e-05,
+      "loss": 0.4023,
+      "step": 52
+    },
+    {
+      "epoch": 0.2122122122122122,
+      "grad_norm": 0.3773253262042999,
+      "learning_rate": 1.9949263614308894e-05,
+      "loss": 0.396,
+      "step": 53
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.4236510992050171,
+      "learning_rate": 1.994475999924987e-05,
+      "loss": 0.4032,
+      "step": 54
+    },
+    {
+      "epoch": 0.22022022022022023,
+      "grad_norm": 0.3507007360458374,
+      "learning_rate": 1.9940065462599394e-05,
+      "loss": 0.3615,
+      "step": 55
+    },
+    {
+      "epoch": 0.22422422422422422,
+      "grad_norm": 0.45851731300354004,
+      "learning_rate": 1.9935180094484164e-05,
+      "loss": 0.3402,
+      "step": 56
+    },
+    {
+      "epoch": 0.22822822822822822,
+      "grad_norm": 0.36239954829216003,
+      "learning_rate": 1.99301039886945e-05,
+      "loss": 0.3668,
+      "step": 57
+    },
+    {
+      "epoch": 0.23223223223223224,
+      "grad_norm": 0.3855980634689331,
+      "learning_rate": 1.992483724268255e-05,
+      "loss": 0.3575,
+      "step": 58
+    },
+    {
+      "epoch": 0.23623623623623624,
+      "grad_norm": 0.4716965854167938,
+      "learning_rate": 1.9919379957560413e-05,
+      "loss": 0.3532,
+      "step": 59
+    },
+    {
+      "epoch": 0.24024024024024024,
+      "grad_norm": 0.45167747139930725,
+      "learning_rate": 1.991373223809819e-05,
+      "loss": 0.3537,
+      "step": 60
+    },
+    {
+      "epoch": 0.24424424424424424,
+      "grad_norm": 0.420512855052948,
+      "learning_rate": 1.990789419272199e-05,
+      "loss": 0.3471,
+      "step": 61
+    },
+    {
+      "epoch": 0.24824824824824826,
+      "grad_norm": 0.4863167107105255,
+      "learning_rate": 1.9901865933511834e-05,
+      "loss": 0.3741,
+      "step": 62
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 0.39653480052948,
+      "learning_rate": 1.9895647576199507e-05,
+      "loss": 0.3799,
+      "step": 63
+    },
+    {
+      "epoch": 0.25625625625625625,
+      "grad_norm": 0.38418489694595337,
+      "learning_rate": 1.988923924016634e-05,
+      "loss": 0.3577,
+      "step": 64
+    },
+    {
+      "epoch": 0.2602602602602603,
+      "grad_norm": 0.39798790216445923,
+      "learning_rate": 1.988264104844091e-05,
+      "loss": 0.3581,
+      "step": 65
+    },
+    {
+      "epoch": 0.26426426426426425,
+      "grad_norm": 0.41135266423225403,
+      "learning_rate": 1.987585312769669e-05,
+      "loss": 0.3306,
+      "step": 66
+    },
+    {
+      "epoch": 0.2682682682682683,
+      "grad_norm": 0.36515921354293823,
+      "learning_rate": 1.9868875608249613e-05,
+      "loss": 0.3628,
+      "step": 67
+    },
+    {
+      "epoch": 0.2722722722722723,
+      "grad_norm": 0.4733300507068634,
+      "learning_rate": 1.986170862405556e-05,
+      "loss": 0.3733,
+      "step": 68
+    },
+    {
+      "epoch": 0.27627627627627627,
+      "grad_norm": 0.4218703806400299,
+      "learning_rate": 1.98543523127078e-05,
+      "loss": 0.371,
+      "step": 69
+    },
+    {
+      "epoch": 0.2802802802802803,
+      "grad_norm": 0.35307836532592773,
+      "learning_rate": 1.984680681543434e-05,
+      "loss": 0.3568,
+      "step": 70
+    },
+    {
+      "epoch": 0.28428428428428426,
+      "grad_norm": 0.4475674629211426,
+      "learning_rate": 1.9839072277095222e-05,
+      "loss": 0.3828,
+      "step": 71
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 0.37015393376350403,
+      "learning_rate": 1.9831148846179743e-05,
+      "loss": 0.3426,
+      "step": 72
+    },
+    {
+      "epoch": 0.2922922922922923,
+      "grad_norm": 0.39532360434532166,
+      "learning_rate": 1.9823036674803585e-05,
+      "loss": 0.3545,
+      "step": 73
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 0.349669486284256,
+      "learning_rate": 1.981473591870593e-05,
+      "loss": 0.332,
+      "step": 74
+    },
+    {
+      "epoch": 0.3003003003003003,
+      "grad_norm": 0.356000155210495,
+      "learning_rate": 1.980624673724643e-05,
+      "loss": 0.3684,
+      "step": 75
+    },
+    {
+      "epoch": 0.30430430430430433,
+      "grad_norm": 0.4027644693851471,
+      "learning_rate": 1.9797569293402174e-05,
+      "loss": 0.3592,
+      "step": 76
+    },
+    {
+      "epoch": 0.3083083083083083,
+      "grad_norm": 0.36116528511047363,
+      "learning_rate": 1.9788703753764554e-05,
+      "loss": 0.3433,
+      "step": 77
+    },
+    {
+      "epoch": 0.3123123123123123,
+      "grad_norm": 0.35624876618385315,
+      "learning_rate": 1.9779650288536057e-05,
+      "loss": 0.3541,
+      "step": 78
+    },
+    {
+      "epoch": 0.3163163163163163,
+      "grad_norm": 0.360287606716156,
+      "learning_rate": 1.977040907152702e-05,
+      "loss": 0.3413,
+      "step": 79
+    },
+    {
+      "epoch": 0.3203203203203203,
+      "grad_norm": 0.412913978099823,
+      "learning_rate": 1.976098028015226e-05,
+      "loss": 0.4142,
+      "step": 80
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.38022932410240173,
+      "learning_rate": 1.9751364095427694e-05,
+      "loss": 0.3481,
+      "step": 81
+    },
+    {
+      "epoch": 0.3283283283283283,
+      "grad_norm": 0.37813299894332886,
+      "learning_rate": 1.974156070196686e-05,
+      "loss": 0.3552,
+      "step": 82
+    },
+    {
+      "epoch": 0.33233233233233234,
+      "grad_norm": 0.3782276511192322,
+      "learning_rate": 1.973157028797737e-05,
+      "loss": 0.3799,
+      "step": 83
+    },
+    {
+      "epoch": 0.33233233233233234,
+      "eval_loss": 0.33762940764427185,
+      "eval_runtime": 6.2209,
+      "eval_samples_per_second": 13.021,
+      "eval_steps_per_second": 1.768,
+      "step": 83
+    },
+    {
+      "epoch": 0.33633633633633636,
+      "grad_norm": 0.3933849036693573,
+      "learning_rate": 1.9721393045257277e-05,
+      "loss": 0.3654,
+      "step": 84
+    },
+    {
+      "epoch": 0.34034034034034033,
+      "grad_norm": 0.35808295011520386,
+      "learning_rate": 1.9711029169191437e-05,
+      "loss": 0.3716,
+      "step": 85
+    },
+    {
+      "epoch": 0.34434434434434436,
+      "grad_norm": 0.3547224700450897,
+      "learning_rate": 1.970047885874771e-05,
+      "loss": 0.3397,
+      "step": 86
+    },
+    {
+      "epoch": 0.3483483483483483,
+      "grad_norm": 0.39883747696876526,
+      "learning_rate": 1.968974231647318e-05,
+      "loss": 0.3645,
+      "step": 87
+    },
+    {
+      "epoch": 0.35235235235235235,
+      "grad_norm": 0.3208443224430084,
+      "learning_rate": 1.9678819748490236e-05,
+      "loss": 0.3431,
+      "step": 88
+    },
+    {
+      "epoch": 0.3563563563563564,
+      "grad_norm": 0.34669625759124756,
+      "learning_rate": 1.9667711364492638e-05,
+      "loss": 0.3613,
+      "step": 89
+    },
+    {
+      "epoch": 0.36036036036036034,
+      "grad_norm": 0.35220351815223694,
+      "learning_rate": 1.965641737774147e-05,
+      "loss": 0.3499,
+      "step": 90
+    },
+    {
+      "epoch": 0.36436436436436437,
+      "grad_norm": 0.3566596806049347,
+      "learning_rate": 1.9644938005061062e-05,
+      "loss": 0.3204,
+      "step": 91
+    },
+    {
+      "epoch": 0.3683683683683684,
+      "grad_norm": 0.3813494145870209,
+      "learning_rate": 1.9633273466834826e-05,
+      "loss": 0.3526,
+      "step": 92
+    },
+    {
+      "epoch": 0.37237237237237236,
+      "grad_norm": 0.3885157108306885,
+      "learning_rate": 1.9621423987001013e-05,
+      "loss": 0.3562,
+      "step": 93
+    },
+    {
+      "epoch": 0.3763763763763764,
+      "grad_norm": 0.4746558666229248,
+      "learning_rate": 1.960938979304843e-05,
+      "loss": 0.3509,
+      "step": 94
+    },
+    {
+      "epoch": 0.38038038038038036,
+      "grad_norm": 0.3623756766319275,
+      "learning_rate": 1.959717111601206e-05,
+      "loss": 0.371,
+      "step": 95
+    },
+    {
+      "epoch": 0.3843843843843844,
+      "grad_norm": 0.3858593702316284,
+      "learning_rate": 1.9584768190468624e-05,
+      "loss": 0.3551,
+      "step": 96
+    },
+    {
+      "epoch": 0.3883883883883884,
+      "grad_norm": 0.3788565993309021,
+      "learning_rate": 1.95721812545321e-05,
+      "loss": 0.3669,
+      "step": 97
+    },
+    {
+      "epoch": 0.3923923923923924,
+      "grad_norm": 0.3729216754436493,
+      "learning_rate": 1.9559410549849125e-05,
+      "loss": 0.34,
+      "step": 98
+    },
+    {
+      "epoch": 0.3963963963963964,
+      "grad_norm": 0.44008028507232666,
+      "learning_rate": 1.9546456321594374e-05,
+      "loss": 0.3898,
+      "step": 99
+    },
+    {
+      "epoch": 0.4004004004004004,
+      "grad_norm": 0.36622726917266846,
+      "learning_rate": 1.9533318818465837e-05,
+      "loss": 0.3624,
+      "step": 100
+    },
+    {
+      "epoch": 0.4044044044044044,
+      "grad_norm": 0.39071497321128845,
+      "learning_rate": 1.9519998292680062e-05,
+      "loss": 0.3518,
+      "step": 101
+    },
+    {
+      "epoch": 0.4084084084084084,
+      "grad_norm": 0.40084153413772583,
+      "learning_rate": 1.9506494999967298e-05,
+      "loss": 0.3483,
+      "step": 102
+    },
+    {
+      "epoch": 0.4124124124124124,
+      "grad_norm": 0.3901435434818268,
+      "learning_rate": 1.94928091995666e-05,
+      "loss": 0.3577,
+      "step": 103
+    },
+    {
+      "epoch": 0.4164164164164164,
+      "grad_norm": 0.34764233231544495,
+      "learning_rate": 1.9478941154220833e-05,
+      "loss": 0.3487,
+      "step": 104
+    },
+    {
+      "epoch": 0.42042042042042044,
+      "grad_norm": 0.4044027030467987,
+      "learning_rate": 1.9464891130171647e-05,
+      "loss": 0.3602,
+      "step": 105
+    },
+    {
+      "epoch": 0.4244244244244244,
+      "grad_norm": 0.3537923991680145,
+      "learning_rate": 1.9450659397154353e-05,
+      "loss": 0.3282,
+      "step": 106
+    },
+    {
+      "epoch": 0.42842842842842843,
+      "grad_norm": 0.35585689544677734,
+      "learning_rate": 1.9436246228392762e-05,
+      "loss": 0.361,
+      "step": 107
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.380585253238678,
+      "learning_rate": 1.94216519005939e-05,
+      "loss": 0.3218,
+      "step": 108
+    },
+    {
+      "epoch": 0.4364364364364364,
+      "grad_norm": 0.3388879597187042,
+      "learning_rate": 1.9406876693942747e-05,
+      "loss": 0.3563,
+      "step": 109
+    },
+    {
+      "epoch": 0.44044044044044045,
+      "grad_norm": 0.359130859375,
+      "learning_rate": 1.939192089209682e-05,
+      "loss": 0.3386,
+      "step": 110
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.35704970359802246,
+      "learning_rate": 1.9376784782180747e-05,
+      "loss": 0.3336,
+      "step": 111
+    },
+    {
+      "epoch": 0.44844844844844844,
+      "grad_norm": 0.3638094961643219,
+      "learning_rate": 1.9361468654780748e-05,
+      "loss": 0.3582,
+      "step": 112
+    },
+    {
+      "epoch": 0.45245245245245247,
+      "grad_norm": 0.3996574878692627,
+      "learning_rate": 1.9345972803939046e-05,
+      "loss": 0.3496,
+      "step": 113
+    },
+    {
+      "epoch": 0.45645645645645644,
+      "grad_norm": 0.3565851151943207,
+      "learning_rate": 1.9330297527148246e-05,
+      "loss": 0.3273,
+      "step": 114
+    },
+    {
+      "epoch": 0.46046046046046046,
+      "grad_norm": 0.3502778112888336,
+      "learning_rate": 1.9314443125345606e-05,
+      "loss": 0.3331,
+      "step": 115
+    },
+    {
+      "epoch": 0.4644644644644645,
+      "grad_norm": 0.4017852544784546,
+      "learning_rate": 1.929840990290726e-05,
+      "loss": 0.3522,
+      "step": 116
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "grad_norm": 0.354686975479126,
+      "learning_rate": 1.928219816764238e-05,
+      "loss": 0.3456,
+      "step": 117
+    },
+    {
+      "epoch": 0.4724724724724725,
+      "grad_norm": 0.34257611632347107,
+      "learning_rate": 1.9265808230787265e-05,
+      "loss": 0.3325,
+      "step": 118
+    },
+    {
+      "epoch": 0.47647647647647645,
+      "grad_norm": 0.3504616320133209,
+      "learning_rate": 1.9249240406999366e-05,
+      "loss": 0.3516,
+      "step": 119
+    },
+    {
+      "epoch": 0.4804804804804805,
+      "grad_norm": 0.36517202854156494,
+      "learning_rate": 1.9232495014351248e-05,
+      "loss": 0.3233,
+      "step": 120
+    },
+    {
+      "epoch": 0.4844844844844845,
+      "grad_norm": 0.33805495500564575,
+      "learning_rate": 1.921557237432447e-05,
+      "loss": 0.3261,
+      "step": 121
+    },
+    {
+      "epoch": 0.48848848848848847,
+      "grad_norm": 0.35013893246650696,
+      "learning_rate": 1.919847281180343e-05,
+      "loss": 0.3361,
+      "step": 122
+    },
+    {
+      "epoch": 0.4924924924924925,
+      "grad_norm": 0.3569040298461914,
+      "learning_rate": 1.9181196655069126e-05,
+      "loss": 0.33,
+      "step": 123
+    },
+    {
+      "epoch": 0.4964964964964965,
+      "grad_norm": 0.3355570137500763,
+      "learning_rate": 1.9163744235792845e-05,
+      "loss": 0.3263,
+      "step": 124
+    },
+    {
+      "epoch": 0.5005005005005005,
+      "grad_norm": 0.39212533831596375,
+      "learning_rate": 1.9146115889029793e-05,
+      "loss": 0.3671,
+      "step": 125
+    },
+    {
+      "epoch": 0.5045045045045045,
+      "grad_norm": 0.3458411395549774,
+      "learning_rate": 1.912831195321268e-05,
+      "loss": 0.3575,
+      "step": 126
+    },
+    {
+      "epoch": 0.5085085085085085,
+      "grad_norm": 0.3606933057308197,
+      "learning_rate": 1.9110332770145198e-05,
+      "loss": 0.342,
+      "step": 127
+    },
+    {
+      "epoch": 0.5125125125125125,
+      "grad_norm": 0.3897223472595215,
+      "learning_rate": 1.9092178684995487e-05,
+      "loss": 0.37,
+      "step": 128
+    },
+    {
+      "epoch": 0.5165165165165165,
+      "grad_norm": 0.30565860867500305,
+      "learning_rate": 1.9073850046289484e-05,
+      "loss": 0.3331,
+      "step": 129
+    },
+    {
+      "epoch": 0.5205205205205206,
+      "grad_norm": 0.3087623119354248,
+      "learning_rate": 1.9055347205904245e-05,
+      "loss": 0.322,
+      "step": 130
+    },
+    {
+      "epoch": 0.5245245245245245,
+      "grad_norm": 0.34453681111335754,
+      "learning_rate": 1.903667051906119e-05,
+      "loss": 0.3405,
+      "step": 131
+    },
+    {
+      "epoch": 0.5285285285285285,
+      "grad_norm": 0.32023486495018005,
+      "learning_rate": 1.901782034431927e-05,
+      "loss": 0.338,
+      "step": 132
+    },
+    {
+      "epoch": 0.5325325325325325,
+      "grad_norm": 0.37953877449035645,
+      "learning_rate": 1.8998797043568102e-05,
+      "loss": 0.3406,
+      "step": 133
+    },
+    {
+      "epoch": 0.5365365365365365,
+      "grad_norm": 0.3680444061756134,
+      "learning_rate": 1.8979600982021014e-05,
+      "loss": 0.324,
+      "step": 134
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.3368768095970154,
+      "learning_rate": 1.896023252820802e-05,
+      "loss": 0.3194,
+      "step": 135
+    },
+    {
+      "epoch": 0.5445445445445446,
+      "grad_norm": 0.37519025802612305,
+      "learning_rate": 1.8940692053968773e-05,
+      "loss": 0.3358,
+      "step": 136
+    },
+    {
+      "epoch": 0.5485485485485485,
+      "grad_norm": 0.33083492517471313,
+      "learning_rate": 1.89209799344454e-05,
+      "loss": 0.3202,
+      "step": 137
+    },
+    {
+      "epoch": 0.5525525525525525,
+      "grad_norm": 0.3258178234100342,
+      "learning_rate": 1.8901096548075305e-05,
+      "loss": 0.3186,
+      "step": 138
+    },
+    {
+      "epoch": 0.5565565565565566,
+      "grad_norm": 0.37846943736076355,
+      "learning_rate": 1.8881042276583924e-05,
+      "loss": 0.3563,
+      "step": 139
+    },
+    {
+      "epoch": 0.5605605605605606,
+      "grad_norm": 0.35151371359825134,
+      "learning_rate": 1.8860817504977374e-05,
+      "loss": 0.337,
+      "step": 140
+    },
+    {
+      "epoch": 0.5645645645645646,
+      "grad_norm": 0.33033043146133423,
+      "learning_rate": 1.8840422621535067e-05,
+      "loss": 0.3042,
+      "step": 141
+    },
+    {
+      "epoch": 0.5685685685685685,
+      "grad_norm": 0.3724791705608368,
+      "learning_rate": 1.881985801780225e-05,
+      "loss": 0.3165,
+      "step": 142
+    },
+    {
+      "epoch": 0.5725725725725725,
+      "grad_norm": 0.355752170085907,
+      "learning_rate": 1.8799124088582523e-05,
+      "loss": 0.3693,
+      "step": 143
+    },
+    {
+      "epoch": 0.5765765765765766,
+      "grad_norm": 0.3422398567199707,
+      "learning_rate": 1.8778221231930204e-05,
+      "loss": 0.3121,
+      "step": 144
+    },
+    {
+      "epoch": 0.5805805805805806,
+      "grad_norm": 0.3716834783554077,
+      "learning_rate": 1.8757149849142724e-05,
+      "loss": 0.3338,
+      "step": 145
+    },
+    {
+      "epoch": 0.5845845845845846,
+      "grad_norm": 0.313723087310791,
+      "learning_rate": 1.8735910344752925e-05,
+      "loss": 0.3294,
+      "step": 146
+    },
+    {
+      "epoch": 0.5885885885885885,
+      "grad_norm": 0.33704790472984314,
+      "learning_rate": 1.871450312652126e-05,
+      "loss": 0.3425,
+      "step": 147
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 0.3442137837409973,
+      "learning_rate": 1.8692928605428016e-05,
+      "loss": 0.3243,
+      "step": 148
+    },
+    {
+      "epoch": 0.5965965965965966,
+      "grad_norm": 0.3326564133167267,
+      "learning_rate": 1.8671187195665373e-05,
+      "loss": 0.3548,
+      "step": 149
+    },
+    {
+      "epoch": 0.6006006006006006,
+      "grad_norm": 0.37658053636550903,
+      "learning_rate": 1.8649279314629484e-05,
+      "loss": 0.3545,
+      "step": 150
+    },
+    {
+      "epoch": 0.6046046046046046,
+      "grad_norm": 0.34806281328201294,
+      "learning_rate": 1.862720538291245e-05,
+      "loss": 0.3365,
+      "step": 151
+    },
+    {
+      "epoch": 0.6086086086086087,
+      "grad_norm": 0.3958812355995178,
+      "learning_rate": 1.8604965824294253e-05,
+      "loss": 0.3682,
+      "step": 152
+    },
+    {
+      "epoch": 0.6126126126126126,
+      "grad_norm": 0.34817248582839966,
+      "learning_rate": 1.8582561065734602e-05,
+      "loss": 0.3454,
+      "step": 153
+    },
+    {
+      "epoch": 0.6166166166166166,
+      "grad_norm": 0.3551209270954132,
+      "learning_rate": 1.8559991537364767e-05,
+      "loss": 0.3466,
+      "step": 154
+    },
+    {
+      "epoch": 0.6206206206206206,
+      "grad_norm": 0.34840643405914307,
+      "learning_rate": 1.8537257672479293e-05,
+      "loss": 0.3186,
+      "step": 155
+    },
+    {
+      "epoch": 0.6246246246246246,
+      "grad_norm": 0.3974769413471222,
+      "learning_rate": 1.8514359907527693e-05,
+      "loss": 0.32,
+      "step": 156
+    },
+    {
+      "epoch": 0.6286286286286287,
+      "grad_norm": 0.36330661177635193,
+      "learning_rate": 1.8491298682106066e-05,
+      "loss": 0.3261,
+      "step": 157
+    },
+    {
+      "epoch": 0.6326326326326326,
+      "grad_norm": 0.3402544856071472,
+      "learning_rate": 1.8468074438948664e-05,
+      "loss": 0.335,
+      "step": 158
+    },
+    {
+      "epoch": 0.6366366366366366,
+      "grad_norm": 0.3627852201461792,
+      "learning_rate": 1.8444687623919388e-05,
+      "loss": 0.318,
+      "step": 159
+    },
+    {
+      "epoch": 0.6406406406406406,
+      "grad_norm": 0.3495313823223114,
+      "learning_rate": 1.842113868600322e-05,
+      "loss": 0.3235,
+      "step": 160
+    },
+    {
+      "epoch": 0.6446446446446447,
+      "grad_norm": 0.34827110171318054,
+      "learning_rate": 1.8397428077297622e-05,
+      "loss": 0.335,
+      "step": 161
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.35586410760879517,
+      "learning_rate": 1.837355625300383e-05,
+      "loss": 0.3163,
+      "step": 162
+    },
+    {
+      "epoch": 0.6526526526526526,
+      "grad_norm": 0.3451946973800659,
+      "learning_rate": 1.834952367141816e-05,
+      "loss": 0.3317,
+      "step": 163
+    },
+    {
+      "epoch": 0.6566566566566566,
+      "grad_norm": 0.36004722118377686,
+      "learning_rate": 1.8325330793923146e-05,
+      "loss": 0.3313,
+      "step": 164
+    },
+    {
+      "epoch": 0.6606606606606606,
+      "grad_norm": 0.32603421807289124,
+      "learning_rate": 1.8300978084978736e-05,
+      "loss": 0.3266,
+      "step": 165
+    },
+    {
+      "epoch": 0.6646646646646647,
+      "grad_norm": 0.33803558349609375,
+      "learning_rate": 1.8276466012113358e-05,
+      "loss": 0.3263,
+      "step": 166
+    },
+    {
+      "epoch": 0.6646646646646647,
+      "eval_loss": 0.3206555247306824,
+      "eval_runtime": 5.9746,
+      "eval_samples_per_second": 13.557,
+      "eval_steps_per_second": 1.841,
+      "step": 166
+    },
+    {
+      "epoch": 0.6686686686686687,
+      "grad_norm": 0.34881776571273804,
+      "learning_rate": 1.8251795045914922e-05,
+      "loss": 0.3575,
+      "step": 167
+    },
+    {
+      "epoch": 0.6726726726726727,
+      "grad_norm": 0.3603871464729309,
+      "learning_rate": 1.8226965660021836e-05,
+      "loss": 0.3215,
+      "step": 168
+    },
+    {
+      "epoch": 0.6766766766766766,
+      "grad_norm": 0.3549259901046753,
+      "learning_rate": 1.8201978331113855e-05,
+      "loss": 0.3302,
+      "step": 169
+    },
+    {
+      "epoch": 0.6806806806806807,
+      "grad_norm": 0.3330039381980896,
+      "learning_rate": 1.817683353890297e-05,
+      "loss": 0.339,
+      "step": 170
+    },
+    {
+      "epoch": 0.6846846846846847,
+      "grad_norm": 0.4140745997428894,
+      "learning_rate": 1.8151531766124186e-05,
+      "loss": 0.4009,
+      "step": 171
+    },
+    {
+      "epoch": 0.6886886886886887,
+      "grad_norm": 0.3429993987083435,
+      "learning_rate": 1.8126073498526254e-05,
+      "loss": 0.3203,
+      "step": 172
+    },
+    {
+      "epoch": 0.6926926926926927,
+      "grad_norm": 0.3445993661880493,
+      "learning_rate": 1.8100459224862336e-05,
+      "loss": 0.3352,
+      "step": 173
+    },
+    {
+      "epoch": 0.6966966966966966,
+      "grad_norm": 0.3043835461139679,
+      "learning_rate": 1.8074689436880643e-05,
+      "loss": 0.3294,
+      "step": 174
+    },
+    {
+      "epoch": 0.7007007007007007,
+      "grad_norm": 0.3373521566390991,
+      "learning_rate": 1.804876462931498e-05,
+      "loss": 0.3204,
+      "step": 175
+    },
+    {
+      "epoch": 0.7047047047047047,
+      "grad_norm": 0.33019447326660156,
+      "learning_rate": 1.8022685299875245e-05,
+      "loss": 0.3339,
+      "step": 176
+    },
+    {
+      "epoch": 0.7087087087087087,
+      "grad_norm": 0.32447516918182373,
+      "learning_rate": 1.799645194923788e-05,
+      "loss": 0.3216,
+      "step": 177
+    },
+    {
+      "epoch": 0.7127127127127127,
+      "grad_norm": 0.3202582597732544,
+      "learning_rate": 1.7970065081036266e-05,
+      "loss": 0.3104,
+      "step": 178
+    },
+    {
+      "epoch": 0.7167167167167167,
+      "grad_norm": 0.3486015200614929,
+      "learning_rate": 1.7943525201851038e-05,
+      "loss": 0.3307,
+      "step": 179
+    },
+    {
+      "epoch": 0.7207207207207207,
+      "grad_norm": 0.36710435152053833,
+      "learning_rate": 1.7916832821200375e-05,
+      "loss": 0.3472,
+      "step": 180
+    },
+    {
+      "epoch": 0.7247247247247247,
+      "grad_norm": 0.347162127494812,
+      "learning_rate": 1.7889988451530208e-05,
+      "loss": 0.3358,
+      "step": 181
+    },
+    {
+      "epoch": 0.7287287287287287,
+      "grad_norm": 0.3109472692012787,
+      "learning_rate": 1.7862992608204384e-05,
+      "loss": 0.2959,
+      "step": 182
+    },
+    {
+      "epoch": 0.7327327327327328,
+      "grad_norm": 0.34007692337036133,
+      "learning_rate": 1.783584580949477e-05,
+      "loss": 0.3199,
+      "step": 183
+    },
+    {
+      "epoch": 0.7367367367367368,
+      "grad_norm": 0.31094032526016235,
+      "learning_rate": 1.7808548576571314e-05,
+      "loss": 0.315,
+      "step": 184
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.367898553609848,
+      "learning_rate": 1.7781101433492026e-05,
+      "loss": 0.346,
+      "step": 185
+    },
+    {
+      "epoch": 0.7447447447447447,
+      "grad_norm": 0.3379008173942566,
+      "learning_rate": 1.7753504907192923e-05,
+      "loss": 0.3299,
+      "step": 186
+    },
+    {
+      "epoch": 0.7487487487487487,
+      "grad_norm": 0.3236868977546692,
+      "learning_rate": 1.7725759527477923e-05,
+      "loss": 0.3081,
+      "step": 187
+    },
+    {
+      "epoch": 0.7527527527527528,
+      "grad_norm": 0.34538552165031433,
+      "learning_rate": 1.769786582700864e-05,
+      "loss": 0.3259,
+      "step": 188
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 0.3106057941913605,
+      "learning_rate": 1.7669824341294203e-05,
+      "loss": 0.3123,
+      "step": 189
+    },
+    {
+      "epoch": 0.7607607607607607,
+      "grad_norm": 0.3219281733036041,
+      "learning_rate": 1.7641635608680942e-05,
+      "loss": 0.3165,
+      "step": 190
+    },
+    {
+      "epoch": 0.7647647647647647,
+      "grad_norm": 0.3259471356868744,
+      "learning_rate": 1.7613300170342073e-05,
+      "loss": 0.3453,
+      "step": 191
+    },
+    {
+      "epoch": 0.7687687687687688,
+      "grad_norm": 0.33405622839927673,
+      "learning_rate": 1.7584818570267287e-05,
+      "loss": 0.3233,
+      "step": 192
+    },
+    {
+      "epoch": 0.7727727727727728,
+      "grad_norm": 0.3420588970184326,
+      "learning_rate": 1.755619135525233e-05,
+      "loss": 0.3229,
+      "step": 193
+    },
+    {
+      "epoch": 0.7767767767767768,
+      "grad_norm": 0.35447970032691956,
+      "learning_rate": 1.7527419074888483e-05,
+      "loss": 0.3356,
+      "step": 194
+    },
+    {
+      "epoch": 0.7807807807807807,
+      "grad_norm": 0.34441742300987244,
+      "learning_rate": 1.749850228155203e-05,
+      "loss": 0.3128,
+      "step": 195
+    },
+    {
+      "epoch": 0.7847847847847848,
+      "grad_norm": 0.33647745847702026,
+      "learning_rate": 1.7469441530393652e-05,
+      "loss": 0.334,
+      "step": 196
+    },
+    {
+      "epoch": 0.7887887887887888,
+      "grad_norm": 0.3477891981601715,
+      "learning_rate": 1.7440237379327745e-05,
+      "loss": 0.3535,
+      "step": 197
+    },
+    {
+      "epoch": 0.7927927927927928,
+      "grad_norm": 0.35994237661361694,
+      "learning_rate": 1.7410890389021737e-05,
+      "loss": 0.3133,
+      "step": 198
+    },
+    {
+      "epoch": 0.7967967967967968,
+      "grad_norm": 0.33523160219192505,
+      "learning_rate": 1.7381401122885316e-05,
+      "loss": 0.3403,
+      "step": 199
+    },
+    {
+      "epoch": 0.8008008008008008,
+      "grad_norm": 0.3123041093349457,
+      "learning_rate": 1.7351770147059604e-05,
+      "loss": 0.3148,
+      "step": 200
+    },
+    {
+      "epoch": 0.8048048048048048,
+      "grad_norm": 0.30986520648002625,
+      "learning_rate": 1.7321998030406303e-05,
+      "loss": 0.3327,
+      "step": 201
+    },
+    {
+      "epoch": 0.8088088088088088,
+      "grad_norm": 0.3334694504737854,
+      "learning_rate": 1.729208534449676e-05,
+      "loss": 0.3346,
+      "step": 202
+    },
+    {
+      "epoch": 0.8128128128128128,
+      "grad_norm": 0.32513362169265747,
+      "learning_rate": 1.7262032663601003e-05,
+      "loss": 0.3137,
+      "step": 203
+    },
+    {
+      "epoch": 0.8168168168168168,
+      "grad_norm": 0.34156233072280884,
+      "learning_rate": 1.723184056467671e-05,
+      "loss": 0.3218,
+      "step": 204
+    },
+    {
+      "epoch": 0.8208208208208209,
+      "grad_norm": 0.32721707224845886,
+      "learning_rate": 1.7201509627358143e-05,
+      "loss": 0.3182,
+      "step": 205
+    },
+    {
+      "epoch": 0.8248248248248248,
+      "grad_norm": 0.35413238406181335,
+      "learning_rate": 1.7171040433945006e-05,
+      "loss": 0.2949,
+      "step": 206
+    },
+    {
+      "epoch": 0.8288288288288288,
+      "grad_norm": 0.30575960874557495,
+      "learning_rate": 1.7140433569391275e-05,
+      "loss": 0.3267,
+      "step": 207
+    },
+    {
+      "epoch": 0.8328328328328328,
+      "grad_norm": 0.3090551793575287,
+      "learning_rate": 1.710968962129396e-05,
+      "loss": 0.3072,
+      "step": 208
+    },
+    {
+      "epoch": 0.8368368368368369,
+      "grad_norm": 0.30769091844558716,
+      "learning_rate": 1.7078809179881847e-05,
+      "loss": 0.3115,
+      "step": 209
+    },
+    {
+      "epoch": 0.8408408408408409,
+      "grad_norm": 0.2806401550769806,
+      "learning_rate": 1.704779283800412e-05,
+      "loss": 0.3172,
+      "step": 210
+    },
+    {
+      "epoch": 0.8448448448448449,
+      "grad_norm": 0.322110652923584,
+      "learning_rate": 1.701664119111904e-05,
+      "loss": 0.319,
+      "step": 211
+    },
+    {
+      "epoch": 0.8488488488488488,
+      "grad_norm": 0.3080562353134155,
+      "learning_rate": 1.6985354837282462e-05,
+      "loss": 0.3247,
+      "step": 212
+    },
+    {
+      "epoch": 0.8528528528528528,
+      "grad_norm": 0.3051875829696655,
+      "learning_rate": 1.6953934377136375e-05,
+      "loss": 0.3195,
+      "step": 213
+    },
+    {
+      "epoch": 0.8568568568568569,
+      "grad_norm": 0.31982743740081787,
+      "learning_rate": 1.6922380413897382e-05,
+      "loss": 0.3202,
+      "step": 214
+    },
+    {
+      "epoch": 0.8608608608608609,
+      "grad_norm": 0.3235834538936615,
+      "learning_rate": 1.689069355334509e-05,
+      "loss": 0.3532,
+      "step": 215
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.32920244336128235,
+      "learning_rate": 1.6858874403810507e-05,
+      "loss": 0.3412,
+      "step": 216
+    },
+    {
+      "epoch": 0.8688688688688688,
+      "grad_norm": 0.3245142102241516,
+      "learning_rate": 1.682692357616435e-05,
+      "loss": 0.3074,
+      "step": 217
+    },
+    {
+      "epoch": 0.8728728728728729,
+      "grad_norm": 0.30122387409210205,
+      "learning_rate": 1.679484168380532e-05,
+      "loss": 0.3166,
+      "step": 218
+    },
+    {
+      "epoch": 0.8768768768768769,
+      "grad_norm": 0.30446872115135193,
+      "learning_rate": 1.676262934264832e-05,
+      "loss": 0.3177,
+      "step": 219
+    },
+    {
+      "epoch": 0.8808808808808809,
+      "grad_norm": 0.3006502091884613,
+      "learning_rate": 1.6730287171112652e-05,
+      "loss": 0.3225,
+      "step": 220
+    },
+    {
+      "epoch": 0.8848848848848849,
+      "grad_norm": 0.33892446756362915,
+      "learning_rate": 1.669781579011011e-05,
+      "loss": 0.335,
+      "step": 221
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.34253889322280884,
+      "learning_rate": 1.666521582303309e-05,
+      "loss": 0.3329,
+      "step": 222
+    },
+    {
+      "epoch": 0.8928928928928929,
+      "grad_norm": 0.291759729385376,
+      "learning_rate": 1.6632487895742612e-05,
+      "loss": 0.3173,
+      "step": 223
+    },
+    {
+      "epoch": 0.8968968968968969,
+      "grad_norm": 0.3063089847564697,
+      "learning_rate": 1.6599632636556292e-05,
+      "loss": 0.3345,
+      "step": 224
+    },
+    {
+      "epoch": 0.9009009009009009,
+      "grad_norm": 0.30597245693206787,
+      "learning_rate": 1.6566650676236307e-05,
+      "loss": 0.3435,
+      "step": 225
+    },
+    {
+      "epoch": 0.9049049049049049,
+      "grad_norm": 0.3654542863368988,
+      "learning_rate": 1.653354264797725e-05,
+      "loss": 0.32,
+      "step": 226
+    },
+    {
+      "epoch": 0.908908908908909,
+      "grad_norm": 0.30520570278167725,
+      "learning_rate": 1.6500309187394005e-05,
+      "loss": 0.3099,
+      "step": 227
+    },
+    {
+      "epoch": 0.9129129129129129,
+      "grad_norm": 0.3210243582725525,
+      "learning_rate": 1.6466950932509532e-05,
+      "loss": 0.316,
+      "step": 228
+    },
+    {
+      "epoch": 0.9169169169169169,
+      "grad_norm": 0.316346138715744,
+      "learning_rate": 1.643346852374261e-05,
+      "loss": 0.2977,
+      "step": 229
+    },
+    {
+      "epoch": 0.9209209209209209,
+      "grad_norm": 0.3127054274082184,
+      "learning_rate": 1.6399862603895563e-05,
+      "loss": 0.2942,
+      "step": 230
+    },
+    {
+      "epoch": 0.924924924924925,
+      "grad_norm": 0.3078126013278961,
+      "learning_rate": 1.6366133818141893e-05,
+      "loss": 0.3008,
+      "step": 231
+    },
+    {
+      "epoch": 0.928928928928929,
+      "grad_norm": 0.3247736096382141,
+      "learning_rate": 1.633228281401392e-05,
+      "loss": 0.3142,
+      "step": 232
+    },
+    {
+      "epoch": 0.9329329329329329,
+      "grad_norm": 0.3431578576564789,
+      "learning_rate": 1.6298310241390326e-05,
+      "loss": 0.3093,
+      "step": 233
+    },
+    {
+      "epoch": 0.9369369369369369,
+      "grad_norm": 0.31279513239860535,
+      "learning_rate": 1.6264216752483697e-05,
+      "loss": 0.3175,
+      "step": 234
+    },
+    {
+      "epoch": 0.9409409409409409,
+      "grad_norm": 0.33988532423973083,
+      "learning_rate": 1.6230003001828e-05,
+      "loss": 0.324,
+      "step": 235
+    },
+    {
+      "epoch": 0.944944944944945,
+      "grad_norm": 0.3213682472705841,
+      "learning_rate": 1.6195669646266003e-05,
+      "loss": 0.3321,
+      "step": 236
+    },
+    {
+      "epoch": 0.948948948948949,
+      "grad_norm": 0.357149213552475,
+      "learning_rate": 1.616121734493668e-05,
+      "loss": 0.3342,
+      "step": 237
+    },
+    {
+      "epoch": 0.9529529529529529,
+      "grad_norm": 0.31976473331451416,
+      "learning_rate": 1.6126646759262548e-05,
+      "loss": 0.3181,
+      "step": 238
+    },
+    {
+      "epoch": 0.9569569569569569,
+      "grad_norm": 0.328274130821228,
+      "learning_rate": 1.609195855293697e-05,
+      "loss": 0.3161,
+      "step": 239
+    },
+    {
+      "epoch": 0.960960960960961,
+      "grad_norm": 0.3303963840007782,
+      "learning_rate": 1.6057153391911422e-05,
+      "loss": 0.3076,
+      "step": 240
+    },
+    {
+      "epoch": 0.964964964964965,
+      "grad_norm": 0.34692683815956116,
+      "learning_rate": 1.6022231944382693e-05,
+      "loss": 0.3351,
+      "step": 241
+    },
+    {
+      "epoch": 0.968968968968969,
+      "grad_norm": 0.3468174636363983,
+      "learning_rate": 1.598719488078007e-05,
+      "loss": 0.3224,
+      "step": 242
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.3575330972671509,
+      "learning_rate": 1.5952042873752463e-05,
+      "loss": 0.3189,
+      "step": 243
+    },
+    {
+      "epoch": 0.9769769769769769,
+      "grad_norm": 0.35199496150016785,
+      "learning_rate": 1.5916776598155478e-05,
+      "loss": 0.3515,
+      "step": 244
+    },
+    {
+      "epoch": 0.980980980980981,
+      "grad_norm": 0.34917882084846497,
+      "learning_rate": 1.5881396731038493e-05,
+      "loss": 0.3354,
+      "step": 245
+    },
+    {
+      "epoch": 0.984984984984985,
+      "grad_norm": 0.32909733057022095,
+      "learning_rate": 1.584590395163162e-05,
+      "loss": 0.3009,
+      "step": 246
+    },
+    {
+      "epoch": 0.988988988988989,
+      "grad_norm": 0.3109247088432312,
+      "learning_rate": 1.5810298941332696e-05,
+      "loss": 0.3164,
+      "step": 247
+    },
+    {
+      "epoch": 0.992992992992993,
+      "grad_norm": 0.30618447065353394,
+      "learning_rate": 1.5774582383694196e-05,
+      "loss": 0.2923,
+      "step": 248
+    },
+    {
+      "epoch": 0.996996996996997,
+      "grad_norm": 0.32330596446990967,
+      "learning_rate": 1.5738754964410084e-05,
+      "loss": 0.3213,
+      "step": 249
+    },
+    {
+      "epoch": 0.996996996996997,
+      "eval_loss": 0.304058700799942,
+      "eval_runtime": 6.1571,
+      "eval_samples_per_second": 13.156,
+      "eval_steps_per_second": 1.787,
+      "step": 249
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 747,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 249,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.146380399234253e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

3b-w-cot/checkpoint-249/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ae10bafaded3f1f05741f3f17290afb1efc74a263062c027a77525fa9902f1e
+size 10744

3b-w-cot/checkpoint-249/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

3b-w-cot/checkpoint-498/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
+}

3b-w-cot/checkpoint-498/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-3B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.3",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

3b-w-cot/checkpoint-498/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.48.3"
+}

3b-w-cot/checkpoint-498/latest

@@ -0,0 +1 @@
+global_step497

3b-w-cot/checkpoint-498/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a482673aeb9035cfc2e8a145055294d02953cff436fcd5b63164f9b40fc9adf
+size 4957560304

3b-w-cot/checkpoint-498/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e832a210ed1cb3d1081c2621ce9afe15bc8f73253d8d03114ccdfd08679c29cd
+size 1836696752

3b-w-cot/checkpoint-498/model.safetensors.index.json

@@ -0,0 +1,442 @@
+{
+  "metadata": {
+    "total_size": 6794207232
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

3b-w-cot/checkpoint-498/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91bd7f619e4cd37883f469c08e90105c4d218fd82ffc43ae58fa9fdbcc37fce5
+size 14512

3b-w-cot/checkpoint-498/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b0a7593f9ab52bf47328c6d50954dce1fcd69866aa6f5f35851aef7f7af3899
+size 14512

3b-w-cot/checkpoint-498/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4826800cd376d53466d2abcda597b68ee000b75ad68cf3b3811bc19eb99665f
+size 1064

3b-w-cot/checkpoint-498/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": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

3b-w-cot/checkpoint-498/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

3b-w-cot/checkpoint-498/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,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

3b-w-cot/checkpoint-498/trainer_state.json

@@ -0,0 +1,3575 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.992992992992993,
+  "eval_steps": 83,
+  "global_step": 498,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004004004004004004,
+      "grad_norm": 6.739165782928467,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.4163,
+      "step": 1
+    },
+    {
+      "epoch": 0.004004004004004004,
+      "eval_loss": 1.4217944145202637,
+      "eval_runtime": 5.678,
+      "eval_samples_per_second": 14.266,
+      "eval_steps_per_second": 1.937,
+      "step": 1
+    },
+    {
+      "epoch": 0.008008008008008008,
+      "grad_norm": 6.897192001342773,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.3767,
+      "step": 2
+    },
+    {
+      "epoch": 0.012012012012012012,
+      "grad_norm": 6.74431037902832,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.3848,
+      "step": 3
+    },
+    {
+      "epoch": 0.016016016016016016,
+      "grad_norm": 6.999237537384033,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.4294,
+      "step": 4
+    },
+    {
+      "epoch": 0.02002002002002002,
+      "grad_norm": 4.510103702545166,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.2646,
+      "step": 5
+    },
+    {
+      "epoch": 0.024024024024024024,
+      "grad_norm": 4.533900737762451,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.2563,
+      "step": 6
+    },
+    {
+      "epoch": 0.028028028028028028,
+      "grad_norm": 3.565216541290283,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.1061,
+      "step": 7
+    },
+    {
+      "epoch": 0.03203203203203203,
+      "grad_norm": 3.1937592029571533,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.0445,
+      "step": 8
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 3.0942018032073975,
+      "learning_rate": 6e-06,
+      "loss": 0.8555,
+      "step": 9
+    },
+    {
+      "epoch": 0.04004004004004004,
+      "grad_norm": 4.628591060638428,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8539,
+      "step": 10
+    },
+    {
+      "epoch": 0.044044044044044044,
+      "grad_norm": 5.402413845062256,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.7684,
+      "step": 11
+    },
+    {
+      "epoch": 0.04804804804804805,
+      "grad_norm": 1.96303391456604,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6347,
+      "step": 12
+    },
+    {
+      "epoch": 0.05205205205205205,
+      "grad_norm": 1.33661687374115,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.6692,
+      "step": 13
+    },
+    {
+      "epoch": 0.056056056056056056,
+      "grad_norm": 0.9704915285110474,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.6385,
+      "step": 14
+    },
+    {
+      "epoch": 0.06006006006006006,
+      "grad_norm": 0.7543495893478394,
+      "learning_rate": 1e-05,
+      "loss": 0.5901,
+      "step": 15
+    },
+    {
+      "epoch": 0.06406406406406406,
+      "grad_norm": 1.3593250513076782,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.5645,
+      "step": 16
+    },
+    {
+      "epoch": 0.06806806806806807,
+      "grad_norm": 1.147750735282898,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 0.5508,
+      "step": 17
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 0.8024477958679199,
+      "learning_rate": 1.2e-05,
+      "loss": 0.5282,
+      "step": 18
+    },
+    {
+      "epoch": 0.07607607607607608,
+      "grad_norm": 0.7931386232376099,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 0.4929,
+      "step": 19
+    },
+    {
+      "epoch": 0.08008008008008008,
+      "grad_norm": 0.6702373623847961,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4868,
+      "step": 20
+    },
+    {
+      "epoch": 0.08408408408408409,
+      "grad_norm": 0.747148871421814,
+      "learning_rate": 1.4e-05,
+      "loss": 0.4857,
+      "step": 21
+    },
+    {
+      "epoch": 0.08808808808808809,
+      "grad_norm": 0.5935061573982239,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 0.4908,
+      "step": 22
+    },
+    {
+      "epoch": 0.0920920920920921,
+      "grad_norm": 0.5696635842323303,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.4437,
+      "step": 23
+    },
+    {
+      "epoch": 0.0960960960960961,
+      "grad_norm": 0.5861401557922363,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.4442,
+      "step": 24
+    },
+    {
+      "epoch": 0.1001001001001001,
+      "grad_norm": 0.5833226442337036,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.4585,
+      "step": 25
+    },
+    {
+      "epoch": 0.1041041041041041,
+      "grad_norm": 0.48214447498321533,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.3994,
+      "step": 26
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 0.49264758825302124,
+      "learning_rate": 1.8e-05,
+      "loss": 0.4609,
+      "step": 27
+    },
+    {
+      "epoch": 0.11211211211211211,
+      "grad_norm": 0.5025641322135925,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.4816,
+      "step": 28
+    },
+    {
+      "epoch": 0.11611611611611612,
+      "grad_norm": 0.46342933177948,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 0.4435,
+      "step": 29
+    },
+    {
+      "epoch": 0.12012012012012012,
+      "grad_norm": 0.46523571014404297,
+      "learning_rate": 2e-05,
+      "loss": 0.4208,
+      "step": 30
+    },
+    {
+      "epoch": 0.12412412412412413,
+      "grad_norm": 0.4298263192176819,
+      "learning_rate": 1.9999904008949705e-05,
+      "loss": 0.4137,
+      "step": 31
+    },
+    {
+      "epoch": 0.12812812812812813,
+      "grad_norm": 0.43352752923965454,
+      "learning_rate": 1.999961603764167e-05,
+      "loss": 0.4096,
+      "step": 32
+    },
+    {
+      "epoch": 0.13213213213213212,
+      "grad_norm": 0.4369907081127167,
+      "learning_rate": 1.9999136091604433e-05,
+      "loss": 0.4128,
+      "step": 33
+    },
+    {
+      "epoch": 0.13613613613613615,
+      "grad_norm": 0.4188078045845032,
+      "learning_rate": 1.99984641800521e-05,
+      "loss": 0.4171,
+      "step": 34
+    },
+    {
+      "epoch": 0.14014014014014015,
+      "grad_norm": 0.4548235237598419,
+      "learning_rate": 1.9997600315884166e-05,
+      "loss": 0.4236,
+      "step": 35
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 0.4751495122909546,
+      "learning_rate": 1.999654451568528e-05,
+      "loss": 0.438,
+      "step": 36
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 0.4333605170249939,
+      "learning_rate": 1.9995296799724914e-05,
+      "loss": 0.4208,
+      "step": 37
+    },
+    {
+      "epoch": 0.15215215215215216,
+      "grad_norm": 0.48084449768066406,
+      "learning_rate": 1.999385719195698e-05,
+      "loss": 0.3969,
+      "step": 38
+    },
+    {
+      "epoch": 0.15615615615615616,
+      "grad_norm": 0.42627257108688354,
+      "learning_rate": 1.9992225720019377e-05,
+      "loss": 0.4503,
+      "step": 39
+    },
+    {
+      "epoch": 0.16016016016016016,
+      "grad_norm": 0.41883718967437744,
+      "learning_rate": 1.9990402415233436e-05,
+      "loss": 0.4322,
+      "step": 40
+    },
+    {
+      "epoch": 0.16416416416416416,
+      "grad_norm": 0.39587706327438354,
+      "learning_rate": 1.998838731260335e-05,
+      "loss": 0.3841,
+      "step": 41
+    },
+    {
+      "epoch": 0.16816816816816818,
+      "grad_norm": 0.40334609150886536,
+      "learning_rate": 1.9986180450815485e-05,
+      "loss": 0.3737,
+      "step": 42
+    },
+    {
+      "epoch": 0.17217217217217218,
+      "grad_norm": 0.3715899884700775,
+      "learning_rate": 1.9983781872237634e-05,
+      "loss": 0.387,
+      "step": 43
+    },
+    {
+      "epoch": 0.17617617617617617,
+      "grad_norm": 0.38875311613082886,
+      "learning_rate": 1.9981191622918217e-05,
+      "loss": 0.3695,
+      "step": 44
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 0.39087679982185364,
+      "learning_rate": 1.997840975258538e-05,
+      "loss": 0.3934,
+      "step": 45
+    },
+    {
+      "epoch": 0.1841841841841842,
+      "grad_norm": 0.41504260897636414,
+      "learning_rate": 1.9975436314646052e-05,
+      "loss": 0.3769,
+      "step": 46
+    },
+    {
+      "epoch": 0.1881881881881882,
+      "grad_norm": 0.36994805932044983,
+      "learning_rate": 1.9972271366184922e-05,
+      "loss": 0.3807,
+      "step": 47
+    },
+    {
+      "epoch": 0.1921921921921922,
+      "grad_norm": 0.39996904134750366,
+      "learning_rate": 1.996891496796334e-05,
+      "loss": 0.3903,
+      "step": 48
+    },
+    {
+      "epoch": 0.1961961961961962,
+      "grad_norm": 0.3904661536216736,
+      "learning_rate": 1.9965367184418138e-05,
+      "loss": 0.3785,
+      "step": 49
+    },
+    {
+      "epoch": 0.2002002002002002,
+      "grad_norm": 0.39596742391586304,
+      "learning_rate": 1.9961628083660406e-05,
+      "loss": 0.3806,
+      "step": 50
+    },
+    {
+      "epoch": 0.2042042042042042,
+      "grad_norm": 0.39424648880958557,
+      "learning_rate": 1.9957697737474198e-05,
+      "loss": 0.3736,
+      "step": 51
+    },
+    {
+      "epoch": 0.2082082082082082,
+      "grad_norm": 0.41719183325767517,
+      "learning_rate": 1.9953576221315116e-05,
+      "loss": 0.4023,
+      "step": 52
+    },
+    {
+      "epoch": 0.2122122122122122,
+      "grad_norm": 0.3773253262042999,
+      "learning_rate": 1.9949263614308894e-05,
+      "loss": 0.396,
+      "step": 53
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.4236510992050171,
+      "learning_rate": 1.994475999924987e-05,
+      "loss": 0.4032,
+      "step": 54
+    },
+    {
+      "epoch": 0.22022022022022023,
+      "grad_norm": 0.3507007360458374,
+      "learning_rate": 1.9940065462599394e-05,
+      "loss": 0.3615,
+      "step": 55
+    },
+    {
+      "epoch": 0.22422422422422422,
+      "grad_norm": 0.45851731300354004,
+      "learning_rate": 1.9935180094484164e-05,
+      "loss": 0.3402,
+      "step": 56
+    },
+    {
+      "epoch": 0.22822822822822822,
+      "grad_norm": 0.36239954829216003,
+      "learning_rate": 1.99301039886945e-05,
+      "loss": 0.3668,
+      "step": 57
+    },
+    {
+      "epoch": 0.23223223223223224,
+      "grad_norm": 0.3855980634689331,
+      "learning_rate": 1.992483724268255e-05,
+      "loss": 0.3575,
+      "step": 58
+    },
+    {
+      "epoch": 0.23623623623623624,
+      "grad_norm": 0.4716965854167938,
+      "learning_rate": 1.9919379957560413e-05,
+      "loss": 0.3532,
+      "step": 59
+    },
+    {
+      "epoch": 0.24024024024024024,
+      "grad_norm": 0.45167747139930725,
+      "learning_rate": 1.991373223809819e-05,
+      "loss": 0.3537,
+      "step": 60
+    },
+    {
+      "epoch": 0.24424424424424424,
+      "grad_norm": 0.420512855052948,
+      "learning_rate": 1.990789419272199e-05,
+      "loss": 0.3471,
+      "step": 61
+    },
+    {
+      "epoch": 0.24824824824824826,
+      "grad_norm": 0.4863167107105255,
+      "learning_rate": 1.9901865933511834e-05,
+      "loss": 0.3741,
+      "step": 62
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 0.39653480052948,
+      "learning_rate": 1.9895647576199507e-05,
+      "loss": 0.3799,
+      "step": 63
+    },
+    {
+      "epoch": 0.25625625625625625,
+      "grad_norm": 0.38418489694595337,
+      "learning_rate": 1.988923924016634e-05,
+      "loss": 0.3577,
+      "step": 64
+    },
+    {
+      "epoch": 0.2602602602602603,
+      "grad_norm": 0.39798790216445923,
+      "learning_rate": 1.988264104844091e-05,
+      "loss": 0.3581,
+      "step": 65
+    },
+    {
+      "epoch": 0.26426426426426425,
+      "grad_norm": 0.41135266423225403,
+      "learning_rate": 1.987585312769669e-05,
+      "loss": 0.3306,
+      "step": 66
+    },
+    {
+      "epoch": 0.2682682682682683,
+      "grad_norm": 0.36515921354293823,
+      "learning_rate": 1.9868875608249613e-05,
+      "loss": 0.3628,
+      "step": 67
+    },
+    {
+      "epoch": 0.2722722722722723,
+      "grad_norm": 0.4733300507068634,
+      "learning_rate": 1.986170862405556e-05,
+      "loss": 0.3733,
+      "step": 68
+    },
+    {
+      "epoch": 0.27627627627627627,
+      "grad_norm": 0.4218703806400299,
+      "learning_rate": 1.98543523127078e-05,
+      "loss": 0.371,
+      "step": 69
+    },
+    {
+      "epoch": 0.2802802802802803,
+      "grad_norm": 0.35307836532592773,
+      "learning_rate": 1.984680681543434e-05,
+      "loss": 0.3568,
+      "step": 70
+    },
+    {
+      "epoch": 0.28428428428428426,
+      "grad_norm": 0.4475674629211426,
+      "learning_rate": 1.9839072277095222e-05,
+      "loss": 0.3828,
+      "step": 71
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 0.37015393376350403,
+      "learning_rate": 1.9831148846179743e-05,
+      "loss": 0.3426,
+      "step": 72
+    },
+    {
+      "epoch": 0.2922922922922923,
+      "grad_norm": 0.39532360434532166,
+      "learning_rate": 1.9823036674803585e-05,
+      "loss": 0.3545,
+      "step": 73
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 0.349669486284256,
+      "learning_rate": 1.981473591870593e-05,
+      "loss": 0.332,
+      "step": 74
+    },
+    {
+      "epoch": 0.3003003003003003,
+      "grad_norm": 0.356000155210495,
+      "learning_rate": 1.980624673724643e-05,
+      "loss": 0.3684,
+      "step": 75
+    },
+    {
+      "epoch": 0.30430430430430433,
+      "grad_norm": 0.4027644693851471,
+      "learning_rate": 1.9797569293402174e-05,
+      "loss": 0.3592,
+      "step": 76
+    },
+    {
+      "epoch": 0.3083083083083083,
+      "grad_norm": 0.36116528511047363,
+      "learning_rate": 1.9788703753764554e-05,
+      "loss": 0.3433,
+      "step": 77
+    },
+    {
+      "epoch": 0.3123123123123123,
+      "grad_norm": 0.35624876618385315,
+      "learning_rate": 1.9779650288536057e-05,
+      "loss": 0.3541,
+      "step": 78
+    },
+    {
+      "epoch": 0.3163163163163163,
+      "grad_norm": 0.360287606716156,
+      "learning_rate": 1.977040907152702e-05,
+      "loss": 0.3413,
+      "step": 79
+    },
+    {
+      "epoch": 0.3203203203203203,
+      "grad_norm": 0.412913978099823,
+      "learning_rate": 1.976098028015226e-05,
+      "loss": 0.4142,
+      "step": 80
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.38022932410240173,
+      "learning_rate": 1.9751364095427694e-05,
+      "loss": 0.3481,
+      "step": 81
+    },
+    {
+      "epoch": 0.3283283283283283,
+      "grad_norm": 0.37813299894332886,
+      "learning_rate": 1.974156070196686e-05,
+      "loss": 0.3552,
+      "step": 82
+    },
+    {
+      "epoch": 0.33233233233233234,
+      "grad_norm": 0.3782276511192322,
+      "learning_rate": 1.973157028797737e-05,
+      "loss": 0.3799,
+      "step": 83
+    },
+    {
+      "epoch": 0.33233233233233234,
+      "eval_loss": 0.33762940764427185,
+      "eval_runtime": 6.2209,
+      "eval_samples_per_second": 13.021,
+      "eval_steps_per_second": 1.768,
+      "step": 83
+    },
+    {
+      "epoch": 0.33633633633633636,
+      "grad_norm": 0.3933849036693573,
+      "learning_rate": 1.9721393045257277e-05,
+      "loss": 0.3654,
+      "step": 84
+    },
+    {
+      "epoch": 0.34034034034034033,
+      "grad_norm": 0.35808295011520386,
+      "learning_rate": 1.9711029169191437e-05,
+      "loss": 0.3716,
+      "step": 85
+    },
+    {
+      "epoch": 0.34434434434434436,
+      "grad_norm": 0.3547224700450897,
+      "learning_rate": 1.970047885874771e-05,
+      "loss": 0.3397,
+      "step": 86
+    },
+    {
+      "epoch": 0.3483483483483483,
+      "grad_norm": 0.39883747696876526,
+      "learning_rate": 1.968974231647318e-05,
+      "loss": 0.3645,
+      "step": 87
+    },
+    {
+      "epoch": 0.35235235235235235,
+      "grad_norm": 0.3208443224430084,
+      "learning_rate": 1.9678819748490236e-05,
+      "loss": 0.3431,
+      "step": 88
+    },
+    {
+      "epoch": 0.3563563563563564,
+      "grad_norm": 0.34669625759124756,
+      "learning_rate": 1.9667711364492638e-05,
+      "loss": 0.3613,
+      "step": 89
+    },
+    {
+      "epoch": 0.36036036036036034,
+      "grad_norm": 0.35220351815223694,
+      "learning_rate": 1.965641737774147e-05,
+      "loss": 0.3499,
+      "step": 90
+    },
+    {
+      "epoch": 0.36436436436436437,
+      "grad_norm": 0.3566596806049347,
+      "learning_rate": 1.9644938005061062e-05,
+      "loss": 0.3204,
+      "step": 91
+    },
+    {
+      "epoch": 0.3683683683683684,
+      "grad_norm": 0.3813494145870209,
+      "learning_rate": 1.9633273466834826e-05,
+      "loss": 0.3526,
+      "step": 92
+    },
+    {
+      "epoch": 0.37237237237237236,
+      "grad_norm": 0.3885157108306885,
+      "learning_rate": 1.9621423987001013e-05,
+      "loss": 0.3562,
+      "step": 93
+    },
+    {
+      "epoch": 0.3763763763763764,
+      "grad_norm": 0.4746558666229248,
+      "learning_rate": 1.960938979304843e-05,
+      "loss": 0.3509,
+      "step": 94
+    },
+    {
+      "epoch": 0.38038038038038036,
+      "grad_norm": 0.3623756766319275,
+      "learning_rate": 1.959717111601206e-05,
+      "loss": 0.371,
+      "step": 95
+    },
+    {
+      "epoch": 0.3843843843843844,
+      "grad_norm": 0.3858593702316284,
+      "learning_rate": 1.9584768190468624e-05,
+      "loss": 0.3551,
+      "step": 96
+    },
+    {
+      "epoch": 0.3883883883883884,
+      "grad_norm": 0.3788565993309021,
+      "learning_rate": 1.95721812545321e-05,
+      "loss": 0.3669,
+      "step": 97
+    },
+    {
+      "epoch": 0.3923923923923924,
+      "grad_norm": 0.3729216754436493,
+      "learning_rate": 1.9559410549849125e-05,
+      "loss": 0.34,
+      "step": 98
+    },
+    {
+      "epoch": 0.3963963963963964,
+      "grad_norm": 0.44008028507232666,
+      "learning_rate": 1.9546456321594374e-05,
+      "loss": 0.3898,
+      "step": 99
+    },
+    {
+      "epoch": 0.4004004004004004,
+      "grad_norm": 0.36622726917266846,
+      "learning_rate": 1.9533318818465837e-05,
+      "loss": 0.3624,
+      "step": 100
+    },
+    {
+      "epoch": 0.4044044044044044,
+      "grad_norm": 0.39071497321128845,
+      "learning_rate": 1.9519998292680062e-05,
+      "loss": 0.3518,
+      "step": 101
+    },
+    {
+      "epoch": 0.4084084084084084,
+      "grad_norm": 0.40084153413772583,
+      "learning_rate": 1.9506494999967298e-05,
+      "loss": 0.3483,
+      "step": 102
+    },
+    {
+      "epoch": 0.4124124124124124,
+      "grad_norm": 0.3901435434818268,
+      "learning_rate": 1.94928091995666e-05,
+      "loss": 0.3577,
+      "step": 103
+    },
+    {
+      "epoch": 0.4164164164164164,
+      "grad_norm": 0.34764233231544495,
+      "learning_rate": 1.9478941154220833e-05,
+      "loss": 0.3487,
+      "step": 104
+    },
+    {
+      "epoch": 0.42042042042042044,
+      "grad_norm": 0.4044027030467987,
+      "learning_rate": 1.9464891130171647e-05,
+      "loss": 0.3602,
+      "step": 105
+    },
+    {
+      "epoch": 0.4244244244244244,
+      "grad_norm": 0.3537923991680145,
+      "learning_rate": 1.9450659397154353e-05,
+      "loss": 0.3282,
+      "step": 106
+    },
+    {
+      "epoch": 0.42842842842842843,
+      "grad_norm": 0.35585689544677734,
+      "learning_rate": 1.9436246228392762e-05,
+      "loss": 0.361,
+      "step": 107
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.380585253238678,
+      "learning_rate": 1.94216519005939e-05,
+      "loss": 0.3218,
+      "step": 108
+    },
+    {
+      "epoch": 0.4364364364364364,
+      "grad_norm": 0.3388879597187042,
+      "learning_rate": 1.9406876693942747e-05,
+      "loss": 0.3563,
+      "step": 109
+    },
+    {
+      "epoch": 0.44044044044044045,
+      "grad_norm": 0.359130859375,
+      "learning_rate": 1.939192089209682e-05,
+      "loss": 0.3386,
+      "step": 110
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.35704970359802246,
+      "learning_rate": 1.9376784782180747e-05,
+      "loss": 0.3336,
+      "step": 111
+    },
+    {
+      "epoch": 0.44844844844844844,
+      "grad_norm": 0.3638094961643219,
+      "learning_rate": 1.9361468654780748e-05,
+      "loss": 0.3582,
+      "step": 112
+    },
+    {
+      "epoch": 0.45245245245245247,
+      "grad_norm": 0.3996574878692627,
+      "learning_rate": 1.9345972803939046e-05,
+      "loss": 0.3496,
+      "step": 113
+    },
+    {
+      "epoch": 0.45645645645645644,
+      "grad_norm": 0.3565851151943207,
+      "learning_rate": 1.9330297527148246e-05,
+      "loss": 0.3273,
+      "step": 114
+    },
+    {
+      "epoch": 0.46046046046046046,
+      "grad_norm": 0.3502778112888336,
+      "learning_rate": 1.9314443125345606e-05,
+      "loss": 0.3331,
+      "step": 115
+    },
+    {
+      "epoch": 0.4644644644644645,
+      "grad_norm": 0.4017852544784546,
+      "learning_rate": 1.929840990290726e-05,
+      "loss": 0.3522,
+      "step": 116
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "grad_norm": 0.354686975479126,
+      "learning_rate": 1.928219816764238e-05,
+      "loss": 0.3456,
+      "step": 117
+    },
+    {
+      "epoch": 0.4724724724724725,
+      "grad_norm": 0.34257611632347107,
+      "learning_rate": 1.9265808230787265e-05,
+      "loss": 0.3325,
+      "step": 118
+    },
+    {
+      "epoch": 0.47647647647647645,
+      "grad_norm": 0.3504616320133209,
+      "learning_rate": 1.9249240406999366e-05,
+      "loss": 0.3516,
+      "step": 119
+    },
+    {
+      "epoch": 0.4804804804804805,
+      "grad_norm": 0.36517202854156494,
+      "learning_rate": 1.9232495014351248e-05,
+      "loss": 0.3233,
+      "step": 120
+    },
+    {
+      "epoch": 0.4844844844844845,
+      "grad_norm": 0.33805495500564575,
+      "learning_rate": 1.921557237432447e-05,
+      "loss": 0.3261,
+      "step": 121
+    },
+    {
+      "epoch": 0.48848848848848847,
+      "grad_norm": 0.35013893246650696,
+      "learning_rate": 1.919847281180343e-05,
+      "loss": 0.3361,
+      "step": 122
+    },
+    {
+      "epoch": 0.4924924924924925,
+      "grad_norm": 0.3569040298461914,
+      "learning_rate": 1.9181196655069126e-05,
+      "loss": 0.33,
+      "step": 123
+    },
+    {
+      "epoch": 0.4964964964964965,
+      "grad_norm": 0.3355570137500763,
+      "learning_rate": 1.9163744235792845e-05,
+      "loss": 0.3263,
+      "step": 124
+    },
+    {
+      "epoch": 0.5005005005005005,
+      "grad_norm": 0.39212533831596375,
+      "learning_rate": 1.9146115889029793e-05,
+      "loss": 0.3671,
+      "step": 125
+    },
+    {
+      "epoch": 0.5045045045045045,
+      "grad_norm": 0.3458411395549774,
+      "learning_rate": 1.912831195321268e-05,
+      "loss": 0.3575,
+      "step": 126
+    },
+    {
+      "epoch": 0.5085085085085085,
+      "grad_norm": 0.3606933057308197,
+      "learning_rate": 1.9110332770145198e-05,
+      "loss": 0.342,
+      "step": 127
+    },
+    {
+      "epoch": 0.5125125125125125,
+      "grad_norm": 0.3897223472595215,
+      "learning_rate": 1.9092178684995487e-05,
+      "loss": 0.37,
+      "step": 128
+    },
+    {
+      "epoch": 0.5165165165165165,
+      "grad_norm": 0.30565860867500305,
+      "learning_rate": 1.9073850046289484e-05,
+      "loss": 0.3331,
+      "step": 129
+    },
+    {
+      "epoch": 0.5205205205205206,
+      "grad_norm": 0.3087623119354248,
+      "learning_rate": 1.9055347205904245e-05,
+      "loss": 0.322,
+      "step": 130
+    },
+    {
+      "epoch": 0.5245245245245245,
+      "grad_norm": 0.34453681111335754,
+      "learning_rate": 1.903667051906119e-05,
+      "loss": 0.3405,
+      "step": 131
+    },
+    {
+      "epoch": 0.5285285285285285,
+      "grad_norm": 0.32023486495018005,
+      "learning_rate": 1.901782034431927e-05,
+      "loss": 0.338,
+      "step": 132
+    },
+    {
+      "epoch": 0.5325325325325325,
+      "grad_norm": 0.37953877449035645,
+      "learning_rate": 1.8998797043568102e-05,
+      "loss": 0.3406,
+      "step": 133
+    },
+    {
+      "epoch": 0.5365365365365365,
+      "grad_norm": 0.3680444061756134,
+      "learning_rate": 1.8979600982021014e-05,
+      "loss": 0.324,
+      "step": 134
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.3368768095970154,
+      "learning_rate": 1.896023252820802e-05,
+      "loss": 0.3194,
+      "step": 135
+    },
+    {
+      "epoch": 0.5445445445445446,
+      "grad_norm": 0.37519025802612305,
+      "learning_rate": 1.8940692053968773e-05,
+      "loss": 0.3358,
+      "step": 136
+    },
+    {
+      "epoch": 0.5485485485485485,
+      "grad_norm": 0.33083492517471313,
+      "learning_rate": 1.89209799344454e-05,
+      "loss": 0.3202,
+      "step": 137
+    },
+    {
+      "epoch": 0.5525525525525525,
+      "grad_norm": 0.3258178234100342,
+      "learning_rate": 1.8901096548075305e-05,
+      "loss": 0.3186,
+      "step": 138
+    },
+    {
+      "epoch": 0.5565565565565566,
+      "grad_norm": 0.37846943736076355,
+      "learning_rate": 1.8881042276583924e-05,
+      "loss": 0.3563,
+      "step": 139
+    },
+    {
+      "epoch": 0.5605605605605606,
+      "grad_norm": 0.35151371359825134,
+      "learning_rate": 1.8860817504977374e-05,
+      "loss": 0.337,
+      "step": 140
+    },
+    {
+      "epoch": 0.5645645645645646,
+      "grad_norm": 0.33033043146133423,
+      "learning_rate": 1.8840422621535067e-05,
+      "loss": 0.3042,
+      "step": 141
+    },
+    {
+      "epoch": 0.5685685685685685,
+      "grad_norm": 0.3724791705608368,
+      "learning_rate": 1.881985801780225e-05,
+      "loss": 0.3165,
+      "step": 142
+    },
+    {
+      "epoch": 0.5725725725725725,
+      "grad_norm": 0.355752170085907,
+      "learning_rate": 1.8799124088582523e-05,
+      "loss": 0.3693,
+      "step": 143
+    },
+    {
+      "epoch": 0.5765765765765766,
+      "grad_norm": 0.3422398567199707,
+      "learning_rate": 1.8778221231930204e-05,
+      "loss": 0.3121,
+      "step": 144
+    },
+    {
+      "epoch": 0.5805805805805806,
+      "grad_norm": 0.3716834783554077,
+      "learning_rate": 1.8757149849142724e-05,
+      "loss": 0.3338,
+      "step": 145
+    },
+    {
+      "epoch": 0.5845845845845846,
+      "grad_norm": 0.313723087310791,
+      "learning_rate": 1.8735910344752925e-05,
+      "loss": 0.3294,
+      "step": 146
+    },
+    {
+      "epoch": 0.5885885885885885,
+      "grad_norm": 0.33704790472984314,
+      "learning_rate": 1.871450312652126e-05,
+      "loss": 0.3425,
+      "step": 147
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 0.3442137837409973,
+      "learning_rate": 1.8692928605428016e-05,
+      "loss": 0.3243,
+      "step": 148
+    },
+    {
+      "epoch": 0.5965965965965966,
+      "grad_norm": 0.3326564133167267,
+      "learning_rate": 1.8671187195665373e-05,
+      "loss": 0.3548,
+      "step": 149
+    },
+    {
+      "epoch": 0.6006006006006006,
+      "grad_norm": 0.37658053636550903,
+      "learning_rate": 1.8649279314629484e-05,
+      "loss": 0.3545,
+      "step": 150
+    },
+    {
+      "epoch": 0.6046046046046046,
+      "grad_norm": 0.34806281328201294,
+      "learning_rate": 1.862720538291245e-05,
+      "loss": 0.3365,
+      "step": 151
+    },
+    {
+      "epoch": 0.6086086086086087,
+      "grad_norm": 0.3958812355995178,
+      "learning_rate": 1.8604965824294253e-05,
+      "loss": 0.3682,
+      "step": 152
+    },
+    {
+      "epoch": 0.6126126126126126,
+      "grad_norm": 0.34817248582839966,
+      "learning_rate": 1.8582561065734602e-05,
+      "loss": 0.3454,
+      "step": 153
+    },
+    {
+      "epoch": 0.6166166166166166,
+      "grad_norm": 0.3551209270954132,
+      "learning_rate": 1.8559991537364767e-05,
+      "loss": 0.3466,
+      "step": 154
+    },
+    {
+      "epoch": 0.6206206206206206,
+      "grad_norm": 0.34840643405914307,
+      "learning_rate": 1.8537257672479293e-05,
+      "loss": 0.3186,
+      "step": 155
+    },
+    {
+      "epoch": 0.6246246246246246,
+      "grad_norm": 0.3974769413471222,
+      "learning_rate": 1.8514359907527693e-05,
+      "loss": 0.32,
+      "step": 156
+    },
+    {
+      "epoch": 0.6286286286286287,
+      "grad_norm": 0.36330661177635193,
+      "learning_rate": 1.8491298682106066e-05,
+      "loss": 0.3261,
+      "step": 157
+    },
+    {
+      "epoch": 0.6326326326326326,
+      "grad_norm": 0.3402544856071472,
+      "learning_rate": 1.8468074438948664e-05,
+      "loss": 0.335,
+      "step": 158
+    },
+    {
+      "epoch": 0.6366366366366366,
+      "grad_norm": 0.3627852201461792,
+      "learning_rate": 1.8444687623919388e-05,
+      "loss": 0.318,
+      "step": 159
+    },
+    {
+      "epoch": 0.6406406406406406,
+      "grad_norm": 0.3495313823223114,
+      "learning_rate": 1.842113868600322e-05,
+      "loss": 0.3235,
+      "step": 160
+    },
+    {
+      "epoch": 0.6446446446446447,
+      "grad_norm": 0.34827110171318054,
+      "learning_rate": 1.8397428077297622e-05,
+      "loss": 0.335,
+      "step": 161
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.35586410760879517,
+      "learning_rate": 1.837355625300383e-05,
+      "loss": 0.3163,
+      "step": 162
+    },
+    {
+      "epoch": 0.6526526526526526,
+      "grad_norm": 0.3451946973800659,
+      "learning_rate": 1.834952367141816e-05,
+      "loss": 0.3317,
+      "step": 163
+    },
+    {
+      "epoch": 0.6566566566566566,
+      "grad_norm": 0.36004722118377686,
+      "learning_rate": 1.8325330793923146e-05,
+      "loss": 0.3313,
+      "step": 164
+    },
+    {
+      "epoch": 0.6606606606606606,
+      "grad_norm": 0.32603421807289124,
+      "learning_rate": 1.8300978084978736e-05,
+      "loss": 0.3266,
+      "step": 165
+    },
+    {
+      "epoch": 0.6646646646646647,
+      "grad_norm": 0.33803558349609375,
+      "learning_rate": 1.8276466012113358e-05,
+      "loss": 0.3263,
+      "step": 166
+    },
+    {
+      "epoch": 0.6646646646646647,
+      "eval_loss": 0.3206555247306824,
+      "eval_runtime": 5.9746,
+      "eval_samples_per_second": 13.557,
+      "eval_steps_per_second": 1.841,
+      "step": 166
+    },
+    {
+      "epoch": 0.6686686686686687,
+      "grad_norm": 0.34881776571273804,
+      "learning_rate": 1.8251795045914922e-05,
+      "loss": 0.3575,
+      "step": 167
+    },
+    {
+      "epoch": 0.6726726726726727,
+      "grad_norm": 0.3603871464729309,
+      "learning_rate": 1.8226965660021836e-05,
+      "loss": 0.3215,
+      "step": 168
+    },
+    {
+      "epoch": 0.6766766766766766,
+      "grad_norm": 0.3549259901046753,
+      "learning_rate": 1.8201978331113855e-05,
+      "loss": 0.3302,
+      "step": 169
+    },
+    {
+      "epoch": 0.6806806806806807,
+      "grad_norm": 0.3330039381980896,
+      "learning_rate": 1.817683353890297e-05,
+      "loss": 0.339,
+      "step": 170
+    },
+    {
+      "epoch": 0.6846846846846847,
+      "grad_norm": 0.4140745997428894,
+      "learning_rate": 1.8151531766124186e-05,
+      "loss": 0.4009,
+      "step": 171
+    },
+    {
+      "epoch": 0.6886886886886887,
+      "grad_norm": 0.3429993987083435,
+      "learning_rate": 1.8126073498526254e-05,
+      "loss": 0.3203,
+      "step": 172
+    },
+    {
+      "epoch": 0.6926926926926927,
+      "grad_norm": 0.3445993661880493,
+      "learning_rate": 1.8100459224862336e-05,
+      "loss": 0.3352,
+      "step": 173
+    },
+    {
+      "epoch": 0.6966966966966966,
+      "grad_norm": 0.3043835461139679,
+      "learning_rate": 1.8074689436880643e-05,
+      "loss": 0.3294,
+      "step": 174
+    },
+    {
+      "epoch": 0.7007007007007007,
+      "grad_norm": 0.3373521566390991,
+      "learning_rate": 1.804876462931498e-05,
+      "loss": 0.3204,
+      "step": 175
+    },
+    {
+      "epoch": 0.7047047047047047,
+      "grad_norm": 0.33019447326660156,
+      "learning_rate": 1.8022685299875245e-05,
+      "loss": 0.3339,
+      "step": 176
+    },
+    {
+      "epoch": 0.7087087087087087,
+      "grad_norm": 0.32447516918182373,
+      "learning_rate": 1.799645194923788e-05,
+      "loss": 0.3216,
+      "step": 177
+    },
+    {
+      "epoch": 0.7127127127127127,
+      "grad_norm": 0.3202582597732544,
+      "learning_rate": 1.7970065081036266e-05,
+      "loss": 0.3104,
+      "step": 178
+    },
+    {
+      "epoch": 0.7167167167167167,
+      "grad_norm": 0.3486015200614929,
+      "learning_rate": 1.7943525201851038e-05,
+      "loss": 0.3307,
+      "step": 179
+    },
+    {
+      "epoch": 0.7207207207207207,
+      "grad_norm": 0.36710435152053833,
+      "learning_rate": 1.7916832821200375e-05,
+      "loss": 0.3472,
+      "step": 180
+    },
+    {
+      "epoch": 0.7247247247247247,
+      "grad_norm": 0.347162127494812,
+      "learning_rate": 1.7889988451530208e-05,
+      "loss": 0.3358,
+      "step": 181
+    },
+    {
+      "epoch": 0.7287287287287287,
+      "grad_norm": 0.3109472692012787,
+      "learning_rate": 1.7862992608204384e-05,
+      "loss": 0.2959,
+      "step": 182
+    },
+    {
+      "epoch": 0.7327327327327328,
+      "grad_norm": 0.34007692337036133,
+      "learning_rate": 1.783584580949477e-05,
+      "loss": 0.3199,
+      "step": 183
+    },
+    {
+      "epoch": 0.7367367367367368,
+      "grad_norm": 0.31094032526016235,
+      "learning_rate": 1.7808548576571314e-05,
+      "loss": 0.315,
+      "step": 184
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.367898553609848,
+      "learning_rate": 1.7781101433492026e-05,
+      "loss": 0.346,
+      "step": 185
+    },
+    {
+      "epoch": 0.7447447447447447,
+      "grad_norm": 0.3379008173942566,
+      "learning_rate": 1.7753504907192923e-05,
+      "loss": 0.3299,
+      "step": 186
+    },
+    {
+      "epoch": 0.7487487487487487,
+      "grad_norm": 0.3236868977546692,
+      "learning_rate": 1.7725759527477923e-05,
+      "loss": 0.3081,
+      "step": 187
+    },
+    {
+      "epoch": 0.7527527527527528,
+      "grad_norm": 0.34538552165031433,
+      "learning_rate": 1.769786582700864e-05,
+      "loss": 0.3259,
+      "step": 188
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 0.3106057941913605,
+      "learning_rate": 1.7669824341294203e-05,
+      "loss": 0.3123,
+      "step": 189
+    },
+    {
+      "epoch": 0.7607607607607607,
+      "grad_norm": 0.3219281733036041,
+      "learning_rate": 1.7641635608680942e-05,
+      "loss": 0.3165,
+      "step": 190
+    },
+    {
+      "epoch": 0.7647647647647647,
+      "grad_norm": 0.3259471356868744,
+      "learning_rate": 1.7613300170342073e-05,
+      "loss": 0.3453,
+      "step": 191
+    },
+    {
+      "epoch": 0.7687687687687688,
+      "grad_norm": 0.33405622839927673,
+      "learning_rate": 1.7584818570267287e-05,
+      "loss": 0.3233,
+      "step": 192
+    },
+    {
+      "epoch": 0.7727727727727728,
+      "grad_norm": 0.3420588970184326,
+      "learning_rate": 1.755619135525233e-05,
+      "loss": 0.3229,
+      "step": 193
+    },
+    {
+      "epoch": 0.7767767767767768,
+      "grad_norm": 0.35447970032691956,
+      "learning_rate": 1.7527419074888483e-05,
+      "loss": 0.3356,
+      "step": 194
+    },
+    {
+      "epoch": 0.7807807807807807,
+      "grad_norm": 0.34441742300987244,
+      "learning_rate": 1.749850228155203e-05,
+      "loss": 0.3128,
+      "step": 195
+    },
+    {
+      "epoch": 0.7847847847847848,
+      "grad_norm": 0.33647745847702026,
+      "learning_rate": 1.7469441530393652e-05,
+      "loss": 0.334,
+      "step": 196
+    },
+    {
+      "epoch": 0.7887887887887888,
+      "grad_norm": 0.3477891981601715,
+      "learning_rate": 1.7440237379327745e-05,
+      "loss": 0.3535,
+      "step": 197
+    },
+    {
+      "epoch": 0.7927927927927928,
+      "grad_norm": 0.35994237661361694,
+      "learning_rate": 1.7410890389021737e-05,
+      "loss": 0.3133,
+      "step": 198
+    },
+    {
+      "epoch": 0.7967967967967968,
+      "grad_norm": 0.33523160219192505,
+      "learning_rate": 1.7381401122885316e-05,
+      "loss": 0.3403,
+      "step": 199
+    },
+    {
+      "epoch": 0.8008008008008008,
+      "grad_norm": 0.3123041093349457,
+      "learning_rate": 1.7351770147059604e-05,
+      "loss": 0.3148,
+      "step": 200
+    },
+    {
+      "epoch": 0.8048048048048048,
+      "grad_norm": 0.30986520648002625,
+      "learning_rate": 1.7321998030406303e-05,
+      "loss": 0.3327,
+      "step": 201
+    },
+    {
+      "epoch": 0.8088088088088088,
+      "grad_norm": 0.3334694504737854,
+      "learning_rate": 1.729208534449676e-05,
+      "loss": 0.3346,
+      "step": 202
+    },
+    {
+      "epoch": 0.8128128128128128,
+      "grad_norm": 0.32513362169265747,
+      "learning_rate": 1.7262032663601003e-05,
+      "loss": 0.3137,
+      "step": 203
+    },
+    {
+      "epoch": 0.8168168168168168,
+      "grad_norm": 0.34156233072280884,
+      "learning_rate": 1.723184056467671e-05,
+      "loss": 0.3218,
+      "step": 204
+    },
+    {
+      "epoch": 0.8208208208208209,
+      "grad_norm": 0.32721707224845886,
+      "learning_rate": 1.7201509627358143e-05,
+      "loss": 0.3182,
+      "step": 205
+    },
+    {
+      "epoch": 0.8248248248248248,
+      "grad_norm": 0.35413238406181335,
+      "learning_rate": 1.7171040433945006e-05,
+      "loss": 0.2949,
+      "step": 206
+    },
+    {
+      "epoch": 0.8288288288288288,
+      "grad_norm": 0.30575960874557495,
+      "learning_rate": 1.7140433569391275e-05,
+      "loss": 0.3267,
+      "step": 207
+    },
+    {
+      "epoch": 0.8328328328328328,
+      "grad_norm": 0.3090551793575287,
+      "learning_rate": 1.710968962129396e-05,
+      "loss": 0.3072,
+      "step": 208
+    },
+    {
+      "epoch": 0.8368368368368369,
+      "grad_norm": 0.30769091844558716,
+      "learning_rate": 1.7078809179881847e-05,
+      "loss": 0.3115,
+      "step": 209
+    },
+    {
+      "epoch": 0.8408408408408409,
+      "grad_norm": 0.2806401550769806,
+      "learning_rate": 1.704779283800412e-05,
+      "loss": 0.3172,
+      "step": 210
+    },
+    {
+      "epoch": 0.8448448448448449,
+      "grad_norm": 0.322110652923584,
+      "learning_rate": 1.701664119111904e-05,
+      "loss": 0.319,
+      "step": 211
+    },
+    {
+      "epoch": 0.8488488488488488,
+      "grad_norm": 0.3080562353134155,
+      "learning_rate": 1.6985354837282462e-05,
+      "loss": 0.3247,
+      "step": 212
+    },
+    {
+      "epoch": 0.8528528528528528,
+      "grad_norm": 0.3051875829696655,
+      "learning_rate": 1.6953934377136375e-05,
+      "loss": 0.3195,
+      "step": 213
+    },
+    {
+      "epoch": 0.8568568568568569,
+      "grad_norm": 0.31982743740081787,
+      "learning_rate": 1.6922380413897382e-05,
+      "loss": 0.3202,
+      "step": 214
+    },
+    {
+      "epoch": 0.8608608608608609,
+      "grad_norm": 0.3235834538936615,
+      "learning_rate": 1.689069355334509e-05,
+      "loss": 0.3532,
+      "step": 215
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.32920244336128235,
+      "learning_rate": 1.6858874403810507e-05,
+      "loss": 0.3412,
+      "step": 216
+    },
+    {
+      "epoch": 0.8688688688688688,
+      "grad_norm": 0.3245142102241516,
+      "learning_rate": 1.682692357616435e-05,
+      "loss": 0.3074,
+      "step": 217
+    },
+    {
+      "epoch": 0.8728728728728729,
+      "grad_norm": 0.30122387409210205,
+      "learning_rate": 1.679484168380532e-05,
+      "loss": 0.3166,
+      "step": 218
+    },
+    {
+      "epoch": 0.8768768768768769,
+      "grad_norm": 0.30446872115135193,
+      "learning_rate": 1.676262934264832e-05,
+      "loss": 0.3177,
+      "step": 219
+    },
+    {
+      "epoch": 0.8808808808808809,
+      "grad_norm": 0.3006502091884613,
+      "learning_rate": 1.6730287171112652e-05,
+      "loss": 0.3225,
+      "step": 220
+    },
+    {
+      "epoch": 0.8848848848848849,
+      "grad_norm": 0.33892446756362915,
+      "learning_rate": 1.669781579011011e-05,
+      "loss": 0.335,
+      "step": 221
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.34253889322280884,
+      "learning_rate": 1.666521582303309e-05,
+      "loss": 0.3329,
+      "step": 222
+    },
+    {
+      "epoch": 0.8928928928928929,
+      "grad_norm": 0.291759729385376,
+      "learning_rate": 1.6632487895742612e-05,
+      "loss": 0.3173,
+      "step": 223
+    },
+    {
+      "epoch": 0.8968968968968969,
+      "grad_norm": 0.3063089847564697,
+      "learning_rate": 1.6599632636556292e-05,
+      "loss": 0.3345,
+      "step": 224
+    },
+    {
+      "epoch": 0.9009009009009009,
+      "grad_norm": 0.30597245693206787,
+      "learning_rate": 1.6566650676236307e-05,
+      "loss": 0.3435,
+      "step": 225
+    },
+    {
+      "epoch": 0.9049049049049049,
+      "grad_norm": 0.3654542863368988,
+      "learning_rate": 1.653354264797725e-05,
+      "loss": 0.32,
+      "step": 226
+    },
+    {
+      "epoch": 0.908908908908909,
+      "grad_norm": 0.30520570278167725,
+      "learning_rate": 1.6500309187394005e-05,
+      "loss": 0.3099,
+      "step": 227
+    },
+    {
+      "epoch": 0.9129129129129129,
+      "grad_norm": 0.3210243582725525,
+      "learning_rate": 1.6466950932509532e-05,
+      "loss": 0.316,
+      "step": 228
+    },
+    {
+      "epoch": 0.9169169169169169,
+      "grad_norm": 0.316346138715744,
+      "learning_rate": 1.643346852374261e-05,
+      "loss": 0.2977,
+      "step": 229
+    },
+    {
+      "epoch": 0.9209209209209209,
+      "grad_norm": 0.3127054274082184,
+      "learning_rate": 1.6399862603895563e-05,
+      "loss": 0.2942,
+      "step": 230
+    },
+    {
+      "epoch": 0.924924924924925,
+      "grad_norm": 0.3078126013278961,
+      "learning_rate": 1.6366133818141893e-05,
+      "loss": 0.3008,
+      "step": 231
+    },
+    {
+      "epoch": 0.928928928928929,
+      "grad_norm": 0.3247736096382141,
+      "learning_rate": 1.633228281401392e-05,
+      "loss": 0.3142,
+      "step": 232
+    },
+    {
+      "epoch": 0.9329329329329329,
+      "grad_norm": 0.3431578576564789,
+      "learning_rate": 1.6298310241390326e-05,
+      "loss": 0.3093,
+      "step": 233
+    },
+    {
+      "epoch": 0.9369369369369369,
+      "grad_norm": 0.31279513239860535,
+      "learning_rate": 1.6264216752483697e-05,
+      "loss": 0.3175,
+      "step": 234
+    },
+    {
+      "epoch": 0.9409409409409409,
+      "grad_norm": 0.33988532423973083,
+      "learning_rate": 1.6230003001828e-05,
+      "loss": 0.324,
+      "step": 235
+    },
+    {
+      "epoch": 0.944944944944945,
+      "grad_norm": 0.3213682472705841,
+      "learning_rate": 1.6195669646266003e-05,
+      "loss": 0.3321,
+      "step": 236
+    },
+    {
+      "epoch": 0.948948948948949,
+      "grad_norm": 0.357149213552475,
+      "learning_rate": 1.616121734493668e-05,
+      "loss": 0.3342,
+      "step": 237
+    },
+    {
+      "epoch": 0.9529529529529529,
+      "grad_norm": 0.31976473331451416,
+      "learning_rate": 1.6126646759262548e-05,
+      "loss": 0.3181,
+      "step": 238
+    },
+    {
+      "epoch": 0.9569569569569569,
+      "grad_norm": 0.328274130821228,
+      "learning_rate": 1.609195855293697e-05,
+      "loss": 0.3161,
+      "step": 239
+    },
+    {
+      "epoch": 0.960960960960961,
+      "grad_norm": 0.3303963840007782,
+      "learning_rate": 1.6057153391911422e-05,
+      "loss": 0.3076,
+      "step": 240
+    },
+    {
+      "epoch": 0.964964964964965,
+      "grad_norm": 0.34692683815956116,
+      "learning_rate": 1.6022231944382693e-05,
+      "loss": 0.3351,
+      "step": 241
+    },
+    {
+      "epoch": 0.968968968968969,
+      "grad_norm": 0.3468174636363983,
+      "learning_rate": 1.598719488078007e-05,
+      "loss": 0.3224,
+      "step": 242
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.3575330972671509,
+      "learning_rate": 1.5952042873752463e-05,
+      "loss": 0.3189,
+      "step": 243
+    },
+    {
+      "epoch": 0.9769769769769769,
+      "grad_norm": 0.35199496150016785,
+      "learning_rate": 1.5916776598155478e-05,
+      "loss": 0.3515,
+      "step": 244
+    },
+    {
+      "epoch": 0.980980980980981,
+      "grad_norm": 0.34917882084846497,
+      "learning_rate": 1.5881396731038493e-05,
+      "loss": 0.3354,
+      "step": 245
+    },
+    {
+      "epoch": 0.984984984984985,
+      "grad_norm": 0.32909733057022095,
+      "learning_rate": 1.584590395163162e-05,
+      "loss": 0.3009,
+      "step": 246
+    },
+    {
+      "epoch": 0.988988988988989,
+      "grad_norm": 0.3109247088432312,
+      "learning_rate": 1.5810298941332696e-05,
+      "loss": 0.3164,
+      "step": 247
+    },
+    {
+      "epoch": 0.992992992992993,
+      "grad_norm": 0.30618447065353394,
+      "learning_rate": 1.5774582383694196e-05,
+      "loss": 0.2923,
+      "step": 248
+    },
+    {
+      "epoch": 0.996996996996997,
+      "grad_norm": 0.32330596446990967,
+      "learning_rate": 1.5738754964410084e-05,
+      "loss": 0.3213,
+      "step": 249
+    },
+    {
+      "epoch": 0.996996996996997,
+      "eval_loss": 0.304058700799942,
+      "eval_runtime": 6.1571,
+      "eval_samples_per_second": 13.156,
+      "eval_steps_per_second": 1.787,
+      "step": 249
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32330596446990967,
+      "learning_rate": 1.5702817371302684e-05,
+      "loss": 0.3252,
+      "step": 250
+    },
+    {
+      "epoch": 1.004004004004004,
+      "grad_norm": 0.1360868364572525,
+      "learning_rate": 1.5666770294309467e-05,
+      "loss": 0.2529,
+      "step": 251
+    },
+    {
+      "epoch": 1.008008008008008,
+      "grad_norm": 0.1317552626132965,
+      "learning_rate": 1.5630614425469776e-05,
+      "loss": 0.238,
+      "step": 252
+    },
+    {
+      "epoch": 1.012012012012012,
+      "grad_norm": 0.12073508650064468,
+      "learning_rate": 1.5594350458911586e-05,
+      "loss": 0.2528,
+      "step": 253
+    },
+    {
+      "epoch": 1.016016016016016,
+      "grad_norm": 0.13055667281150818,
+      "learning_rate": 1.5557979090838136e-05,
+      "loss": 0.2519,
+      "step": 254
+    },
+    {
+      "epoch": 1.02002002002002,
+      "grad_norm": 0.1332577019929886,
+      "learning_rate": 1.55215010195146e-05,
+      "loss": 0.2483,
+      "step": 255
+    },
+    {
+      "epoch": 1.024024024024024,
+      "grad_norm": 0.14699849486351013,
+      "learning_rate": 1.5484916945254642e-05,
+      "loss": 0.2513,
+      "step": 256
+    },
+    {
+      "epoch": 1.028028028028028,
+      "grad_norm": 0.12925884127616882,
+      "learning_rate": 1.5448227570407012e-05,
+      "loss": 0.2426,
+      "step": 257
+    },
+    {
+      "epoch": 1.032032032032032,
+      "grad_norm": 0.14779068529605865,
+      "learning_rate": 1.5411433599342038e-05,
+      "loss": 0.2453,
+      "step": 258
+    },
+    {
+      "epoch": 1.0360360360360361,
+      "grad_norm": 0.11822405457496643,
+      "learning_rate": 1.5374535738438105e-05,
+      "loss": 0.2399,
+      "step": 259
+    },
+    {
+      "epoch": 1.04004004004004,
+      "grad_norm": 0.13844071328639984,
+      "learning_rate": 1.5337534696068088e-05,
+      "loss": 0.2291,
+      "step": 260
+    },
+    {
+      "epoch": 1.044044044044044,
+      "grad_norm": 0.12040767818689346,
+      "learning_rate": 1.5300431182585777e-05,
+      "loss": 0.2331,
+      "step": 261
+    },
+    {
+      "epoch": 1.048048048048048,
+      "grad_norm": 0.1192127987742424,
+      "learning_rate": 1.5263225910312222e-05,
+      "loss": 0.2424,
+      "step": 262
+    },
+    {
+      "epoch": 1.052052052052052,
+      "grad_norm": 0.14699995517730713,
+      "learning_rate": 1.5225919593522049e-05,
+      "loss": 0.2595,
+      "step": 263
+    },
+    {
+      "epoch": 1.0560560560560561,
+      "grad_norm": 0.1524747908115387,
+      "learning_rate": 1.5188512948429765e-05,
+      "loss": 0.232,
+      "step": 264
+    },
+    {
+      "epoch": 1.06006006006006,
+      "grad_norm": 0.12861889600753784,
+      "learning_rate": 1.5151006693176005e-05,
+      "loss": 0.245,
+      "step": 265
+    },
+    {
+      "epoch": 1.064064064064064,
+      "grad_norm": 0.15535978972911835,
+      "learning_rate": 1.5113401547813732e-05,
+      "loss": 0.2277,
+      "step": 266
+    },
+    {
+      "epoch": 1.068068068068068,
+      "grad_norm": 0.1480809897184372,
+      "learning_rate": 1.5075698234294424e-05,
+      "loss": 0.2265,
+      "step": 267
+    },
+    {
+      "epoch": 1.072072072072072,
+      "grad_norm": 0.1344352513551712,
+      "learning_rate": 1.5037897476454219e-05,
+      "loss": 0.2397,
+      "step": 268
+    },
+    {
+      "epoch": 1.0760760760760761,
+      "grad_norm": 0.1348842978477478,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.2443,
+      "step": 269
+    },
+    {
+      "epoch": 1.08008008008008,
+      "grad_norm": 0.1489027589559555,
+      "learning_rate": 1.496200653249549e-05,
+      "loss": 0.2358,
+      "step": 270
+    },
+    {
+      "epoch": 1.0840840840840842,
+      "grad_norm": 0.1447959840297699,
+      "learning_rate": 1.492391780334725e-05,
+      "loss": 0.2416,
+      "step": 271
+    },
+    {
+      "epoch": 1.088088088088088,
+      "grad_norm": 0.13745814561843872,
+      "learning_rate": 1.4885734543790707e-05,
+      "loss": 0.2494,
+      "step": 272
+    },
+    {
+      "epoch": 1.092092092092092,
+      "grad_norm": 0.14297857880592346,
+      "learning_rate": 1.4847457486876097e-05,
+      "loss": 0.2383,
+      "step": 273
+    },
+    {
+      "epoch": 1.0960960960960962,
+      "grad_norm": 0.14657525718212128,
+      "learning_rate": 1.4809087367454402e-05,
+      "loss": 0.2339,
+      "step": 274
+    },
+    {
+      "epoch": 1.1001001001001,
+      "grad_norm": 0.13032834231853485,
+      "learning_rate": 1.4770624922163233e-05,
+      "loss": 0.2337,
+      "step": 275
+    },
+    {
+      "epoch": 1.1041041041041042,
+      "grad_norm": 0.12887558341026306,
+      "learning_rate": 1.4732070889412693e-05,
+      "loss": 0.2461,
+      "step": 276
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 0.1498020589351654,
+      "learning_rate": 1.4693426009371203e-05,
+      "loss": 0.2622,
+      "step": 277
+    },
+    {
+      "epoch": 1.112112112112112,
+      "grad_norm": 0.13239090144634247,
+      "learning_rate": 1.4654691023951289e-05,
+      "loss": 0.2422,
+      "step": 278
+    },
+    {
+      "epoch": 1.1161161161161162,
+      "grad_norm": 0.14478150010108948,
+      "learning_rate": 1.4615866676795334e-05,
+      "loss": 0.233,
+      "step": 279
+    },
+    {
+      "epoch": 1.12012012012012,
+      "grad_norm": 0.12946031987667084,
+      "learning_rate": 1.4576953713261313e-05,
+      "loss": 0.235,
+      "step": 280
+    },
+    {
+      "epoch": 1.1241241241241242,
+      "grad_norm": 0.1348564624786377,
+      "learning_rate": 1.4537952880408472e-05,
+      "loss": 0.2386,
+      "step": 281
+    },
+    {
+      "epoch": 1.1281281281281281,
+      "grad_norm": 0.13074958324432373,
+      "learning_rate": 1.4498864926982996e-05,
+      "loss": 0.234,
+      "step": 282
+    },
+    {
+      "epoch": 1.132132132132132,
+      "grad_norm": 0.130098357796669,
+      "learning_rate": 1.4459690603403623e-05,
+      "loss": 0.2329,
+      "step": 283
+    },
+    {
+      "epoch": 1.1361361361361362,
+      "grad_norm": 0.1510149985551834,
+      "learning_rate": 1.4420430661747245e-05,
+      "loss": 0.2507,
+      "step": 284
+    },
+    {
+      "epoch": 1.14014014014014,
+      "grad_norm": 0.12636184692382812,
+      "learning_rate": 1.4381085855734468e-05,
+      "loss": 0.2305,
+      "step": 285
+    },
+    {
+      "epoch": 1.1441441441441442,
+      "grad_norm": 0.14648716151714325,
+      "learning_rate": 1.4341656940715147e-05,
+      "loss": 0.2263,
+      "step": 286
+    },
+    {
+      "epoch": 1.1481481481481481,
+      "grad_norm": 0.12833812832832336,
+      "learning_rate": 1.4302144673653875e-05,
+      "loss": 0.201,
+      "step": 287
+    },
+    {
+      "epoch": 1.1521521521521523,
+      "grad_norm": 0.14879822731018066,
+      "learning_rate": 1.426254981311545e-05,
+      "loss": 0.2452,
+      "step": 288
+    },
+    {
+      "epoch": 1.1561561561561562,
+      "grad_norm": 0.1363959014415741,
+      "learning_rate": 1.4222873119250325e-05,
+      "loss": 0.2366,
+      "step": 289
+    },
+    {
+      "epoch": 1.16016016016016,
+      "grad_norm": 0.12407127767801285,
+      "learning_rate": 1.4183115353780001e-05,
+      "loss": 0.2276,
+      "step": 290
+    },
+    {
+      "epoch": 1.1641641641641642,
+      "grad_norm": 0.1341707408428192,
+      "learning_rate": 1.4143277279982415e-05,
+      "loss": 0.2202,
+      "step": 291
+    },
+    {
+      "epoch": 1.1681681681681682,
+      "grad_norm": 0.13136214017868042,
+      "learning_rate": 1.4103359662677276e-05,
+      "loss": 0.2365,
+      "step": 292
+    },
+    {
+      "epoch": 1.1721721721721723,
+      "grad_norm": 0.13455568253993988,
+      "learning_rate": 1.406336326821138e-05,
+      "loss": 0.2329,
+      "step": 293
+    },
+    {
+      "epoch": 1.1761761761761762,
+      "grad_norm": 0.130329892039299,
+      "learning_rate": 1.4023288864443915e-05,
+      "loss": 0.2182,
+      "step": 294
+    },
+    {
+      "epoch": 1.1801801801801801,
+      "grad_norm": 0.1454066038131714,
+      "learning_rate": 1.3983137220731702e-05,
+      "loss": 0.232,
+      "step": 295
+    },
+    {
+      "epoch": 1.1841841841841843,
+      "grad_norm": 0.12771300971508026,
+      "learning_rate": 1.3942909107914431e-05,
+      "loss": 0.2459,
+      "step": 296
+    },
+    {
+      "epoch": 1.1881881881881882,
+      "grad_norm": 0.1486714780330658,
+      "learning_rate": 1.390260529829986e-05,
+      "loss": 0.2382,
+      "step": 297
+    },
+    {
+      "epoch": 1.1921921921921923,
+      "grad_norm": 0.12321332842111588,
+      "learning_rate": 1.3862226565648996e-05,
+      "loss": 0.2226,
+      "step": 298
+    },
+    {
+      "epoch": 1.1961961961961962,
+      "grad_norm": 0.11899197101593018,
+      "learning_rate": 1.3821773685161224e-05,
+      "loss": 0.2292,
+      "step": 299
+    },
+    {
+      "epoch": 1.2002002002002001,
+      "grad_norm": 0.13106873631477356,
+      "learning_rate": 1.3781247433459447e-05,
+      "loss": 0.2115,
+      "step": 300
+    },
+    {
+      "epoch": 1.2042042042042043,
+      "grad_norm": 0.1518913358449936,
+      "learning_rate": 1.3740648588575156e-05,
+      "loss": 0.2346,
+      "step": 301
+    },
+    {
+      "epoch": 1.2082082082082082,
+      "grad_norm": 0.13578097522258759,
+      "learning_rate": 1.3699977929933503e-05,
+      "loss": 0.2208,
+      "step": 302
+    },
+    {
+      "epoch": 1.2122122122122123,
+      "grad_norm": 0.13382072746753693,
+      "learning_rate": 1.3659236238338339e-05,
+      "loss": 0.229,
+      "step": 303
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 0.14184784889221191,
+      "learning_rate": 1.361842429595721e-05,
+      "loss": 0.2418,
+      "step": 304
+    },
+    {
+      "epoch": 1.2202202202202201,
+      "grad_norm": 0.13084499537944794,
+      "learning_rate": 1.3577542886306367e-05,
+      "loss": 0.2323,
+      "step": 305
+    },
+    {
+      "epoch": 1.2242242242242243,
+      "grad_norm": 0.13019020855426788,
+      "learning_rate": 1.3536592794235696e-05,
+      "loss": 0.2155,
+      "step": 306
+    },
+    {
+      "epoch": 1.2282282282282282,
+      "grad_norm": 0.14093178510665894,
+      "learning_rate": 1.3495574805913669e-05,
+      "loss": 0.236,
+      "step": 307
+    },
+    {
+      "epoch": 1.2322322322322323,
+      "grad_norm": 0.13968797028064728,
+      "learning_rate": 1.3454489708812237e-05,
+      "loss": 0.2192,
+      "step": 308
+    },
+    {
+      "epoch": 1.2362362362362362,
+      "grad_norm": 0.13795237243175507,
+      "learning_rate": 1.3413338291691726e-05,
+      "loss": 0.2236,
+      "step": 309
+    },
+    {
+      "epoch": 1.2402402402402402,
+      "grad_norm": 0.16295593976974487,
+      "learning_rate": 1.3372121344585694e-05,
+      "loss": 0.2429,
+      "step": 310
+    },
+    {
+      "epoch": 1.2442442442442443,
+      "grad_norm": 0.1423620730638504,
+      "learning_rate": 1.3330839658785739e-05,
+      "loss": 0.2411,
+      "step": 311
+    },
+    {
+      "epoch": 1.2482482482482482,
+      "grad_norm": 0.1591283529996872,
+      "learning_rate": 1.3289494026826337e-05,
+      "loss": 0.2446,
+      "step": 312
+    },
+    {
+      "epoch": 1.2522522522522523,
+      "grad_norm": 0.136406809091568,
+      "learning_rate": 1.3248085242469629e-05,
+      "loss": 0.223,
+      "step": 313
+    },
+    {
+      "epoch": 1.2562562562562563,
+      "grad_norm": 0.12121517956256866,
+      "learning_rate": 1.3206614100690139e-05,
+      "loss": 0.2238,
+      "step": 314
+    },
+    {
+      "epoch": 1.2602602602602602,
+      "grad_norm": 0.12819747626781464,
+      "learning_rate": 1.3165081397659563e-05,
+      "loss": 0.2352,
+      "step": 315
+    },
+    {
+      "epoch": 1.2642642642642643,
+      "grad_norm": 0.12027258425951004,
+      "learning_rate": 1.3123487930731464e-05,
+      "loss": 0.2181,
+      "step": 316
+    },
+    {
+      "epoch": 1.2682682682682682,
+      "grad_norm": 0.1466352492570877,
+      "learning_rate": 1.3081834498425952e-05,
+      "loss": 0.2352,
+      "step": 317
+    },
+    {
+      "epoch": 1.2722722722722724,
+      "grad_norm": 0.12040114402770996,
+      "learning_rate": 1.3040121900414371e-05,
+      "loss": 0.2104,
+      "step": 318
+    },
+    {
+      "epoch": 1.2762762762762763,
+      "grad_norm": 0.12626713514328003,
+      "learning_rate": 1.2998350937503939e-05,
+      "loss": 0.231,
+      "step": 319
+    },
+    {
+      "epoch": 1.2802802802802802,
+      "grad_norm": 0.13121092319488525,
+      "learning_rate": 1.2956522411622377e-05,
+      "loss": 0.2376,
+      "step": 320
+    },
+    {
+      "epoch": 1.2842842842842843,
+      "grad_norm": 0.13681887090206146,
+      "learning_rate": 1.2914637125802514e-05,
+      "loss": 0.2219,
+      "step": 321
+    },
+    {
+      "epoch": 1.2882882882882882,
+      "grad_norm": 0.15647488832473755,
+      "learning_rate": 1.287269588416686e-05,
+      "loss": 0.2372,
+      "step": 322
+    },
+    {
+      "epoch": 1.2922922922922924,
+      "grad_norm": 0.14653246104717255,
+      "learning_rate": 1.2830699491912186e-05,
+      "loss": 0.2245,
+      "step": 323
+    },
+    {
+      "epoch": 1.2962962962962963,
+      "grad_norm": 0.12728944420814514,
+      "learning_rate": 1.2788648755294056e-05,
+      "loss": 0.2186,
+      "step": 324
+    },
+    {
+      "epoch": 1.3003003003003002,
+      "grad_norm": 0.13707587122917175,
+      "learning_rate": 1.2746544481611336e-05,
+      "loss": 0.2236,
+      "step": 325
+    },
+    {
+      "epoch": 1.3043043043043043,
+      "grad_norm": 0.1246773898601532,
+      "learning_rate": 1.270438747919072e-05,
+      "loss": 0.2195,
+      "step": 326
+    },
+    {
+      "epoch": 1.3083083083083082,
+      "grad_norm": 0.13304340839385986,
+      "learning_rate": 1.2662178557371198e-05,
+      "loss": 0.2395,
+      "step": 327
+    },
+    {
+      "epoch": 1.3123123123123124,
+      "grad_norm": 0.15779724717140198,
+      "learning_rate": 1.261991852648852e-05,
+      "loss": 0.2435,
+      "step": 328
+    },
+    {
+      "epoch": 1.3163163163163163,
+      "grad_norm": 0.13654161989688873,
+      "learning_rate": 1.2577608197859627e-05,
+      "loss": 0.2316,
+      "step": 329
+    },
+    {
+      "epoch": 1.3203203203203202,
+      "grad_norm": 0.15000677108764648,
+      "learning_rate": 1.2535248383767102e-05,
+      "loss": 0.2292,
+      "step": 330
+    },
+    {
+      "epoch": 1.3243243243243243,
+      "grad_norm": 0.14126931130886078,
+      "learning_rate": 1.2492839897443554e-05,
+      "loss": 0.2364,
+      "step": 331
+    },
+    {
+      "epoch": 1.3283283283283283,
+      "grad_norm": 0.1323651373386383,
+      "learning_rate": 1.2450383553056011e-05,
+      "loss": 0.2369,
+      "step": 332
+    },
+    {
+      "epoch": 1.3283283283283283,
+      "eval_loss": 0.31280121207237244,
+      "eval_runtime": 6.244,
+      "eval_samples_per_second": 12.972,
+      "eval_steps_per_second": 1.762,
+      "step": 332
+    },
+    {
+      "epoch": 1.3323323323323324,
+      "grad_norm": 0.13480286300182343,
+      "learning_rate": 1.2407880165690289e-05,
+      "loss": 0.2203,
+      "step": 333
+    },
+    {
+      "epoch": 1.3363363363363363,
+      "grad_norm": 0.13146036863327026,
+      "learning_rate": 1.2365330551335348e-05,
+      "loss": 0.2261,
+      "step": 334
+    },
+    {
+      "epoch": 1.3403403403403402,
+      "grad_norm": 0.1318022459745407,
+      "learning_rate": 1.2322735526867624e-05,
+      "loss": 0.2328,
+      "step": 335
+    },
+    {
+      "epoch": 1.3443443443443444,
+      "grad_norm": 0.11951415240764618,
+      "learning_rate": 1.2280095910035343e-05,
+      "loss": 0.2183,
+      "step": 336
+    },
+    {
+      "epoch": 1.3483483483483483,
+      "grad_norm": 0.13236179947853088,
+      "learning_rate": 1.2237412519442828e-05,
+      "loss": 0.2292,
+      "step": 337
+    },
+    {
+      "epoch": 1.3523523523523524,
+      "grad_norm": 0.16215181350708008,
+      "learning_rate": 1.2194686174534771e-05,
+      "loss": 0.2459,
+      "step": 338
+    },
+    {
+      "epoch": 1.3563563563563563,
+      "grad_norm": 0.1451573669910431,
+      "learning_rate": 1.2151917695580523e-05,
+      "loss": 0.2276,
+      "step": 339
+    },
+    {
+      "epoch": 1.3603603603603602,
+      "grad_norm": 0.14320491254329681,
+      "learning_rate": 1.2109107903658326e-05,
+      "loss": 0.2364,
+      "step": 340
+    },
+    {
+      "epoch": 1.3643643643643644,
+      "grad_norm": 0.13618679344654083,
+      "learning_rate": 1.2066257620639557e-05,
+      "loss": 0.2227,
+      "step": 341
+    },
+    {
+      "epoch": 1.3683683683683685,
+      "grad_norm": 0.1250023990869522,
+      "learning_rate": 1.2023367669172947e-05,
+      "loss": 0.2303,
+      "step": 342
+    },
+    {
+      "epoch": 1.3723723723723724,
+      "grad_norm": 0.14609824120998383,
+      "learning_rate": 1.1980438872668803e-05,
+      "loss": 0.2223,
+      "step": 343
+    },
+    {
+      "epoch": 1.3763763763763763,
+      "grad_norm": 0.15771915018558502,
+      "learning_rate": 1.1937472055283168e-05,
+      "loss": 0.2295,
+      "step": 344
+    },
+    {
+      "epoch": 1.3803803803803802,
+      "grad_norm": 0.1346847116947174,
+      "learning_rate": 1.189446804190203e-05,
+      "loss": 0.2275,
+      "step": 345
+    },
+    {
+      "epoch": 1.3843843843843844,
+      "grad_norm": 0.1324576437473297,
+      "learning_rate": 1.1851427658125474e-05,
+      "loss": 0.2387,
+      "step": 346
+    },
+    {
+      "epoch": 1.3883883883883885,
+      "grad_norm": 0.11786400526762009,
+      "learning_rate": 1.180835173025183e-05,
+      "loss": 0.2151,
+      "step": 347
+    },
+    {
+      "epoch": 1.3923923923923924,
+      "grad_norm": 0.12627920508384705,
+      "learning_rate": 1.1765241085261802e-05,
+      "loss": 0.2249,
+      "step": 348
+    },
+    {
+      "epoch": 1.3963963963963963,
+      "grad_norm": 0.15186864137649536,
+      "learning_rate": 1.172209655080262e-05,
+      "loss": 0.2319,
+      "step": 349
+    },
+    {
+      "epoch": 1.4004004004004005,
+      "grad_norm": 0.17141692340373993,
+      "learning_rate": 1.1678918955172112e-05,
+      "loss": 0.2421,
+      "step": 350
+    },
+    {
+      "epoch": 1.4044044044044044,
+      "grad_norm": 0.13851796090602875,
+      "learning_rate": 1.163570912730283e-05,
+      "loss": 0.2168,
+      "step": 351
+    },
+    {
+      "epoch": 1.4084084084084085,
+      "grad_norm": 0.12362354248762131,
+      "learning_rate": 1.1592467896746122e-05,
+      "loss": 0.2173,
+      "step": 352
+    },
+    {
+      "epoch": 1.4124124124124124,
+      "grad_norm": 0.12605208158493042,
+      "learning_rate": 1.1549196093656223e-05,
+      "loss": 0.2299,
+      "step": 353
+    },
+    {
+      "epoch": 1.4164164164164164,
+      "grad_norm": 0.13899150490760803,
+      "learning_rate": 1.1505894548774294e-05,
+      "loss": 0.2214,
+      "step": 354
+    },
+    {
+      "epoch": 1.4204204204204205,
+      "grad_norm": 0.13161474466323853,
+      "learning_rate": 1.1462564093412493e-05,
+      "loss": 0.2322,
+      "step": 355
+    },
+    {
+      "epoch": 1.4244244244244244,
+      "grad_norm": 0.13784383237361908,
+      "learning_rate": 1.1419205559437998e-05,
+      "loss": 0.2399,
+      "step": 356
+    },
+    {
+      "epoch": 1.4284284284284285,
+      "grad_norm": 0.13900108635425568,
+      "learning_rate": 1.1375819779257058e-05,
+      "loss": 0.2332,
+      "step": 357
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 0.13527807593345642,
+      "learning_rate": 1.1332407585798992e-05,
+      "loss": 0.2294,
+      "step": 358
+    },
+    {
+      "epoch": 1.4364364364364364,
+      "grad_norm": 0.11528453975915909,
+      "learning_rate": 1.1288969812500209e-05,
+      "loss": 0.2131,
+      "step": 359
+    },
+    {
+      "epoch": 1.4404404404404405,
+      "grad_norm": 0.15680839121341705,
+      "learning_rate": 1.1245507293288204e-05,
+      "loss": 0.2296,
+      "step": 360
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.13539738953113556,
+      "learning_rate": 1.1202020862565555e-05,
+      "loss": 0.2326,
+      "step": 361
+    },
+    {
+      "epoch": 1.4484484484484486,
+      "grad_norm": 0.14512887597084045,
+      "learning_rate": 1.1158511355193888e-05,
+      "loss": 0.2254,
+      "step": 362
+    },
+    {
+      "epoch": 1.4524524524524525,
+      "grad_norm": 0.15710808336734772,
+      "learning_rate": 1.1114979606477867e-05,
+      "loss": 0.2395,
+      "step": 363
+    },
+    {
+      "epoch": 1.4564564564564564,
+      "grad_norm": 0.1274499148130417,
+      "learning_rate": 1.1071426452149152e-05,
+      "loss": 0.2203,
+      "step": 364
+    },
+    {
+      "epoch": 1.4604604604604605,
+      "grad_norm": 0.1460108906030655,
+      "learning_rate": 1.1027852728350343e-05,
+      "loss": 0.2477,
+      "step": 365
+    },
+    {
+      "epoch": 1.4644644644644644,
+      "grad_norm": 0.1274898201227188,
+      "learning_rate": 1.0984259271618947e-05,
+      "loss": 0.218,
+      "step": 366
+    },
+    {
+      "epoch": 1.4684684684684686,
+      "grad_norm": 0.14969086647033691,
+      "learning_rate": 1.09406469188713e-05,
+      "loss": 0.2307,
+      "step": 367
+    },
+    {
+      "epoch": 1.4724724724724725,
+      "grad_norm": 0.1416291445493698,
+      "learning_rate": 1.0897016507386513e-05,
+      "loss": 0.2208,
+      "step": 368
+    },
+    {
+      "epoch": 1.4764764764764764,
+      "grad_norm": 0.11650918424129486,
+      "learning_rate": 1.0853368874790392e-05,
+      "loss": 0.228,
+      "step": 369
+    },
+    {
+      "epoch": 1.4804804804804805,
+      "grad_norm": 0.14399082958698273,
+      "learning_rate": 1.0809704859039357e-05,
+      "loss": 0.2297,
+      "step": 370
+    },
+    {
+      "epoch": 1.4844844844844844,
+      "grad_norm": 0.1333630383014679,
+      "learning_rate": 1.0766025298404346e-05,
+      "loss": 0.2391,
+      "step": 371
+    },
+    {
+      "epoch": 1.4884884884884886,
+      "grad_norm": 0.1538272351026535,
+      "learning_rate": 1.0722331031454749e-05,
+      "loss": 0.228,
+      "step": 372
+    },
+    {
+      "epoch": 1.4924924924924925,
+      "grad_norm": 0.14864470064640045,
+      "learning_rate": 1.0678622897042279e-05,
+      "loss": 0.2307,
+      "step": 373
+    },
+    {
+      "epoch": 1.4964964964964964,
+      "grad_norm": 0.14386948943138123,
+      "learning_rate": 1.063490173428488e-05,
+      "loss": 0.2491,
+      "step": 374
+    },
+    {
+      "epoch": 1.5005005005005005,
+      "grad_norm": 0.13960744440555573,
+      "learning_rate": 1.0591168382550617e-05,
+      "loss": 0.2443,
+      "step": 375
+    },
+    {
+      "epoch": 1.5045045045045045,
+      "grad_norm": 0.12918636202812195,
+      "learning_rate": 1.0547423681441567e-05,
+      "loss": 0.237,
+      "step": 376
+    },
+    {
+      "epoch": 1.5085085085085086,
+      "grad_norm": 0.13551214337348938,
+      "learning_rate": 1.050366847077769e-05,
+      "loss": 0.2185,
+      "step": 377
+    },
+    {
+      "epoch": 1.5125125125125125,
+      "grad_norm": 0.14278538525104523,
+      "learning_rate": 1.0459903590580706e-05,
+      "loss": 0.2287,
+      "step": 378
+    },
+    {
+      "epoch": 1.5165165165165164,
+      "grad_norm": 0.1472548395395279,
+      "learning_rate": 1.0416129881057987e-05,
+      "loss": 0.2315,
+      "step": 379
+    },
+    {
+      "epoch": 1.5205205205205206,
+      "grad_norm": 0.13086426258087158,
+      "learning_rate": 1.03723481825864e-05,
+      "loss": 0.2176,
+      "step": 380
+    },
+    {
+      "epoch": 1.5245245245245245,
+      "grad_norm": 0.1289745569229126,
+      "learning_rate": 1.0328559335696188e-05,
+      "loss": 0.2279,
+      "step": 381
+    },
+    {
+      "epoch": 1.5285285285285286,
+      "grad_norm": 0.1345757395029068,
+      "learning_rate": 1.028476418105483e-05,
+      "loss": 0.2195,
+      "step": 382
+    },
+    {
+      "epoch": 1.5325325325325325,
+      "grad_norm": 0.1544564962387085,
+      "learning_rate": 1.0240963559450909e-05,
+      "loss": 0.2431,
+      "step": 383
+    },
+    {
+      "epoch": 1.5365365365365364,
+      "grad_norm": 0.13208632171154022,
+      "learning_rate": 1.0197158311777957e-05,
+      "loss": 0.2159,
+      "step": 384
+    },
+    {
+      "epoch": 1.5405405405405406,
+      "grad_norm": 0.12936881184577942,
+      "learning_rate": 1.015334927901832e-05,
+      "loss": 0.2085,
+      "step": 385
+    },
+    {
+      "epoch": 1.5445445445445447,
+      "grad_norm": 0.12098892778158188,
+      "learning_rate": 1.0109537302227012e-05,
+      "loss": 0.2299,
+      "step": 386
+    },
+    {
+      "epoch": 1.5485485485485486,
+      "grad_norm": 0.1374930441379547,
+      "learning_rate": 1.0065723222515566e-05,
+      "loss": 0.223,
+      "step": 387
+    },
+    {
+      "epoch": 1.5525525525525525,
+      "grad_norm": 0.14887316524982452,
+      "learning_rate": 1.0021907881035891e-05,
+      "loss": 0.2266,
+      "step": 388
+    },
+    {
+      "epoch": 1.5565565565565564,
+      "grad_norm": 0.13595186173915863,
+      "learning_rate": 9.97809211896411e-06,
+      "loss": 0.2182,
+      "step": 389
+    },
+    {
+      "epoch": 1.5605605605605606,
+      "grad_norm": 0.144562229514122,
+      "learning_rate": 9.934276777484436e-06,
+      "loss": 0.2271,
+      "step": 390
+    },
+    {
+      "epoch": 1.5645645645645647,
+      "grad_norm": 0.13522008061408997,
+      "learning_rate": 9.89046269777299e-06,
+      "loss": 0.228,
+      "step": 391
+    },
+    {
+      "epoch": 1.5685685685685686,
+      "grad_norm": 0.12187031656503677,
+      "learning_rate": 9.846650720981682e-06,
+      "loss": 0.2112,
+      "step": 392
+    },
+    {
+      "epoch": 1.5725725725725725,
+      "grad_norm": 0.14943794906139374,
+      "learning_rate": 9.802841688222043e-06,
+      "loss": 0.2148,
+      "step": 393
+    },
+    {
+      "epoch": 1.5765765765765765,
+      "grad_norm": 0.13538117706775665,
+      "learning_rate": 9.759036440549093e-06,
+      "loss": 0.2204,
+      "step": 394
+    },
+    {
+      "epoch": 1.5805805805805806,
+      "grad_norm": 0.12736183404922485,
+      "learning_rate": 9.715235818945171e-06,
+      "loss": 0.2118,
+      "step": 395
+    },
+    {
+      "epoch": 1.5845845845845847,
+      "grad_norm": 0.1528167724609375,
+      "learning_rate": 9.671440664303813e-06,
+      "loss": 0.235,
+      "step": 396
+    },
+    {
+      "epoch": 1.5885885885885886,
+      "grad_norm": 0.12496887892484665,
+      "learning_rate": 9.627651817413605e-06,
+      "loss": 0.2222,
+      "step": 397
+    },
+    {
+      "epoch": 1.5925925925925926,
+      "grad_norm": 0.11347772926092148,
+      "learning_rate": 9.583870118942014e-06,
+      "loss": 0.2125,
+      "step": 398
+    },
+    {
+      "epoch": 1.5965965965965965,
+      "grad_norm": 0.12634208798408508,
+      "learning_rate": 9.540096409419295e-06,
+      "loss": 0.209,
+      "step": 399
+    },
+    {
+      "epoch": 1.6006006006006006,
+      "grad_norm": 0.1373746246099472,
+      "learning_rate": 9.496331529222313e-06,
+      "loss": 0.2223,
+      "step": 400
+    },
+    {
+      "epoch": 1.6046046046046047,
+      "grad_norm": 0.13776709139347076,
+      "learning_rate": 9.452576318558437e-06,
+      "loss": 0.212,
+      "step": 401
+    },
+    {
+      "epoch": 1.6086086086086087,
+      "grad_norm": 0.13330040872097015,
+      "learning_rate": 9.408831617449385e-06,
+      "loss": 0.2319,
+      "step": 402
+    },
+    {
+      "epoch": 1.6126126126126126,
+      "grad_norm": 0.1234062984585762,
+      "learning_rate": 9.365098265715124e-06,
+      "loss": 0.2195,
+      "step": 403
+    },
+    {
+      "epoch": 1.6166166166166165,
+      "grad_norm": 0.13824118673801422,
+      "learning_rate": 9.321377102957723e-06,
+      "loss": 0.2119,
+      "step": 404
+    },
+    {
+      "epoch": 1.6206206206206206,
+      "grad_norm": 0.13618600368499756,
+      "learning_rate": 9.277668968545253e-06,
+      "loss": 0.2179,
+      "step": 405
+    },
+    {
+      "epoch": 1.6246246246246248,
+      "grad_norm": 0.15361671149730682,
+      "learning_rate": 9.233974701595654e-06,
+      "loss": 0.2162,
+      "step": 406
+    },
+    {
+      "epoch": 1.6286286286286287,
+      "grad_norm": 0.15466053783893585,
+      "learning_rate": 9.190295140960649e-06,
+      "loss": 0.2199,
+      "step": 407
+    },
+    {
+      "epoch": 1.6326326326326326,
+      "grad_norm": 0.1456158608198166,
+      "learning_rate": 9.146631125209608e-06,
+      "loss": 0.2192,
+      "step": 408
+    },
+    {
+      "epoch": 1.6366366366366365,
+      "grad_norm": 0.1501886397600174,
+      "learning_rate": 9.102983492613489e-06,
+      "loss": 0.2269,
+      "step": 409
+    },
+    {
+      "epoch": 1.6406406406406406,
+      "grad_norm": 0.13893765211105347,
+      "learning_rate": 9.059353081128702e-06,
+      "loss": 0.2322,
+      "step": 410
+    },
+    {
+      "epoch": 1.6446446446446448,
+      "grad_norm": 0.15337541699409485,
+      "learning_rate": 9.015740728381055e-06,
+      "loss": 0.2262,
+      "step": 411
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 0.13367938995361328,
+      "learning_rate": 8.972147271649662e-06,
+      "loss": 0.2095,
+      "step": 412
+    },
+    {
+      "epoch": 1.6526526526526526,
+      "grad_norm": 0.1440502554178238,
+      "learning_rate": 8.928573547850852e-06,
+      "loss": 0.2287,
+      "step": 413
+    },
+    {
+      "epoch": 1.6566566566566565,
+      "grad_norm": 0.1278400868177414,
+      "learning_rate": 8.885020393522136e-06,
+      "loss": 0.2233,
+      "step": 414
+    },
+    {
+      "epoch": 1.6606606606606606,
+      "grad_norm": 0.15888282656669617,
+      "learning_rate": 8.841488644806115e-06,
+      "loss": 0.2436,
+      "step": 415
+    },
+    {
+      "epoch": 1.6606606606606606,
+      "eval_loss": 0.30414843559265137,
+      "eval_runtime": 6.1188,
+      "eval_samples_per_second": 13.238,
+      "eval_steps_per_second": 1.798,
+      "step": 415
+    },
+    {
+      "epoch": 1.6646646646646648,
+      "grad_norm": 0.1494596004486084,
+      "learning_rate": 8.797979137434452e-06,
+      "loss": 0.2171,
+      "step": 416
+    },
+    {
+      "epoch": 1.6686686686686687,
+      "grad_norm": 0.14288803935050964,
+      "learning_rate": 8.754492706711798e-06,
+      "loss": 0.218,
+      "step": 417
+    },
+    {
+      "epoch": 1.6726726726726726,
+      "grad_norm": 0.14627306163311005,
+      "learning_rate": 8.711030187499795e-06,
+      "loss": 0.2196,
+      "step": 418
+    },
+    {
+      "epoch": 1.6766766766766765,
+      "grad_norm": 0.15222109854221344,
+      "learning_rate": 8.66759241420101e-06,
+      "loss": 0.2323,
+      "step": 419
+    },
+    {
+      "epoch": 1.6806806806806807,
+      "grad_norm": 0.1480226367712021,
+      "learning_rate": 8.624180220742945e-06,
+      "loss": 0.2168,
+      "step": 420
+    },
+    {
+      "epoch": 1.6846846846846848,
+      "grad_norm": 0.14224019646644592,
+      "learning_rate": 8.580794440562003e-06,
+      "loss": 0.2228,
+      "step": 421
+    },
+    {
+      "epoch": 1.6886886886886887,
+      "grad_norm": 0.13837142288684845,
+      "learning_rate": 8.53743590658751e-06,
+      "loss": 0.2175,
+      "step": 422
+    },
+    {
+      "epoch": 1.6926926926926926,
+      "grad_norm": 0.1258038580417633,
+      "learning_rate": 8.494105451225706e-06,
+      "loss": 0.2078,
+      "step": 423
+    },
+    {
+      "epoch": 1.6966966966966965,
+      "grad_norm": 0.1426319181919098,
+      "learning_rate": 8.45080390634378e-06,
+      "loss": 0.2228,
+      "step": 424
+    },
+    {
+      "epoch": 1.7007007007007007,
+      "grad_norm": 0.14483590424060822,
+      "learning_rate": 8.407532103253878e-06,
+      "loss": 0.2249,
+      "step": 425
+    },
+    {
+      "epoch": 1.7047047047047048,
+      "grad_norm": 0.1365538090467453,
+      "learning_rate": 8.364290872697175e-06,
+      "loss": 0.2241,
+      "step": 426
+    },
+    {
+      "epoch": 1.7087087087087087,
+      "grad_norm": 0.15289802849292755,
+      "learning_rate": 8.321081044827894e-06,
+      "loss": 0.2182,
+      "step": 427
+    },
+    {
+      "epoch": 1.7127127127127126,
+      "grad_norm": 0.13033711910247803,
+      "learning_rate": 8.277903449197383e-06,
+      "loss": 0.216,
+      "step": 428
+    },
+    {
+      "epoch": 1.7167167167167166,
+      "grad_norm": 0.13258293271064758,
+      "learning_rate": 8.2347589147382e-06,
+      "loss": 0.1956,
+      "step": 429
+    },
+    {
+      "epoch": 1.7207207207207207,
+      "grad_norm": 0.13521750271320343,
+      "learning_rate": 8.191648269748173e-06,
+      "loss": 0.2258,
+      "step": 430
+    },
+    {
+      "epoch": 1.7247247247247248,
+      "grad_norm": 0.15456490218639374,
+      "learning_rate": 8.14857234187453e-06,
+      "loss": 0.2256,
+      "step": 431
+    },
+    {
+      "epoch": 1.7287287287287287,
+      "grad_norm": 0.13333886861801147,
+      "learning_rate": 8.105531958097973e-06,
+      "loss": 0.2209,
+      "step": 432
+    },
+    {
+      "epoch": 1.7327327327327327,
+      "grad_norm": 0.1298181712627411,
+      "learning_rate": 8.062527944716837e-06,
+      "loss": 0.2106,
+      "step": 433
+    },
+    {
+      "epoch": 1.7367367367367368,
+      "grad_norm": 0.13067086040973663,
+      "learning_rate": 8.019561127331202e-06,
+      "loss": 0.224,
+      "step": 434
+    },
+    {
+      "epoch": 1.7407407407407407,
+      "grad_norm": 0.13905219733715057,
+      "learning_rate": 7.976632330827056e-06,
+      "loss": 0.2206,
+      "step": 435
+    },
+    {
+      "epoch": 1.7447447447447448,
+      "grad_norm": 0.13740921020507812,
+      "learning_rate": 7.933742379360446e-06,
+      "loss": 0.217,
+      "step": 436
+    },
+    {
+      "epoch": 1.7487487487487487,
+      "grad_norm": 0.12991660833358765,
+      "learning_rate": 7.890892096341677e-06,
+      "loss": 0.2173,
+      "step": 437
+    },
+    {
+      "epoch": 1.7527527527527527,
+      "grad_norm": 0.13946564495563507,
+      "learning_rate": 7.848082304419478e-06,
+      "loss": 0.2293,
+      "step": 438
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 0.12789933383464813,
+      "learning_rate": 7.805313825465232e-06,
+      "loss": 0.2116,
+      "step": 439
+    },
+    {
+      "epoch": 1.7607607607607607,
+      "grad_norm": 0.14676792919635773,
+      "learning_rate": 7.762587480557175e-06,
+      "loss": 0.2293,
+      "step": 440
+    },
+    {
+      "epoch": 1.7647647647647648,
+      "grad_norm": 0.14003834128379822,
+      "learning_rate": 7.719904089964658e-06,
+      "loss": 0.2137,
+      "step": 441
+    },
+    {
+      "epoch": 1.7687687687687688,
+      "grad_norm": 0.14215737581253052,
+      "learning_rate": 7.67726447313238e-06,
+      "loss": 0.2221,
+      "step": 442
+    },
+    {
+      "epoch": 1.7727727727727727,
+      "grad_norm": 0.12992604076862335,
+      "learning_rate": 7.634669448664655e-06,
+      "loss": 0.2006,
+      "step": 443
+    },
+    {
+      "epoch": 1.7767767767767768,
+      "grad_norm": 0.13772279024124146,
+      "learning_rate": 7.5921198343097145e-06,
+      "loss": 0.2266,
+      "step": 444
+    },
+    {
+      "epoch": 1.7807807807807807,
+      "grad_norm": 0.14676547050476074,
+      "learning_rate": 7.549616446943992e-06,
+      "loss": 0.2248,
+      "step": 445
+    },
+    {
+      "epoch": 1.7847847847847849,
+      "grad_norm": 0.13406066596508026,
+      "learning_rate": 7.507160102556451e-06,
+      "loss": 0.2028,
+      "step": 446
+    },
+    {
+      "epoch": 1.7887887887887888,
+      "grad_norm": 0.14275118708610535,
+      "learning_rate": 7.464751616232902e-06,
+      "loss": 0.2129,
+      "step": 447
+    },
+    {
+      "epoch": 1.7927927927927927,
+      "grad_norm": 0.1265939474105835,
+      "learning_rate": 7.422391802140376e-06,
+      "loss": 0.2102,
+      "step": 448
+    },
+    {
+      "epoch": 1.7967967967967968,
+      "grad_norm": 0.13890117406845093,
+      "learning_rate": 7.380081473511484e-06,
+      "loss": 0.2094,
+      "step": 449
+    },
+    {
+      "epoch": 1.800800800800801,
+      "grad_norm": 0.14124225080013275,
+      "learning_rate": 7.337821442628805e-06,
+      "loss": 0.2323,
+      "step": 450
+    },
+    {
+      "epoch": 1.8048048048048049,
+      "grad_norm": 0.15107014775276184,
+      "learning_rate": 7.295612520809281e-06,
+      "loss": 0.224,
+      "step": 451
+    },
+    {
+      "epoch": 1.8088088088088088,
+      "grad_norm": 0.1297285109758377,
+      "learning_rate": 7.253455518388668e-06,
+      "loss": 0.2053,
+      "step": 452
+    },
+    {
+      "epoch": 1.8128128128128127,
+      "grad_norm": 0.12204065918922424,
+      "learning_rate": 7.211351244705947e-06,
+      "loss": 0.2189,
+      "step": 453
+    },
+    {
+      "epoch": 1.8168168168168168,
+      "grad_norm": 0.15427479147911072,
+      "learning_rate": 7.169300508087815e-06,
+      "loss": 0.2337,
+      "step": 454
+    },
+    {
+      "epoch": 1.820820820820821,
+      "grad_norm": 0.13939781486988068,
+      "learning_rate": 7.127304115833141e-06,
+      "loss": 0.2175,
+      "step": 455
+    },
+    {
+      "epoch": 1.8248248248248249,
+      "grad_norm": 0.1337432563304901,
+      "learning_rate": 7.08536287419749e-06,
+      "loss": 0.2198,
+      "step": 456
+    },
+    {
+      "epoch": 1.8288288288288288,
+      "grad_norm": 0.19405747950077057,
+      "learning_rate": 7.043477588377623e-06,
+      "loss": 0.2263,
+      "step": 457
+    },
+    {
+      "epoch": 1.8328328328328327,
+      "grad_norm": 0.13952726125717163,
+      "learning_rate": 7.001649062496065e-06,
+      "loss": 0.2408,
+      "step": 458
+    },
+    {
+      "epoch": 1.8368368368368369,
+      "grad_norm": 0.14409776031970978,
+      "learning_rate": 6.959878099585634e-06,
+      "loss": 0.2216,
+      "step": 459
+    },
+    {
+      "epoch": 1.840840840840841,
+      "grad_norm": 0.13670873641967773,
+      "learning_rate": 6.918165501574051e-06,
+      "loss": 0.2225,
+      "step": 460
+    },
+    {
+      "epoch": 1.844844844844845,
+      "grad_norm": 0.11937157064676285,
+      "learning_rate": 6.876512069268541e-06,
+      "loss": 0.2097,
+      "step": 461
+    },
+    {
+      "epoch": 1.8488488488488488,
+      "grad_norm": 0.12589742243289948,
+      "learning_rate": 6.834918602340439e-06,
+      "loss": 0.2189,
+      "step": 462
+    },
+    {
+      "epoch": 1.8528528528528527,
+      "grad_norm": 0.12316793203353882,
+      "learning_rate": 6.793385899309866e-06,
+      "loss": 0.2098,
+      "step": 463
+    },
+    {
+      "epoch": 1.8568568568568569,
+      "grad_norm": 0.13508102297782898,
+      "learning_rate": 6.751914757530375e-06,
+      "loss": 0.2311,
+      "step": 464
+    },
+    {
+      "epoch": 1.860860860860861,
+      "grad_norm": 0.13518530130386353,
+      "learning_rate": 6.7105059731736645e-06,
+      "loss": 0.2244,
+      "step": 465
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 0.1370282769203186,
+      "learning_rate": 6.669160341214265e-06,
+      "loss": 0.2175,
+      "step": 466
+    },
+    {
+      "epoch": 1.8688688688688688,
+      "grad_norm": 0.13752111792564392,
+      "learning_rate": 6.627878655414311e-06,
+      "loss": 0.2211,
+      "step": 467
+    },
+    {
+      "epoch": 1.8728728728728727,
+      "grad_norm": 0.1328887790441513,
+      "learning_rate": 6.586661708308273e-06,
+      "loss": 0.2147,
+      "step": 468
+    },
+    {
+      "epoch": 1.8768768768768769,
+      "grad_norm": 0.13455736637115479,
+      "learning_rate": 6.5455102911877665e-06,
+      "loss": 0.2226,
+      "step": 469
+    },
+    {
+      "epoch": 1.880880880880881,
+      "grad_norm": 0.1263863891363144,
+      "learning_rate": 6.504425194086334e-06,
+      "loss": 0.2085,
+      "step": 470
+    },
+    {
+      "epoch": 1.884884884884885,
+      "grad_norm": 0.14515650272369385,
+      "learning_rate": 6.4634072057643045e-06,
+      "loss": 0.2175,
+      "step": 471
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.12257494777441025,
+      "learning_rate": 6.422457113693633e-06,
+      "loss": 0.2098,
+      "step": 472
+    },
+    {
+      "epoch": 1.8928928928928928,
+      "grad_norm": 0.1487148255109787,
+      "learning_rate": 6.381575704042792e-06,
+      "loss": 0.2098,
+      "step": 473
+    },
+    {
+      "epoch": 1.896896896896897,
+      "grad_norm": 0.13246439397335052,
+      "learning_rate": 6.340763761661665e-06,
+      "loss": 0.2169,
+      "step": 474
+    },
+    {
+      "epoch": 1.900900900900901,
+      "grad_norm": 0.13017022609710693,
+      "learning_rate": 6.3000220700664985e-06,
+      "loss": 0.2119,
+      "step": 475
+    },
+    {
+      "epoch": 1.904904904904905,
+      "grad_norm": 0.13715329766273499,
+      "learning_rate": 6.259351411424849e-06,
+      "loss": 0.2291,
+      "step": 476
+    },
+    {
+      "epoch": 1.9089089089089089,
+      "grad_norm": 0.12503132224082947,
+      "learning_rate": 6.218752566540555e-06,
+      "loss": 0.2172,
+      "step": 477
+    },
+    {
+      "epoch": 1.9129129129129128,
+      "grad_norm": 0.13519714772701263,
+      "learning_rate": 6.17822631483878e-06,
+      "loss": 0.2085,
+      "step": 478
+    },
+    {
+      "epoch": 1.916916916916917,
+      "grad_norm": 0.12226010113954544,
+      "learning_rate": 6.137773434351009e-06,
+      "loss": 0.1983,
+      "step": 479
+    },
+    {
+      "epoch": 1.920920920920921,
+      "grad_norm": 0.14085441827774048,
+      "learning_rate": 6.097394701700146e-06,
+      "loss": 0.2175,
+      "step": 480
+    },
+    {
+      "epoch": 1.924924924924925,
+      "grad_norm": 0.1396440863609314,
+      "learning_rate": 6.057090892085571e-06,
+      "loss": 0.2281,
+      "step": 481
+    },
+    {
+      "epoch": 1.9289289289289289,
+      "grad_norm": 0.13764607906341553,
+      "learning_rate": 6.016862779268301e-06,
+      "loss": 0.2241,
+      "step": 482
+    },
+    {
+      "epoch": 1.9329329329329328,
+      "grad_norm": 0.1312662959098816,
+      "learning_rate": 5.976711135556086e-06,
+      "loss": 0.2237,
+      "step": 483
+    },
+    {
+      "epoch": 1.936936936936937,
+      "grad_norm": 0.15344005823135376,
+      "learning_rate": 5.936636731788621e-06,
+      "loss": 0.215,
+      "step": 484
+    },
+    {
+      "epoch": 1.940940940940941,
+      "grad_norm": 0.1357364058494568,
+      "learning_rate": 5.896640337322725e-06,
+      "loss": 0.2284,
+      "step": 485
+    },
+    {
+      "epoch": 1.944944944944945,
+      "grad_norm": 0.14679481089115143,
+      "learning_rate": 5.8567227200175865e-06,
+      "loss": 0.2471,
+      "step": 486
+    },
+    {
+      "epoch": 1.9489489489489489,
+      "grad_norm": 0.13242673873901367,
+      "learning_rate": 5.816884646219997e-06,
+      "loss": 0.2207,
+      "step": 487
+    },
+    {
+      "epoch": 1.9529529529529528,
+      "grad_norm": 0.13282544910907745,
+      "learning_rate": 5.7771268807496794e-06,
+      "loss": 0.2133,
+      "step": 488
+    },
+    {
+      "epoch": 1.956956956956957,
+      "grad_norm": 0.12988749146461487,
+      "learning_rate": 5.737450186884555e-06,
+      "loss": 0.2119,
+      "step": 489
+    },
+    {
+      "epoch": 1.960960960960961,
+      "grad_norm": 0.1413298100233078,
+      "learning_rate": 5.6978553263461265e-06,
+      "loss": 0.2197,
+      "step": 490
+    },
+    {
+      "epoch": 1.964964964964965,
+      "grad_norm": 0.1392560601234436,
+      "learning_rate": 5.6583430592848565e-06,
+      "loss": 0.2118,
+      "step": 491
+    },
+    {
+      "epoch": 1.968968968968969,
+      "grad_norm": 0.1412695348262787,
+      "learning_rate": 5.618914144265532e-06,
+      "loss": 0.2146,
+      "step": 492
+    },
+    {
+      "epoch": 1.972972972972973,
+      "grad_norm": 0.14450322091579437,
+      "learning_rate": 5.579569338252758e-06,
+      "loss": 0.214,
+      "step": 493
+    },
+    {
+      "epoch": 1.976976976976977,
+      "grad_norm": 0.1321956366300583,
+      "learning_rate": 5.5403093965963806e-06,
+      "loss": 0.2171,
+      "step": 494
+    },
+    {
+      "epoch": 1.980980980980981,
+      "grad_norm": 0.15199291706085205,
+      "learning_rate": 5.501135073017008e-06,
+      "loss": 0.2186,
+      "step": 495
+    },
+    {
+      "epoch": 1.984984984984985,
+      "grad_norm": 0.13251198828220367,
+      "learning_rate": 5.4620471195915304e-06,
+      "loss": 0.2041,
+      "step": 496
+    },
+    {
+      "epoch": 1.988988988988989,
+      "grad_norm": 0.16512924432754517,
+      "learning_rate": 5.42304628673869e-06,
+      "loss": 0.2219,
+      "step": 497
+    },
+    {
+      "epoch": 1.992992992992993,
+      "grad_norm": 0.13640455901622772,
+      "learning_rate": 5.384133323204666e-06,
+      "loss": 0.2159,
+      "step": 498
+    },
+    {
+      "epoch": 1.992992992992993,
+      "eval_loss": 0.2962433695793152,
+      "eval_runtime": 6.2218,
+      "eval_samples_per_second": 13.019,
+      "eval_steps_per_second": 1.768,
+      "step": 498
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 747,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 249,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.308189414870221e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

3b-w-cot/checkpoint-498/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ae10bafaded3f1f05741f3f17290afb1efc74a263062c027a77525fa9902f1e
+size 10744

3b-w-cot/checkpoint-498/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

3b-w-cot/checkpoint-747/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
+}

3b-w-cot/checkpoint-747/config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "Qwen/Qwen2.5-3B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 70,
+  "model_type": "qwen2",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 36,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.3",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

3b-w-cot/checkpoint-747/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.48.3"
+}

3b-w-cot/checkpoint-747/latest

@@ -0,0 +1 @@
+global_step746

3b-w-cot/checkpoint-747/model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d864a3ede4291dbf8954edc730f490c139147b7997a05c84493a7ae77553fe0d
+size 4957560304

3b-w-cot/checkpoint-747/model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7732a9d9c741b7b288991f75baa7da5bd2ef0ace964fe0a2a63a19f001ff6734
+size 1836696752

3b-w-cot/checkpoint-747/model.safetensors.index.json

@@ -0,0 +1,442 @@
+{
+  "metadata": {
+    "total_size": 6794207232
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00002-of-00002.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.bias": "model-00002-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

3b-w-cot/checkpoint-747/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71792d9986abf333291d25825245eca92628cefa5f54c3852cce3ae98163a606
+size 14512

3b-w-cot/checkpoint-747/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8746eb25faee6c63bdad38e5ccce008abfe09b6a67f278aadc8f5b3e48f5a137
+size 14512

3b-w-cot/checkpoint-747/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b3a54b902ddde79dee7411c1b56d36e553d20b71c14e027984774fd8aa1d553
+size 1064

3b-w-cot/checkpoint-747/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": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}