Upload folder using huggingface_hub

README.md

@@ -0,0 +1,138 @@
+---
+library_name: peft
+license: apache-2.0
+base_model: Qwen/Qwen2.5-7B-Instruct
+tags:
+- generated_from_trainer
+datasets:
+- ugaoo/medqa_train
+model-index:
+- name: out/Qwen_Qwen2.5_7B_Instruct_ugaoo_medqa_train
+  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: `0.8.0.dev0`
+```yaml
+base_model: Qwen/Qwen2.5-7B-Instruct
+model_type: AutoModelForCausalLM
+tokenizer_type: AutoTokenizer
+trust_remote_code: true
+
+load_in_8bit: false
+load_in_4bit: true
+strict: false
+
+datasets:
+  - path: ugaoo/medqa_train
+    type: alpaca
+val_set_size: 0
+output_dir: ./out/Qwen_Qwen2.5_7B_Instruct_ugaoo_medqa_train
+
+sequence_len: 4000
+sample_packing: true
+pad_to_sequence_len: true
+
+adapter: qlora
+lora_r: 256
+lora_alpha: 512
+lora_dropout: 0.05
+lora_target_linear: true
+lora_target_modules:
+  - q_proj
+  - k_proj
+  - v_proj
+  - o_proj
+  - up_proj
+  - down_proj
+  - gate_proj
+ 
+wandb_project: testsearch
+wandb_entity:
+wandb_watch: 
+wandb_name: Qwen_Qwen2.5_7B_Instruct_ugaoo_medqa_train
+wandb_log_model:
+
+gradient_accumulation_steps: 3
+micro_batch_size: 4
+num_epochs: 6
+optimizer: adamw_torch
+lr_scheduler: cosine
+learning_rate: 5e-6
+
+train_on_inputs: false
+group_by_length: false
+bf16: auto
+fp16: false
+tf32: false
+
+gradient_checkpointing: true
+early_stopping_patience:
+resume_from_checkpoint:
+logging_steps: 1
+xformers_attention:
+flash_attention: true
+
+warmup_steps: 100
+evals_per_epoch: 6
+eval_table_size:
+saves_per_epoch: 1
+debug:
+deepspeed:
+weight_decay: 0.0
+fsdp:
+fsdp_config:
+save_total_limit: 6
+```
+
+</details><br>
+
+# out/Qwen_Qwen2.5_7B_Instruct_ugaoo_medqa_train
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the ugaoo/medqa_train dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 5e-06
+- train_batch_size: 4
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 3
+- total_train_batch_size: 12
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 100
+- num_epochs: 6.0
+
+### Training results
+
+
+
+### Framework versions
+
+- PEFT 0.14.0
+- Transformers 4.49.0
+- Pytorch 2.5.1+cu124
+- Datasets 3.2.0
+- Tokenizers 0.21.0

adapter_config.json

@@ -6,7 +6,7 @@
   "eva_config": null,
   "exclude_modules": null,
   "fan_in_fan_out": null,
-  "inference_mode": false,
+  "inference_mode": true,
   "init_lora_weights": true,
   "layer_replication": null,
   "layers_pattern": null,

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:530d8bfe012a1060fee5f3afb02770e9c499f0365acb31c6150abed5e723293a
+size 1291899552

checkpoint-168/README.md

@@ -0,0 +1,202 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.14.0

checkpoint-168/adapter_config.json

@@ -0,0 +1,37 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 512,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "v_proj",
+    "q_proj",
+    "up_proj",
+    "down_proj",
+    "k_proj",
+    "o_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-168/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:156d7eb2d4b441b266fef248ac659671b0b2b5744936bb9c2b13268a5b079d6e
+size 1291899552

checkpoint-168/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoint-168/global_step168/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71da898a176b195a8efe25872ad3f89b1af9285caf20dbacf4efc52d89e33189
+size 7751120668

checkpoint-168/global_step168/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3308ecb90d400b80b2699540e1f95fa8e85611de14f25401d27b63f37d1916ce
+size 1396311192

checkpoint-168/latest

@@ -0,0 +1 @@
+global_step168

checkpoint-168/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23510c2691f4817baac34aa7d9ef9863e2d975c56be7de7b139e8f4a6feff1f2
+size 14244

checkpoint-168/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e7b47f19470c49545e8b9d7f687db153957f1e1eb9baefe8460577ae6653930
+size 1064

checkpoint-168/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
+  }
+}

checkpoint-168/tokenizer.json

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

checkpoint-168/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
+}

checkpoint-168/trainer_state.json

@@ -0,0 +1,1209 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 168,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.023809523809523808,
+      "grad_norm": 26.742956161499023,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 4.1563,
+      "step": 1
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 26.336389541625977,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 4.0633,
+      "step": 2
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 28.836008071899414,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 4.3575,
+      "step": 3
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 27.459413528442383,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 4.1943,
+      "step": 4
+    },
+    {
+      "epoch": 0.11904761904761904,
+      "grad_norm": 27.914960861206055,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 4.3336,
+      "step": 5
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 27.47397232055664,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 4.2354,
+      "step": 6
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 26.212989807128906,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 4.1001,
+      "step": 7
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 26.80431365966797,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 4.1827,
+      "step": 8
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 27.641605377197266,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 4.2596,
+      "step": 9
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 27.783071517944336,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 4.2694,
+      "step": 10
+    },
+    {
+      "epoch": 0.2619047619047619,
+      "grad_norm": 26.553335189819336,
+      "learning_rate": 5.5e-07,
+      "loss": 4.0739,
+      "step": 11
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 26.77140998840332,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 4.1049,
+      "step": 12
+    },
+    {
+      "epoch": 0.30952380952380953,
+      "grad_norm": 25.07087516784668,
+      "learning_rate": 6.5e-07,
+      "loss": 3.8982,
+      "step": 13
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 25.834062576293945,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 4.0768,
+      "step": 14
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 25.00474739074707,
+      "learning_rate": 7.5e-07,
+      "loss": 4.0104,
+      "step": 15
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 23.692975997924805,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 3.8469,
+      "step": 16
+    },
+    {
+      "epoch": 0.40476190476190477,
+      "grad_norm": 22.197919845581055,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 3.7738,
+      "step": 17
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 20.92680549621582,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 3.6514,
+      "step": 18
+    },
+    {
+      "epoch": 0.4523809523809524,
+      "grad_norm": 20.251178741455078,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 3.6739,
+      "step": 19
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 17.55536460876465,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 3.2675,
+      "step": 20
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.066797256469727,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 3.4232,
+      "step": 21
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 16.1475887298584,
+      "learning_rate": 1.1e-06,
+      "loss": 3.1988,
+      "step": 22
+    },
+    {
+      "epoch": 0.5476190476190477,
+      "grad_norm": 15.61026382446289,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 3.1338,
+      "step": 23
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 15.409480094909668,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.9836,
+      "step": 24
+    },
+    {
+      "epoch": 0.5952380952380952,
+      "grad_norm": 15.391901969909668,
+      "learning_rate": 1.25e-06,
+      "loss": 2.9064,
+      "step": 25
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 16.92401885986328,
+      "learning_rate": 1.3e-06,
+      "loss": 2.906,
+      "step": 26
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 17.880958557128906,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 2.702,
+      "step": 27
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 18.114517211914062,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.4876,
+      "step": 28
+    },
+    {
+      "epoch": 0.6904761904761905,
+      "grad_norm": 17.608840942382812,
+      "learning_rate": 1.45e-06,
+      "loss": 2.2996,
+      "step": 29
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 17.055673599243164,
+      "learning_rate": 1.5e-06,
+      "loss": 2.2709,
+      "step": 30
+    },
+    {
+      "epoch": 0.7380952380952381,
+      "grad_norm": 14.92151927947998,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 2.0406,
+      "step": 31
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 13.657073020935059,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.8564,
+      "step": 32
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 13.274576187133789,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 1.7382,
+      "step": 33
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 13.728348731994629,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.6629,
+      "step": 34
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 13.521151542663574,
+      "learning_rate": 1.75e-06,
+      "loss": 1.5301,
+      "step": 35
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 13.709525108337402,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.4122,
+      "step": 36
+    },
+    {
+      "epoch": 0.8809523809523809,
+      "grad_norm": 12.587928771972656,
+      "learning_rate": 1.85e-06,
+      "loss": 1.2014,
+      "step": 37
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 13.494888305664062,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.1793,
+      "step": 38
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 13.417922019958496,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.0282,
+      "step": 39
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 12.754359245300293,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8525,
+      "step": 40
+    },
+    {
+      "epoch": 0.9761904761904762,
+      "grad_norm": 12.766407012939453,
+      "learning_rate": 2.05e-06,
+      "loss": 0.7043,
+      "step": 41
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.440421104431152,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.6245,
+      "step": 42
+    },
+    {
+      "epoch": 1.0238095238095237,
+      "grad_norm": 9.672205924987793,
+      "learning_rate": 2.15e-06,
+      "loss": 0.4768,
+      "step": 43
+    },
+    {
+      "epoch": 1.0476190476190477,
+      "grad_norm": 7.8501057624816895,
+      "learning_rate": 2.2e-06,
+      "loss": 0.3679,
+      "step": 44
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 6.751816749572754,
+      "learning_rate": 2.25e-06,
+      "loss": 0.2708,
+      "step": 45
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 5.267884731292725,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.2143,
+      "step": 46
+    },
+    {
+      "epoch": 1.119047619047619,
+      "grad_norm": 3.251101016998291,
+      "learning_rate": 2.35e-06,
+      "loss": 0.1759,
+      "step": 47
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 2.585360050201416,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.1855,
+      "step": 48
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 2.0107483863830566,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.1251,
+      "step": 49
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 1.941689133644104,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1441,
+      "step": 50
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 1.4036344289779663,
+      "learning_rate": 2.55e-06,
+      "loss": 0.1097,
+      "step": 51
+    },
+    {
+      "epoch": 1.2380952380952381,
+      "grad_norm": 1.2856179475784302,
+      "learning_rate": 2.6e-06,
+      "loss": 0.1303,
+      "step": 52
+    },
+    {
+      "epoch": 1.2619047619047619,
+      "grad_norm": 1.3184994459152222,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.1021,
+      "step": 53
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 1.144294261932373,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.115,
+      "step": 54
+    },
+    {
+      "epoch": 1.3095238095238095,
+      "grad_norm": 0.9276831150054932,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.1028,
+      "step": 55
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.9152742028236389,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.0979,
+      "step": 56
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.8525308966636658,
+      "learning_rate": 2.85e-06,
+      "loss": 0.0908,
+      "step": 57
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.9806348085403442,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0817,
+      "step": 58
+    },
+    {
+      "epoch": 1.4047619047619047,
+      "grad_norm": 0.606792151927948,
+      "learning_rate": 2.95e-06,
+      "loss": 0.0904,
+      "step": 59
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.7274054884910583,
+      "learning_rate": 3e-06,
+      "loss": 0.0811,
+      "step": 60
+    },
+    {
+      "epoch": 1.4523809523809523,
+      "grad_norm": 1.0523946285247803,
+      "learning_rate": 3.05e-06,
+      "loss": 0.0881,
+      "step": 61
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 0.5840473175048828,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.0848,
+      "step": 62
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7410831451416016,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.0788,
+      "step": 63
+    },
+    {
+      "epoch": 1.5238095238095237,
+      "grad_norm": 0.828996479511261,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.0866,
+      "step": 64
+    },
+    {
+      "epoch": 1.5476190476190477,
+      "grad_norm": 0.7505109310150146,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.076,
+      "step": 65
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.9672189354896545,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.0765,
+      "step": 66
+    },
+    {
+      "epoch": 1.5952380952380953,
+      "grad_norm": 0.5925746560096741,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.0706,
+      "step": 67
+    },
+    {
+      "epoch": 1.619047619047619,
+      "grad_norm": 0.6671133637428284,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0713,
+      "step": 68
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.5542609095573425,
+      "learning_rate": 3.45e-06,
+      "loss": 0.0715,
+      "step": 69
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.541200578212738,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0701,
+      "step": 70
+    },
+    {
+      "epoch": 1.6904761904761905,
+      "grad_norm": 0.4222320020198822,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.0669,
+      "step": 71
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.717410147190094,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.0782,
+      "step": 72
+    },
+    {
+      "epoch": 1.7380952380952381,
+      "grad_norm": 0.6776471734046936,
+      "learning_rate": 3.65e-06,
+      "loss": 0.0695,
+      "step": 73
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 0.5480474829673767,
+      "learning_rate": 3.7e-06,
+      "loss": 0.0662,
+      "step": 74
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.4779343605041504,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.0743,
+      "step": 75
+    },
+    {
+      "epoch": 1.8095238095238095,
+      "grad_norm": 0.43138471245765686,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.0666,
+      "step": 76
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.6058762669563293,
+      "learning_rate": 3.85e-06,
+      "loss": 0.0696,
+      "step": 77
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 1.3352755308151245,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0891,
+      "step": 78
+    },
+    {
+      "epoch": 1.880952380952381,
+      "grad_norm": 0.5319089293479919,
+      "learning_rate": 3.95e-06,
+      "loss": 0.0617,
+      "step": 79
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.5629184246063232,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0622,
+      "step": 80
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.37953704595565796,
+      "learning_rate": 4.05e-06,
+      "loss": 0.0676,
+      "step": 81
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.37576770782470703,
+      "learning_rate": 4.1e-06,
+      "loss": 0.0719,
+      "step": 82
+    },
+    {
+      "epoch": 1.9761904761904763,
+      "grad_norm": 0.4720636010169983,
+      "learning_rate": 4.15e-06,
+      "loss": 0.0662,
+      "step": 83
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4793304204940796,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.0664,
+      "step": 84
+    },
+    {
+      "epoch": 2.0238095238095237,
+      "grad_norm": 0.44540268182754517,
+      "learning_rate": 4.25e-06,
+      "loss": 0.0627,
+      "step": 85
+    },
+    {
+      "epoch": 2.0476190476190474,
+      "grad_norm": 0.40029338002204895,
+      "learning_rate": 4.3e-06,
+      "loss": 0.0615,
+      "step": 86
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.44975095987319946,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.0556,
+      "step": 87
+    },
+    {
+      "epoch": 2.0952380952380953,
+      "grad_norm": 0.49270153045654297,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0509,
+      "step": 88
+    },
+    {
+      "epoch": 2.119047619047619,
+      "grad_norm": 0.5457497239112854,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.0654,
+      "step": 89
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.5662055611610413,
+      "learning_rate": 4.5e-06,
+      "loss": 0.0584,
+      "step": 90
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.4589175581932068,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.05,
+      "step": 91
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.4754093289375305,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.0409,
+      "step": 92
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.5139522552490234,
+      "learning_rate": 4.65e-06,
+      "loss": 0.0548,
+      "step": 93
+    },
+    {
+      "epoch": 2.238095238095238,
+      "grad_norm": 0.6404117345809937,
+      "learning_rate": 4.7e-06,
+      "loss": 0.0575,
+      "step": 94
+    },
+    {
+      "epoch": 2.261904761904762,
+      "grad_norm": 0.4840497672557831,
+      "learning_rate": 4.75e-06,
+      "loss": 0.0545,
+      "step": 95
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.468397319316864,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.0532,
+      "step": 96
+    },
+    {
+      "epoch": 2.3095238095238093,
+      "grad_norm": 0.41116851568222046,
+      "learning_rate": 4.85e-06,
+      "loss": 0.0508,
+      "step": 97
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.4745863080024719,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0571,
+      "step": 98
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.5628818869590759,
+      "learning_rate": 4.95e-06,
+      "loss": 0.0527,
+      "step": 99
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.5258535146713257,
+      "learning_rate": 5e-06,
+      "loss": 0.0602,
+      "step": 100
+    },
+    {
+      "epoch": 2.4047619047619047,
+      "grad_norm": 0.4047287702560425,
+      "learning_rate": 4.999466041969828e-06,
+      "loss": 0.0511,
+      "step": 101
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.493963360786438,
+      "learning_rate": 4.997864395968252e-06,
+      "loss": 0.0496,
+      "step": 102
+    },
+    {
+      "epoch": 2.4523809523809526,
+      "grad_norm": 0.4637805223464966,
+      "learning_rate": 4.9951957461646705e-06,
+      "loss": 0.047,
+      "step": 103
+    },
+    {
+      "epoch": 2.4761904761904763,
+      "grad_norm": 0.4990857243537903,
+      "learning_rate": 4.991461232516675e-06,
+      "loss": 0.0566,
+      "step": 104
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.41547098755836487,
+      "learning_rate": 4.986662450283107e-06,
+      "loss": 0.0496,
+      "step": 105
+    },
+    {
+      "epoch": 2.5238095238095237,
+      "grad_norm": 0.5748231410980225,
+      "learning_rate": 4.9808014493426124e-06,
+      "loss": 0.0362,
+      "step": 106
+    },
+    {
+      "epoch": 2.5476190476190474,
+      "grad_norm": 0.481645792722702,
+      "learning_rate": 4.973880733318007e-06,
+      "loss": 0.0418,
+      "step": 107
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.49727940559387207,
+      "learning_rate": 4.965903258506806e-06,
+      "loss": 0.0436,
+      "step": 108
+    },
+    {
+      "epoch": 2.5952380952380953,
+      "grad_norm": 0.6753897070884705,
+      "learning_rate": 4.956872432618399e-06,
+      "loss": 0.0486,
+      "step": 109
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.6681185960769653,
+      "learning_rate": 4.9467921133183864e-06,
+      "loss": 0.0426,
+      "step": 110
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.6316296458244324,
+      "learning_rate": 4.935666606580719e-06,
+      "loss": 0.0377,
+      "step": 111
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.43960002064704895,
+      "learning_rate": 4.923500664848327e-06,
+      "loss": 0.0355,
+      "step": 112
+    },
+    {
+      "epoch": 2.6904761904761907,
+      "grad_norm": 0.437860906124115,
+      "learning_rate": 4.910299485003034e-06,
+      "loss": 0.0369,
+      "step": 113
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.5523399710655212,
+      "learning_rate": 4.896068706145632e-06,
+      "loss": 0.0488,
+      "step": 114
+    },
+    {
+      "epoch": 2.738095238095238,
+      "grad_norm": 0.6876885890960693,
+      "learning_rate": 4.880814407187037e-06,
+      "loss": 0.0492,
+      "step": 115
+    },
+    {
+      "epoch": 2.761904761904762,
+      "grad_norm": 0.48951950669288635,
+      "learning_rate": 4.864543104251587e-06,
+      "loss": 0.0464,
+      "step": 116
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.6003612875938416,
+      "learning_rate": 4.8472617478935744e-06,
+      "loss": 0.0376,
+      "step": 117
+    },
+    {
+      "epoch": 2.8095238095238093,
+      "grad_norm": 0.5665768384933472,
+      "learning_rate": 4.828977720128198e-06,
+      "loss": 0.0364,
+      "step": 118
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.6751538515090942,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.0415,
+      "step": 119
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.6295377612113953,
+      "learning_rate": 4.789433316637644e-06,
+      "loss": 0.0364,
+      "step": 120
+    },
+    {
+      "epoch": 2.880952380952381,
+      "grad_norm": 0.4297006130218506,
+      "learning_rate": 4.7681898329539004e-06,
+      "loss": 0.0282,
+      "step": 121
+    },
+    {
+      "epoch": 2.9047619047619047,
+      "grad_norm": 0.48039913177490234,
+      "learning_rate": 4.745977454729947e-06,
+      "loss": 0.0319,
+      "step": 122
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.5199178457260132,
+      "learning_rate": 4.722805670347963e-06,
+      "loss": 0.0299,
+      "step": 123
+    },
+    {
+      "epoch": 2.9523809523809526,
+      "grad_norm": 0.5075429677963257,
+      "learning_rate": 4.698684378016223e-06,
+      "loss": 0.0265,
+      "step": 124
+    },
+    {
+      "epoch": 2.9761904761904763,
+      "grad_norm": 0.5364718437194824,
+      "learning_rate": 4.673623881540917e-06,
+      "loss": 0.0351,
+      "step": 125
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5693520903587341,
+      "learning_rate": 4.647634885924713e-06,
+      "loss": 0.0286,
+      "step": 126
+    },
+    {
+      "epoch": 3.0238095238095237,
+      "grad_norm": 0.502586305141449,
+      "learning_rate": 4.620728492793934e-06,
+      "loss": 0.0222,
+      "step": 127
+    },
+    {
+      "epoch": 3.0476190476190474,
+      "grad_norm": 0.42680972814559937,
+      "learning_rate": 4.592916195656322e-06,
+      "loss": 0.0183,
+      "step": 128
+    },
+    {
+      "epoch": 3.0714285714285716,
+      "grad_norm": 0.6365283131599426,
+      "learning_rate": 4.56420987499139e-06,
+      "loss": 0.025,
+      "step": 129
+    },
+    {
+      "epoch": 3.0952380952380953,
+      "grad_norm": 0.4085887670516968,
+      "learning_rate": 4.534621793175488e-06,
+      "loss": 0.0166,
+      "step": 130
+    },
+    {
+      "epoch": 3.119047619047619,
+      "grad_norm": 0.4094793498516083,
+      "learning_rate": 4.504164589243721e-06,
+      "loss": 0.0158,
+      "step": 131
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.6991594433784485,
+      "learning_rate": 4.472851273490985e-06,
+      "loss": 0.0253,
+      "step": 132
+    },
+    {
+      "epoch": 3.1666666666666665,
+      "grad_norm": 0.4174569547176361,
+      "learning_rate": 4.440695221914394e-06,
+      "loss": 0.0165,
+      "step": 133
+    },
+    {
+      "epoch": 3.1904761904761907,
+      "grad_norm": 0.4691311717033386,
+      "learning_rate": 4.407710170499517e-06,
+      "loss": 0.0187,
+      "step": 134
+    },
+    {
+      "epoch": 3.2142857142857144,
+      "grad_norm": 0.6574212312698364,
+      "learning_rate": 4.373910209352816e-06,
+      "loss": 0.018,
+      "step": 135
+    },
+    {
+      "epoch": 3.238095238095238,
+      "grad_norm": 0.40424057841300964,
+      "learning_rate": 4.33930977668283e-06,
+      "loss": 0.0115,
+      "step": 136
+    },
+    {
+      "epoch": 3.261904761904762,
+      "grad_norm": 0.5574136972427368,
+      "learning_rate": 4.303923652632656e-06,
+      "loss": 0.0143,
+      "step": 137
+    },
+    {
+      "epoch": 3.2857142857142856,
+      "grad_norm": 0.7671014666557312,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.0123,
+      "step": 138
+    },
+    {
+      "epoch": 3.3095238095238093,
+      "grad_norm": 0.5041195750236511,
+      "learning_rate": 4.2308551226120745e-06,
+      "loss": 0.0105,
+      "step": 139
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.7552913427352905,
+      "learning_rate": 4.1932039290643534e-06,
+      "loss": 0.0132,
+      "step": 140
+    },
+    {
+      "epoch": 3.357142857142857,
+      "grad_norm": 0.4549916982650757,
+      "learning_rate": 4.154829455648916e-06,
+      "loss": 0.0073,
+      "step": 141
+    },
+    {
+      "epoch": 3.380952380952381,
+      "grad_norm": 0.5453060269355774,
+      "learning_rate": 4.115748094652352e-06,
+      "loss": 0.0065,
+      "step": 142
+    },
+    {
+      "epoch": 3.4047619047619047,
+      "grad_norm": 0.6225312948226929,
+      "learning_rate": 4.075976540319888e-06,
+      "loss": 0.0087,
+      "step": 143
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.629115641117096,
+      "learning_rate": 4.0355317817241705e-06,
+      "loss": 0.0118,
+      "step": 144
+    },
+    {
+      "epoch": 3.4523809523809526,
+      "grad_norm": 0.537651538848877,
+      "learning_rate": 3.994431095508102e-06,
+      "loss": 0.0066,
+      "step": 145
+    },
+    {
+      "epoch": 3.4761904761904763,
+      "grad_norm": 0.9304003715515137,
+      "learning_rate": 3.9526920385048465e-06,
+      "loss": 0.0082,
+      "step": 146
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.557716965675354,
+      "learning_rate": 3.9103324402381285e-06,
+      "loss": 0.0063,
+      "step": 147
+    },
+    {
+      "epoch": 3.5238095238095237,
+      "grad_norm": 0.3688383102416992,
+      "learning_rate": 3.8673703953060685e-06,
+      "loss": 0.0049,
+      "step": 148
+    },
+    {
+      "epoch": 3.5476190476190474,
+      "grad_norm": 0.44352808594703674,
+      "learning_rate": 3.8238242556517725e-06,
+      "loss": 0.0043,
+      "step": 149
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.5826281905174255,
+      "learning_rate": 3.779712622724003e-06,
+      "loss": 0.0063,
+      "step": 150
+    },
+    {
+      "epoch": 3.5952380952380953,
+      "grad_norm": 0.600183367729187,
+      "learning_rate": 3.7350543395312604e-06,
+      "loss": 0.0057,
+      "step": 151
+    },
+    {
+      "epoch": 3.619047619047619,
+      "grad_norm": 0.6133769750595093,
+      "learning_rate": 3.6898684825926845e-06,
+      "loss": 0.007,
+      "step": 152
+    },
+    {
+      "epoch": 3.642857142857143,
+      "grad_norm": 0.5637931227684021,
+      "learning_rate": 3.6441743537892045e-06,
+      "loss": 0.0059,
+      "step": 153
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 0.27519500255584717,
+      "learning_rate": 3.5979914721184263e-06,
+      "loss": 0.002,
+      "step": 154
+    },
+    {
+      "epoch": 3.6904761904761907,
+      "grad_norm": 0.3693721294403076,
+      "learning_rate": 3.551339565356769e-06,
+      "loss": 0.0042,
+      "step": 155
+    },
+    {
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.733023464679718,
+      "learning_rate": 3.5042385616324243e-06,
+      "loss": 0.0031,
+      "step": 156
+    },
+    {
+      "epoch": 3.738095238095238,
+      "grad_norm": 0.4825158417224884,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.0034,
+      "step": 157
+    },
+    {
+      "epoch": 3.761904761904762,
+      "grad_norm": 0.745896577835083,
+      "learning_rate": 3.4087699264095746e-06,
+      "loss": 0.006,
+      "step": 158
+    },
+    {
+      "epoch": 3.7857142857142856,
+      "grad_norm": 0.606246829032898,
+      "learning_rate": 3.360443075906597e-06,
+      "loss": 0.0028,
+      "step": 159
+    },
+    {
+      "epoch": 3.8095238095238093,
+      "grad_norm": 0.35037049651145935,
+      "learning_rate": 3.3117486730117092e-06,
+      "loss": 0.0022,
+      "step": 160
+    },
+    {
+      "epoch": 3.8333333333333335,
+      "grad_norm": 0.3873816132545471,
+      "learning_rate": 3.2627075183388725e-06,
+      "loss": 0.0024,
+      "step": 161
+    },
+    {
+      "epoch": 3.857142857142857,
+      "grad_norm": 0.2003440409898758,
+      "learning_rate": 3.2133405606227636e-06,
+      "loss": 0.0014,
+      "step": 162
+    },
+    {
+      "epoch": 3.880952380952381,
+      "grad_norm": 1.2881035804748535,
+      "learning_rate": 3.163668887770181e-06,
+      "loss": 0.0052,
+      "step": 163
+    },
+    {
+      "epoch": 3.9047619047619047,
+      "grad_norm": 0.16532732546329498,
+      "learning_rate": 3.1137137178519983e-06,
+      "loss": 0.0006,
+      "step": 164
+    },
+    {
+      "epoch": 3.928571428571429,
+      "grad_norm": 0.42558538913726807,
+      "learning_rate": 3.063496390039516e-06,
+      "loss": 0.0018,
+      "step": 165
+    },
+    {
+      "epoch": 3.9523809523809526,
+      "grad_norm": 0.5848471522331238,
+      "learning_rate": 3.013038355489086e-06,
+      "loss": 0.002,
+      "step": 166
+    },
+    {
+      "epoch": 3.9761904761904763,
+      "grad_norm": 0.18496190011501312,
+      "learning_rate": 2.9623611681788967e-06,
+      "loss": 0.0012,
+      "step": 167
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9864285588264465,
+      "learning_rate": 2.911486475701835e-06,
+      "loss": 0.0021,
+      "step": 168
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 252,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 42,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.7634402467341926e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-168/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcddbcdf99a4c44b8679b782b2f2c338b06bf1242af2189e08ac28c3f5a753ca
+size 7992

checkpoint-168/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/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 . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+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)
+
+        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 files:
+        state_dict = torch.load(f, map_location=device)
+        # 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}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) 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")
+
+
+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 = fp32_flat_groups[0].numel() * 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
+    for name, shape in param_shapes.items():
+
+        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}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        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 get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=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
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    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.
+
+    """
+    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")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, 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_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``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
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+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_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    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_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

checkpoint-210/README.md

@@ -0,0 +1,202 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.14.0

checkpoint-210/adapter_config.json

@@ -0,0 +1,37 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 512,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "v_proj",
+    "q_proj",
+    "up_proj",
+    "down_proj",
+    "k_proj",
+    "o_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-210/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fe390b27ddc982fb149767f003f279787a9a71f57dfc6faf5d93b798cb59bee
+size 1291899552

checkpoint-210/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoint-210/global_step210/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c75ea5309acfcf244df6f5e11185328b886c3f321560c405e595a62356a55133
+size 7751120668

checkpoint-210/global_step210/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08d1bf1f9a081b329dd8f9776d65c90fb929e2f594a81c7907d4614ba1eacaea
+size 1396311192

checkpoint-210/latest

@@ -0,0 +1 @@
+global_step210

checkpoint-210/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb0ef418b4ac5c8b78986b0e43e92dba5129dc78a0da0c8d61a6eab8451dc89d
+size 14244

checkpoint-210/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91731ef41aea5443f294c0fcd55e3484b4208ac30b16b66f89fa85dd75194b6d
+size 1064

checkpoint-210/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
+  }
+}

checkpoint-210/tokenizer.json

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

checkpoint-210/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
+}

checkpoint-210/trainer_state.json

@@ -0,0 +1,1503 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 210,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.023809523809523808,
+      "grad_norm": 26.742956161499023,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 4.1563,
+      "step": 1
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 26.336389541625977,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 4.0633,
+      "step": 2
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 28.836008071899414,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 4.3575,
+      "step": 3
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 27.459413528442383,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 4.1943,
+      "step": 4
+    },
+    {
+      "epoch": 0.11904761904761904,
+      "grad_norm": 27.914960861206055,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 4.3336,
+      "step": 5
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 27.47397232055664,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 4.2354,
+      "step": 6
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 26.212989807128906,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 4.1001,
+      "step": 7
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 26.80431365966797,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 4.1827,
+      "step": 8
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 27.641605377197266,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 4.2596,
+      "step": 9
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 27.783071517944336,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 4.2694,
+      "step": 10
+    },
+    {
+      "epoch": 0.2619047619047619,
+      "grad_norm": 26.553335189819336,
+      "learning_rate": 5.5e-07,
+      "loss": 4.0739,
+      "step": 11
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 26.77140998840332,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 4.1049,
+      "step": 12
+    },
+    {
+      "epoch": 0.30952380952380953,
+      "grad_norm": 25.07087516784668,
+      "learning_rate": 6.5e-07,
+      "loss": 3.8982,
+      "step": 13
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 25.834062576293945,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 4.0768,
+      "step": 14
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 25.00474739074707,
+      "learning_rate": 7.5e-07,
+      "loss": 4.0104,
+      "step": 15
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 23.692975997924805,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 3.8469,
+      "step": 16
+    },
+    {
+      "epoch": 0.40476190476190477,
+      "grad_norm": 22.197919845581055,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 3.7738,
+      "step": 17
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 20.92680549621582,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 3.6514,
+      "step": 18
+    },
+    {
+      "epoch": 0.4523809523809524,
+      "grad_norm": 20.251178741455078,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 3.6739,
+      "step": 19
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 17.55536460876465,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 3.2675,
+      "step": 20
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.066797256469727,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 3.4232,
+      "step": 21
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 16.1475887298584,
+      "learning_rate": 1.1e-06,
+      "loss": 3.1988,
+      "step": 22
+    },
+    {
+      "epoch": 0.5476190476190477,
+      "grad_norm": 15.61026382446289,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 3.1338,
+      "step": 23
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 15.409480094909668,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.9836,
+      "step": 24
+    },
+    {
+      "epoch": 0.5952380952380952,
+      "grad_norm": 15.391901969909668,
+      "learning_rate": 1.25e-06,
+      "loss": 2.9064,
+      "step": 25
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 16.92401885986328,
+      "learning_rate": 1.3e-06,
+      "loss": 2.906,
+      "step": 26
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 17.880958557128906,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 2.702,
+      "step": 27
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 18.114517211914062,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.4876,
+      "step": 28
+    },
+    {
+      "epoch": 0.6904761904761905,
+      "grad_norm": 17.608840942382812,
+      "learning_rate": 1.45e-06,
+      "loss": 2.2996,
+      "step": 29
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 17.055673599243164,
+      "learning_rate": 1.5e-06,
+      "loss": 2.2709,
+      "step": 30
+    },
+    {
+      "epoch": 0.7380952380952381,
+      "grad_norm": 14.92151927947998,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 2.0406,
+      "step": 31
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 13.657073020935059,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.8564,
+      "step": 32
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 13.274576187133789,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 1.7382,
+      "step": 33
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 13.728348731994629,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.6629,
+      "step": 34
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 13.521151542663574,
+      "learning_rate": 1.75e-06,
+      "loss": 1.5301,
+      "step": 35
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 13.709525108337402,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.4122,
+      "step": 36
+    },
+    {
+      "epoch": 0.8809523809523809,
+      "grad_norm": 12.587928771972656,
+      "learning_rate": 1.85e-06,
+      "loss": 1.2014,
+      "step": 37
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 13.494888305664062,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.1793,
+      "step": 38
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 13.417922019958496,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.0282,
+      "step": 39
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 12.754359245300293,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8525,
+      "step": 40
+    },
+    {
+      "epoch": 0.9761904761904762,
+      "grad_norm": 12.766407012939453,
+      "learning_rate": 2.05e-06,
+      "loss": 0.7043,
+      "step": 41
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.440421104431152,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.6245,
+      "step": 42
+    },
+    {
+      "epoch": 1.0238095238095237,
+      "grad_norm": 9.672205924987793,
+      "learning_rate": 2.15e-06,
+      "loss": 0.4768,
+      "step": 43
+    },
+    {
+      "epoch": 1.0476190476190477,
+      "grad_norm": 7.8501057624816895,
+      "learning_rate": 2.2e-06,
+      "loss": 0.3679,
+      "step": 44
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 6.751816749572754,
+      "learning_rate": 2.25e-06,
+      "loss": 0.2708,
+      "step": 45
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 5.267884731292725,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.2143,
+      "step": 46
+    },
+    {
+      "epoch": 1.119047619047619,
+      "grad_norm": 3.251101016998291,
+      "learning_rate": 2.35e-06,
+      "loss": 0.1759,
+      "step": 47
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 2.585360050201416,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.1855,
+      "step": 48
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 2.0107483863830566,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.1251,
+      "step": 49
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 1.941689133644104,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1441,
+      "step": 50
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 1.4036344289779663,
+      "learning_rate": 2.55e-06,
+      "loss": 0.1097,
+      "step": 51
+    },
+    {
+      "epoch": 1.2380952380952381,
+      "grad_norm": 1.2856179475784302,
+      "learning_rate": 2.6e-06,
+      "loss": 0.1303,
+      "step": 52
+    },
+    {
+      "epoch": 1.2619047619047619,
+      "grad_norm": 1.3184994459152222,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.1021,
+      "step": 53
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 1.144294261932373,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.115,
+      "step": 54
+    },
+    {
+      "epoch": 1.3095238095238095,
+      "grad_norm": 0.9276831150054932,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.1028,
+      "step": 55
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.9152742028236389,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.0979,
+      "step": 56
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.8525308966636658,
+      "learning_rate": 2.85e-06,
+      "loss": 0.0908,
+      "step": 57
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.9806348085403442,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0817,
+      "step": 58
+    },
+    {
+      "epoch": 1.4047619047619047,
+      "grad_norm": 0.606792151927948,
+      "learning_rate": 2.95e-06,
+      "loss": 0.0904,
+      "step": 59
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.7274054884910583,
+      "learning_rate": 3e-06,
+      "loss": 0.0811,
+      "step": 60
+    },
+    {
+      "epoch": 1.4523809523809523,
+      "grad_norm": 1.0523946285247803,
+      "learning_rate": 3.05e-06,
+      "loss": 0.0881,
+      "step": 61
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 0.5840473175048828,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.0848,
+      "step": 62
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7410831451416016,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.0788,
+      "step": 63
+    },
+    {
+      "epoch": 1.5238095238095237,
+      "grad_norm": 0.828996479511261,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.0866,
+      "step": 64
+    },
+    {
+      "epoch": 1.5476190476190477,
+      "grad_norm": 0.7505109310150146,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.076,
+      "step": 65
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.9672189354896545,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.0765,
+      "step": 66
+    },
+    {
+      "epoch": 1.5952380952380953,
+      "grad_norm": 0.5925746560096741,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.0706,
+      "step": 67
+    },
+    {
+      "epoch": 1.619047619047619,
+      "grad_norm": 0.6671133637428284,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0713,
+      "step": 68
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.5542609095573425,
+      "learning_rate": 3.45e-06,
+      "loss": 0.0715,
+      "step": 69
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.541200578212738,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0701,
+      "step": 70
+    },
+    {
+      "epoch": 1.6904761904761905,
+      "grad_norm": 0.4222320020198822,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.0669,
+      "step": 71
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.717410147190094,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.0782,
+      "step": 72
+    },
+    {
+      "epoch": 1.7380952380952381,
+      "grad_norm": 0.6776471734046936,
+      "learning_rate": 3.65e-06,
+      "loss": 0.0695,
+      "step": 73
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 0.5480474829673767,
+      "learning_rate": 3.7e-06,
+      "loss": 0.0662,
+      "step": 74
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.4779343605041504,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.0743,
+      "step": 75
+    },
+    {
+      "epoch": 1.8095238095238095,
+      "grad_norm": 0.43138471245765686,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.0666,
+      "step": 76
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.6058762669563293,
+      "learning_rate": 3.85e-06,
+      "loss": 0.0696,
+      "step": 77
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 1.3352755308151245,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0891,
+      "step": 78
+    },
+    {
+      "epoch": 1.880952380952381,
+      "grad_norm": 0.5319089293479919,
+      "learning_rate": 3.95e-06,
+      "loss": 0.0617,
+      "step": 79
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.5629184246063232,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0622,
+      "step": 80
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.37953704595565796,
+      "learning_rate": 4.05e-06,
+      "loss": 0.0676,
+      "step": 81
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.37576770782470703,
+      "learning_rate": 4.1e-06,
+      "loss": 0.0719,
+      "step": 82
+    },
+    {
+      "epoch": 1.9761904761904763,
+      "grad_norm": 0.4720636010169983,
+      "learning_rate": 4.15e-06,
+      "loss": 0.0662,
+      "step": 83
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4793304204940796,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.0664,
+      "step": 84
+    },
+    {
+      "epoch": 2.0238095238095237,
+      "grad_norm": 0.44540268182754517,
+      "learning_rate": 4.25e-06,
+      "loss": 0.0627,
+      "step": 85
+    },
+    {
+      "epoch": 2.0476190476190474,
+      "grad_norm": 0.40029338002204895,
+      "learning_rate": 4.3e-06,
+      "loss": 0.0615,
+      "step": 86
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.44975095987319946,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.0556,
+      "step": 87
+    },
+    {
+      "epoch": 2.0952380952380953,
+      "grad_norm": 0.49270153045654297,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0509,
+      "step": 88
+    },
+    {
+      "epoch": 2.119047619047619,
+      "grad_norm": 0.5457497239112854,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.0654,
+      "step": 89
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.5662055611610413,
+      "learning_rate": 4.5e-06,
+      "loss": 0.0584,
+      "step": 90
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.4589175581932068,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.05,
+      "step": 91
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.4754093289375305,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.0409,
+      "step": 92
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.5139522552490234,
+      "learning_rate": 4.65e-06,
+      "loss": 0.0548,
+      "step": 93
+    },
+    {
+      "epoch": 2.238095238095238,
+      "grad_norm": 0.6404117345809937,
+      "learning_rate": 4.7e-06,
+      "loss": 0.0575,
+      "step": 94
+    },
+    {
+      "epoch": 2.261904761904762,
+      "grad_norm": 0.4840497672557831,
+      "learning_rate": 4.75e-06,
+      "loss": 0.0545,
+      "step": 95
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.468397319316864,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.0532,
+      "step": 96
+    },
+    {
+      "epoch": 2.3095238095238093,
+      "grad_norm": 0.41116851568222046,
+      "learning_rate": 4.85e-06,
+      "loss": 0.0508,
+      "step": 97
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.4745863080024719,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0571,
+      "step": 98
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.5628818869590759,
+      "learning_rate": 4.95e-06,
+      "loss": 0.0527,
+      "step": 99
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.5258535146713257,
+      "learning_rate": 5e-06,
+      "loss": 0.0602,
+      "step": 100
+    },
+    {
+      "epoch": 2.4047619047619047,
+      "grad_norm": 0.4047287702560425,
+      "learning_rate": 4.999466041969828e-06,
+      "loss": 0.0511,
+      "step": 101
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.493963360786438,
+      "learning_rate": 4.997864395968252e-06,
+      "loss": 0.0496,
+      "step": 102
+    },
+    {
+      "epoch": 2.4523809523809526,
+      "grad_norm": 0.4637805223464966,
+      "learning_rate": 4.9951957461646705e-06,
+      "loss": 0.047,
+      "step": 103
+    },
+    {
+      "epoch": 2.4761904761904763,
+      "grad_norm": 0.4990857243537903,
+      "learning_rate": 4.991461232516675e-06,
+      "loss": 0.0566,
+      "step": 104
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.41547098755836487,
+      "learning_rate": 4.986662450283107e-06,
+      "loss": 0.0496,
+      "step": 105
+    },
+    {
+      "epoch": 2.5238095238095237,
+      "grad_norm": 0.5748231410980225,
+      "learning_rate": 4.9808014493426124e-06,
+      "loss": 0.0362,
+      "step": 106
+    },
+    {
+      "epoch": 2.5476190476190474,
+      "grad_norm": 0.481645792722702,
+      "learning_rate": 4.973880733318007e-06,
+      "loss": 0.0418,
+      "step": 107
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.49727940559387207,
+      "learning_rate": 4.965903258506806e-06,
+      "loss": 0.0436,
+      "step": 108
+    },
+    {
+      "epoch": 2.5952380952380953,
+      "grad_norm": 0.6753897070884705,
+      "learning_rate": 4.956872432618399e-06,
+      "loss": 0.0486,
+      "step": 109
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.6681185960769653,
+      "learning_rate": 4.9467921133183864e-06,
+      "loss": 0.0426,
+      "step": 110
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.6316296458244324,
+      "learning_rate": 4.935666606580719e-06,
+      "loss": 0.0377,
+      "step": 111
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.43960002064704895,
+      "learning_rate": 4.923500664848327e-06,
+      "loss": 0.0355,
+      "step": 112
+    },
+    {
+      "epoch": 2.6904761904761907,
+      "grad_norm": 0.437860906124115,
+      "learning_rate": 4.910299485003034e-06,
+      "loss": 0.0369,
+      "step": 113
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.5523399710655212,
+      "learning_rate": 4.896068706145632e-06,
+      "loss": 0.0488,
+      "step": 114
+    },
+    {
+      "epoch": 2.738095238095238,
+      "grad_norm": 0.6876885890960693,
+      "learning_rate": 4.880814407187037e-06,
+      "loss": 0.0492,
+      "step": 115
+    },
+    {
+      "epoch": 2.761904761904762,
+      "grad_norm": 0.48951950669288635,
+      "learning_rate": 4.864543104251587e-06,
+      "loss": 0.0464,
+      "step": 116
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.6003612875938416,
+      "learning_rate": 4.8472617478935744e-06,
+      "loss": 0.0376,
+      "step": 117
+    },
+    {
+      "epoch": 2.8095238095238093,
+      "grad_norm": 0.5665768384933472,
+      "learning_rate": 4.828977720128198e-06,
+      "loss": 0.0364,
+      "step": 118
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.6751538515090942,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.0415,
+      "step": 119
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.6295377612113953,
+      "learning_rate": 4.789433316637644e-06,
+      "loss": 0.0364,
+      "step": 120
+    },
+    {
+      "epoch": 2.880952380952381,
+      "grad_norm": 0.4297006130218506,
+      "learning_rate": 4.7681898329539004e-06,
+      "loss": 0.0282,
+      "step": 121
+    },
+    {
+      "epoch": 2.9047619047619047,
+      "grad_norm": 0.48039913177490234,
+      "learning_rate": 4.745977454729947e-06,
+      "loss": 0.0319,
+      "step": 122
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.5199178457260132,
+      "learning_rate": 4.722805670347963e-06,
+      "loss": 0.0299,
+      "step": 123
+    },
+    {
+      "epoch": 2.9523809523809526,
+      "grad_norm": 0.5075429677963257,
+      "learning_rate": 4.698684378016223e-06,
+      "loss": 0.0265,
+      "step": 124
+    },
+    {
+      "epoch": 2.9761904761904763,
+      "grad_norm": 0.5364718437194824,
+      "learning_rate": 4.673623881540917e-06,
+      "loss": 0.0351,
+      "step": 125
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5693520903587341,
+      "learning_rate": 4.647634885924713e-06,
+      "loss": 0.0286,
+      "step": 126
+    },
+    {
+      "epoch": 3.0238095238095237,
+      "grad_norm": 0.502586305141449,
+      "learning_rate": 4.620728492793934e-06,
+      "loss": 0.0222,
+      "step": 127
+    },
+    {
+      "epoch": 3.0476190476190474,
+      "grad_norm": 0.42680972814559937,
+      "learning_rate": 4.592916195656322e-06,
+      "loss": 0.0183,
+      "step": 128
+    },
+    {
+      "epoch": 3.0714285714285716,
+      "grad_norm": 0.6365283131599426,
+      "learning_rate": 4.56420987499139e-06,
+      "loss": 0.025,
+      "step": 129
+    },
+    {
+      "epoch": 3.0952380952380953,
+      "grad_norm": 0.4085887670516968,
+      "learning_rate": 4.534621793175488e-06,
+      "loss": 0.0166,
+      "step": 130
+    },
+    {
+      "epoch": 3.119047619047619,
+      "grad_norm": 0.4094793498516083,
+      "learning_rate": 4.504164589243721e-06,
+      "loss": 0.0158,
+      "step": 131
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.6991594433784485,
+      "learning_rate": 4.472851273490985e-06,
+      "loss": 0.0253,
+      "step": 132
+    },
+    {
+      "epoch": 3.1666666666666665,
+      "grad_norm": 0.4174569547176361,
+      "learning_rate": 4.440695221914394e-06,
+      "loss": 0.0165,
+      "step": 133
+    },
+    {
+      "epoch": 3.1904761904761907,
+      "grad_norm": 0.4691311717033386,
+      "learning_rate": 4.407710170499517e-06,
+      "loss": 0.0187,
+      "step": 134
+    },
+    {
+      "epoch": 3.2142857142857144,
+      "grad_norm": 0.6574212312698364,
+      "learning_rate": 4.373910209352816e-06,
+      "loss": 0.018,
+      "step": 135
+    },
+    {
+      "epoch": 3.238095238095238,
+      "grad_norm": 0.40424057841300964,
+      "learning_rate": 4.33930977668283e-06,
+      "loss": 0.0115,
+      "step": 136
+    },
+    {
+      "epoch": 3.261904761904762,
+      "grad_norm": 0.5574136972427368,
+      "learning_rate": 4.303923652632656e-06,
+      "loss": 0.0143,
+      "step": 137
+    },
+    {
+      "epoch": 3.2857142857142856,
+      "grad_norm": 0.7671014666557312,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.0123,
+      "step": 138
+    },
+    {
+      "epoch": 3.3095238095238093,
+      "grad_norm": 0.5041195750236511,
+      "learning_rate": 4.2308551226120745e-06,
+      "loss": 0.0105,
+      "step": 139
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.7552913427352905,
+      "learning_rate": 4.1932039290643534e-06,
+      "loss": 0.0132,
+      "step": 140
+    },
+    {
+      "epoch": 3.357142857142857,
+      "grad_norm": 0.4549916982650757,
+      "learning_rate": 4.154829455648916e-06,
+      "loss": 0.0073,
+      "step": 141
+    },
+    {
+      "epoch": 3.380952380952381,
+      "grad_norm": 0.5453060269355774,
+      "learning_rate": 4.115748094652352e-06,
+      "loss": 0.0065,
+      "step": 142
+    },
+    {
+      "epoch": 3.4047619047619047,
+      "grad_norm": 0.6225312948226929,
+      "learning_rate": 4.075976540319888e-06,
+      "loss": 0.0087,
+      "step": 143
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.629115641117096,
+      "learning_rate": 4.0355317817241705e-06,
+      "loss": 0.0118,
+      "step": 144
+    },
+    {
+      "epoch": 3.4523809523809526,
+      "grad_norm": 0.537651538848877,
+      "learning_rate": 3.994431095508102e-06,
+      "loss": 0.0066,
+      "step": 145
+    },
+    {
+      "epoch": 3.4761904761904763,
+      "grad_norm": 0.9304003715515137,
+      "learning_rate": 3.9526920385048465e-06,
+      "loss": 0.0082,
+      "step": 146
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.557716965675354,
+      "learning_rate": 3.9103324402381285e-06,
+      "loss": 0.0063,
+      "step": 147
+    },
+    {
+      "epoch": 3.5238095238095237,
+      "grad_norm": 0.3688383102416992,
+      "learning_rate": 3.8673703953060685e-06,
+      "loss": 0.0049,
+      "step": 148
+    },
+    {
+      "epoch": 3.5476190476190474,
+      "grad_norm": 0.44352808594703674,
+      "learning_rate": 3.8238242556517725e-06,
+      "loss": 0.0043,
+      "step": 149
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.5826281905174255,
+      "learning_rate": 3.779712622724003e-06,
+      "loss": 0.0063,
+      "step": 150
+    },
+    {
+      "epoch": 3.5952380952380953,
+      "grad_norm": 0.600183367729187,
+      "learning_rate": 3.7350543395312604e-06,
+      "loss": 0.0057,
+      "step": 151
+    },
+    {
+      "epoch": 3.619047619047619,
+      "grad_norm": 0.6133769750595093,
+      "learning_rate": 3.6898684825926845e-06,
+      "loss": 0.007,
+      "step": 152
+    },
+    {
+      "epoch": 3.642857142857143,
+      "grad_norm": 0.5637931227684021,
+      "learning_rate": 3.6441743537892045e-06,
+      "loss": 0.0059,
+      "step": 153
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 0.27519500255584717,
+      "learning_rate": 3.5979914721184263e-06,
+      "loss": 0.002,
+      "step": 154
+    },
+    {
+      "epoch": 3.6904761904761907,
+      "grad_norm": 0.3693721294403076,
+      "learning_rate": 3.551339565356769e-06,
+      "loss": 0.0042,
+      "step": 155
+    },
+    {
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.733023464679718,
+      "learning_rate": 3.5042385616324243e-06,
+      "loss": 0.0031,
+      "step": 156
+    },
+    {
+      "epoch": 3.738095238095238,
+      "grad_norm": 0.4825158417224884,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.0034,
+      "step": 157
+    },
+    {
+      "epoch": 3.761904761904762,
+      "grad_norm": 0.745896577835083,
+      "learning_rate": 3.4087699264095746e-06,
+      "loss": 0.006,
+      "step": 158
+    },
+    {
+      "epoch": 3.7857142857142856,
+      "grad_norm": 0.606246829032898,
+      "learning_rate": 3.360443075906597e-06,
+      "loss": 0.0028,
+      "step": 159
+    },
+    {
+      "epoch": 3.8095238095238093,
+      "grad_norm": 0.35037049651145935,
+      "learning_rate": 3.3117486730117092e-06,
+      "loss": 0.0022,
+      "step": 160
+    },
+    {
+      "epoch": 3.8333333333333335,
+      "grad_norm": 0.3873816132545471,
+      "learning_rate": 3.2627075183388725e-06,
+      "loss": 0.0024,
+      "step": 161
+    },
+    {
+      "epoch": 3.857142857142857,
+      "grad_norm": 0.2003440409898758,
+      "learning_rate": 3.2133405606227636e-06,
+      "loss": 0.0014,
+      "step": 162
+    },
+    {
+      "epoch": 3.880952380952381,
+      "grad_norm": 1.2881035804748535,
+      "learning_rate": 3.163668887770181e-06,
+      "loss": 0.0052,
+      "step": 163
+    },
+    {
+      "epoch": 3.9047619047619047,
+      "grad_norm": 0.16532732546329498,
+      "learning_rate": 3.1137137178519983e-06,
+      "loss": 0.0006,
+      "step": 164
+    },
+    {
+      "epoch": 3.928571428571429,
+      "grad_norm": 0.42558538913726807,
+      "learning_rate": 3.063496390039516e-06,
+      "loss": 0.0018,
+      "step": 165
+    },
+    {
+      "epoch": 3.9523809523809526,
+      "grad_norm": 0.5848471522331238,
+      "learning_rate": 3.013038355489086e-06,
+      "loss": 0.002,
+      "step": 166
+    },
+    {
+      "epoch": 3.9761904761904763,
+      "grad_norm": 0.18496190011501312,
+      "learning_rate": 2.9623611681788967e-06,
+      "loss": 0.0012,
+      "step": 167
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9864285588264465,
+      "learning_rate": 2.911486475701835e-06,
+      "loss": 0.0021,
+      "step": 168
+    },
+    {
+      "epoch": 4.023809523809524,
+      "grad_norm": 0.28038740158081055,
+      "learning_rate": 2.860436010018367e-06,
+      "loss": 0.0016,
+      "step": 169
+    },
+    {
+      "epoch": 4.0476190476190474,
+      "grad_norm": 0.174965500831604,
+      "learning_rate": 2.80923157817337e-06,
+      "loss": 0.0006,
+      "step": 170
+    },
+    {
+      "epoch": 4.071428571428571,
+      "grad_norm": 0.111990787088871,
+      "learning_rate": 2.7578950529808927e-06,
+      "loss": 0.0003,
+      "step": 171
+    },
+    {
+      "epoch": 4.095238095238095,
+      "grad_norm": 0.0902976393699646,
+      "learning_rate": 2.7064483636808314e-06,
+      "loss": 0.0004,
+      "step": 172
+    },
+    {
+      "epoch": 4.119047619047619,
+      "grad_norm": 0.23398929834365845,
+      "learning_rate": 2.654913486571487e-06,
+      "loss": 0.0007,
+      "step": 173
+    },
+    {
+      "epoch": 4.142857142857143,
+      "grad_norm": 0.3977113664150238,
+      "learning_rate": 2.603312435622033e-06,
+      "loss": 0.0012,
+      "step": 174
+    },
+    {
+      "epoch": 4.166666666666667,
+      "grad_norm": 0.18411217629909515,
+      "learning_rate": 2.5516672530688864e-06,
+      "loss": 0.0006,
+      "step": 175
+    },
+    {
+      "epoch": 4.190476190476191,
+      "grad_norm": 0.3521343171596527,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0017,
+      "step": 176
+    },
+    {
+      "epoch": 4.214285714285714,
+      "grad_norm": 0.1081409603357315,
+      "learning_rate": 2.448332746931115e-06,
+      "loss": 0.0004,
+      "step": 177
+    },
+    {
+      "epoch": 4.238095238095238,
+      "grad_norm": 0.10767076164484024,
+      "learning_rate": 2.396687564377967e-06,
+      "loss": 0.0003,
+      "step": 178
+    },
+    {
+      "epoch": 4.261904761904762,
+      "grad_norm": 0.0657818615436554,
+      "learning_rate": 2.345086513428514e-06,
+      "loss": 0.0003,
+      "step": 179
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.03795298561453819,
+      "learning_rate": 2.2935516363191695e-06,
+      "loss": 0.0002,
+      "step": 180
+    },
+    {
+      "epoch": 4.309523809523809,
+      "grad_norm": 0.09580468386411667,
+      "learning_rate": 2.2421049470191077e-06,
+      "loss": 0.0002,
+      "step": 181
+    },
+    {
+      "epoch": 4.333333333333333,
+      "grad_norm": 0.13761642575263977,
+      "learning_rate": 2.190768421826631e-06,
+      "loss": 0.0004,
+      "step": 182
+    },
+    {
+      "epoch": 4.357142857142857,
+      "grad_norm": 0.10877614468336105,
+      "learning_rate": 2.139563989981633e-06,
+      "loss": 0.0003,
+      "step": 183
+    },
+    {
+      "epoch": 4.380952380952381,
+      "grad_norm": 0.14662612974643707,
+      "learning_rate": 2.088513524298165e-06,
+      "loss": 0.0003,
+      "step": 184
+    },
+    {
+      "epoch": 4.404761904761905,
+      "grad_norm": 0.29953065514564514,
+      "learning_rate": 2.037638831821104e-06,
+      "loss": 0.0004,
+      "step": 185
+    },
+    {
+      "epoch": 4.428571428571429,
+      "grad_norm": 0.20808857679367065,
+      "learning_rate": 1.9869616445109146e-06,
+      "loss": 0.0003,
+      "step": 186
+    },
+    {
+      "epoch": 4.4523809523809526,
+      "grad_norm": 0.03407036140561104,
+      "learning_rate": 1.9365036099604853e-06,
+      "loss": 0.0002,
+      "step": 187
+    },
+    {
+      "epoch": 4.476190476190476,
+      "grad_norm": 0.3132052421569824,
+      "learning_rate": 1.8862862821480023e-06,
+      "loss": 0.0006,
+      "step": 188
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.2567755877971649,
+      "learning_rate": 1.83633111222982e-06,
+      "loss": 0.0007,
+      "step": 189
+    },
+    {
+      "epoch": 4.523809523809524,
+      "grad_norm": 0.11175425350666046,
+      "learning_rate": 1.7866594393772375e-06,
+      "loss": 0.0003,
+      "step": 190
+    },
+    {
+      "epoch": 4.5476190476190474,
+      "grad_norm": 0.594926118850708,
+      "learning_rate": 1.7372924816611283e-06,
+      "loss": 0.001,
+      "step": 191
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.03871627524495125,
+      "learning_rate": 1.6882513269882916e-06,
+      "loss": 0.0002,
+      "step": 192
+    },
+    {
+      "epoch": 4.595238095238095,
+      "grad_norm": 0.06972040235996246,
+      "learning_rate": 1.6395569240934042e-06,
+      "loss": 0.0002,
+      "step": 193
+    },
+    {
+      "epoch": 4.619047619047619,
+      "grad_norm": 0.057279668748378754,
+      "learning_rate": 1.5912300735904252e-06,
+      "loss": 0.0002,
+      "step": 194
+    },
+    {
+      "epoch": 4.642857142857143,
+      "grad_norm": 0.09757455438375473,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.0003,
+      "step": 195
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 0.1020384356379509,
+      "learning_rate": 1.495761438367577e-06,
+      "loss": 0.0003,
+      "step": 196
+    },
+    {
+      "epoch": 4.690476190476191,
+      "grad_norm": 0.08006434142589569,
+      "learning_rate": 1.4486604346432311e-06,
+      "loss": 0.0003,
+      "step": 197
+    },
+    {
+      "epoch": 4.714285714285714,
+      "grad_norm": 0.0277456846088171,
+      "learning_rate": 1.4020085278815745e-06,
+      "loss": 0.0001,
+      "step": 198
+    },
+    {
+      "epoch": 4.738095238095238,
+      "grad_norm": 0.018923059105873108,
+      "learning_rate": 1.3558256462107965e-06,
+      "loss": 0.0002,
+      "step": 199
+    },
+    {
+      "epoch": 4.761904761904762,
+      "grad_norm": 0.01575353741645813,
+      "learning_rate": 1.3101315174073162e-06,
+      "loss": 0.0001,
+      "step": 200
+    },
+    {
+      "epoch": 4.785714285714286,
+      "grad_norm": 0.014921327121555805,
+      "learning_rate": 1.2649456604687404e-06,
+      "loss": 0.0002,
+      "step": 201
+    },
+    {
+      "epoch": 4.809523809523809,
+      "grad_norm": 0.03840240091085434,
+      "learning_rate": 1.2202873772759983e-06,
+      "loss": 0.0002,
+      "step": 202
+    },
+    {
+      "epoch": 4.833333333333333,
+      "grad_norm": 0.04131830111145973,
+      "learning_rate": 1.1761757443482285e-06,
+      "loss": 0.0002,
+      "step": 203
+    },
+    {
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.02138919197022915,
+      "learning_rate": 1.1326296046939334e-06,
+      "loss": 0.0002,
+      "step": 204
+    },
+    {
+      "epoch": 4.880952380952381,
+      "grad_norm": 0.016792714595794678,
+      "learning_rate": 1.0896675597618725e-06,
+      "loss": 0.0001,
+      "step": 205
+    },
+    {
+      "epoch": 4.904761904761905,
+      "grad_norm": 0.017704278230667114,
+      "learning_rate": 1.0473079614951546e-06,
+      "loss": 0.0002,
+      "step": 206
+    },
+    {
+      "epoch": 4.928571428571429,
+      "grad_norm": 0.020070459693670273,
+      "learning_rate": 1.0055689044918979e-06,
+      "loss": 0.0001,
+      "step": 207
+    },
+    {
+      "epoch": 4.9523809523809526,
+      "grad_norm": 0.035451311618089676,
+      "learning_rate": 9.644682182758305e-07,
+      "loss": 0.0002,
+      "step": 208
+    },
+    {
+      "epoch": 4.976190476190476,
+      "grad_norm": 0.02921333909034729,
+      "learning_rate": 9.240234596801125e-07,
+      "loss": 0.0002,
+      "step": 209
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.006285260431468487,
+      "learning_rate": 8.842519053476476e-07,
+      "loss": 0.0001,
+      "step": 210
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 252,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 42,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.704300308417741e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-210/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcddbcdf99a4c44b8679b782b2f2c338b06bf1242af2189e08ac28c3f5a753ca
+size 7992

checkpoint-210/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/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 . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+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)
+
+        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 files:
+        state_dict = torch.load(f, map_location=device)
+        # 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}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) 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")
+
+
+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 = fp32_flat_groups[0].numel() * 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
+    for name, shape in param_shapes.items():
+
+        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}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        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 get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=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
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    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.
+
+    """
+    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")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, 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_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``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
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+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_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    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_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

checkpoint-252/README.md

@@ -0,0 +1,202 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.14.0

checkpoint-252/adapter_config.json

@@ -0,0 +1,37 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 512,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "v_proj",
+    "q_proj",
+    "up_proj",
+    "down_proj",
+    "k_proj",
+    "o_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-252/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:530d8bfe012a1060fee5f3afb02770e9c499f0365acb31c6150abed5e723293a
+size 1291899552

checkpoint-252/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoint-252/global_step252/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43274db9770b09bdc6b14d6a652b1ed1d1a76ea8c05c5255ca5db85d246e86f8
+size 7751120668

checkpoint-252/global_step252/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b32a55f8b898000c17818b5b14272ad064f25d550ce4a465312986dbdd3e1f21
+size 1396311192

checkpoint-252/latest

@@ -0,0 +1 @@
+global_step252

checkpoint-252/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27958d276becc559b73a720bfcb2c3d742ae17b8c63666d15f0cffada987483e
+size 14244

checkpoint-252/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6e059e53be2326d60a4aba04bdbb91831223cd1740c916267ec664986bf39a0
+size 1064

checkpoint-252/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
+  }
+}

checkpoint-252/tokenizer.json

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

checkpoint-252/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
+}