Upload folder using huggingface_hub

16/README.md

@@ -0,0 +1,202 @@
+---
+base_model: ibm-granite/granite-3.3-8b-base
+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 0.15.2

16/adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 256,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "v_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "o_proj",
+    "k_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

16/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ddfdefc04a4b97efd6e73d7e420ff2991014d3f01374623f4db0d4855058851
+size 791751704

16/latest

@@ -0,0 +1 @@
+global_step834

16/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b35ea62d2fe4062b9b7389e74ceb955a4bd5ab99f0194c75d371f3de08916b0
+size 15920

16/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:284098c6a74458ba63a335c59693f1638518eb8f158d44038b7aefa3482f492e
+size 15920

16/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39dca782a7cc4f3027fc55106bdae2c279cbf161fa82f054f2109f90ef556bc6
+size 15920

16/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4eafc14bf035d708f065e1428b2b9bf015020c016de78203c45b5128ec6c490b
+size 15920

16/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d42f4e8ec154b13139f0ef86f980e54bcbc96d7131ab3928ac04d8a43971a791
+size 15920

16/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c85f3601a96330d70cb4e3990fc22aad91f3cf8a0d3f4c082be0ad6a9739907a
+size 15920

16/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cca38bc7688e02e9e1d945f06ca7d7cb9f13a64c6e4c15c4a27d89e1bb8f2c7
+size 15920

16/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80429aba363bb6a4d1ac7ab8dfa8a02bf9c38f1f9c08e16d640af25b6fe9b498
+size 15920

16/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bc2679d72a43470d06a0c0282d6315539464d9a2d847691a0e39e94a1727f4c
+size 1064

16/special_tokens_map.json

@@ -0,0 +1,51 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

16/tokenizer_config.json

@@ -0,0 +1,188 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<fim_prefix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<fim_middle>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<fim_suffix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<fim_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<filename>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "<gh_stars>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "7": {
+      "content": "<issue_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "8": {
+      "content": "<issue_comment>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "9": {
+      "content": "<issue_closed>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "10": {
+      "content": "<jupyter_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "11": {
+      "content": "<jupyter_text>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "12": {
+      "content": "<jupyter_code>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "13": {
+      "content": "<jupyter_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "14": {
+      "content": "<empty_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "15": {
+      "content": "<commit_before>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "16": {
+      "content": "<commit_msg>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "17": {
+      "content": "<commit_after>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "18": {
+      "content": "<reponame>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "left",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

16/trainer_state.json

@@ -0,0 +1,1196 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 834,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.012004801920768308,
+      "grad_norm": 1.126847743988037,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 1.2938,
+      "step": 5
+    },
+    {
+      "epoch": 0.024009603841536616,
+      "grad_norm": 0.9763688445091248,
+      "learning_rate": 2.5714285714285716e-06,
+      "loss": 1.2541,
+      "step": 10
+    },
+    {
+      "epoch": 0.03601440576230492,
+      "grad_norm": 0.7441720366477966,
+      "learning_rate": 4e-06,
+      "loss": 1.2122,
+      "step": 15
+    },
+    {
+      "epoch": 0.04801920768307323,
+      "grad_norm": 0.6003404259681702,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 1.2006,
+      "step": 20
+    },
+    {
+      "epoch": 0.060024009603841535,
+      "grad_norm": 0.5410875678062439,
+      "learning_rate": 6.857142857142857e-06,
+      "loss": 1.2223,
+      "step": 25
+    },
+    {
+      "epoch": 0.07202881152460984,
+      "grad_norm": 0.5632787346839905,
+      "learning_rate": 8.285714285714287e-06,
+      "loss": 1.155,
+      "step": 30
+    },
+    {
+      "epoch": 0.08403361344537816,
+      "grad_norm": 0.5073860287666321,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 1.1372,
+      "step": 35
+    },
+    {
+      "epoch": 0.09603841536614646,
+      "grad_norm": 0.4458916187286377,
+      "learning_rate": 1.1142857142857143e-05,
+      "loss": 1.1612,
+      "step": 40
+    },
+    {
+      "epoch": 0.10804321728691477,
+      "grad_norm": 0.4466385841369629,
+      "learning_rate": 1.2571428571428572e-05,
+      "loss": 1.15,
+      "step": 45
+    },
+    {
+      "epoch": 0.12004801920768307,
+      "grad_norm": 0.5031300187110901,
+      "learning_rate": 1.4e-05,
+      "loss": 1.1447,
+      "step": 50
+    },
+    {
+      "epoch": 0.13205282112845138,
+      "grad_norm": 0.44504547119140625,
+      "learning_rate": 1.5428571428571428e-05,
+      "loss": 1.1461,
+      "step": 55
+    },
+    {
+      "epoch": 0.14405762304921968,
+      "grad_norm": 0.5289918184280396,
+      "learning_rate": 1.6857142857142858e-05,
+      "loss": 1.1246,
+      "step": 60
+    },
+    {
+      "epoch": 0.15606242496998798,
+      "grad_norm": 0.4286825954914093,
+      "learning_rate": 1.8285714285714288e-05,
+      "loss": 1.08,
+      "step": 65
+    },
+    {
+      "epoch": 0.16806722689075632,
+      "grad_norm": 0.4959764778614044,
+      "learning_rate": 1.9714285714285714e-05,
+      "loss": 1.0991,
+      "step": 70
+    },
+    {
+      "epoch": 0.18007202881152462,
+      "grad_norm": 0.4261907935142517,
+      "learning_rate": 2.1142857142857144e-05,
+      "loss": 1.108,
+      "step": 75
+    },
+    {
+      "epoch": 0.19207683073229292,
+      "grad_norm": 0.39412474632263184,
+      "learning_rate": 2.257142857142857e-05,
+      "loss": 1.0524,
+      "step": 80
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.457573264837265,
+      "learning_rate": 2.4e-05,
+      "loss": 1.09,
+      "step": 85
+    },
+    {
+      "epoch": 0.21608643457382953,
+      "grad_norm": 0.4494602084159851,
+      "learning_rate": 2.5428571428571427e-05,
+      "loss": 1.0803,
+      "step": 90
+    },
+    {
+      "epoch": 0.22809123649459784,
+      "grad_norm": 0.5574950575828552,
+      "learning_rate": 2.685714285714286e-05,
+      "loss": 1.0802,
+      "step": 95
+    },
+    {
+      "epoch": 0.24009603841536614,
+      "grad_norm": 0.581778883934021,
+      "learning_rate": 2.8285714285714287e-05,
+      "loss": 1.0872,
+      "step": 100
+    },
+    {
+      "epoch": 0.25210084033613445,
+      "grad_norm": 0.5143195986747742,
+      "learning_rate": 2.9714285714285717e-05,
+      "loss": 1.0241,
+      "step": 105
+    },
+    {
+      "epoch": 0.26410564225690275,
+      "grad_norm": 0.6538636684417725,
+      "learning_rate": 2.9999697901093597e-05,
+      "loss": 1.0171,
+      "step": 110
+    },
+    {
+      "epoch": 0.27611044417767105,
+      "grad_norm": 0.46839553117752075,
+      "learning_rate": 2.99984706451415e-05,
+      "loss": 1.0474,
+      "step": 115
+    },
+    {
+      "epoch": 0.28811524609843936,
+      "grad_norm": 0.45771801471710205,
+      "learning_rate": 2.999629942814264e-05,
+      "loss": 1.0063,
+      "step": 120
+    },
+    {
+      "epoch": 0.30012004801920766,
+      "grad_norm": 0.5769961476325989,
+      "learning_rate": 2.9993184386747226e-05,
+      "loss": 1.0608,
+      "step": 125
+    },
+    {
+      "epoch": 0.31212484993997597,
+      "grad_norm": 0.5172315835952759,
+      "learning_rate": 2.9989125717007107e-05,
+      "loss": 0.9765,
+      "step": 130
+    },
+    {
+      "epoch": 0.3241296518607443,
+      "grad_norm": 0.6754844784736633,
+      "learning_rate": 2.9984123674363393e-05,
+      "loss": 1.0156,
+      "step": 135
+    },
+    {
+      "epoch": 0.33613445378151263,
+      "grad_norm": 0.8809567093849182,
+      "learning_rate": 2.9978178573630414e-05,
+      "loss": 0.9547,
+      "step": 140
+    },
+    {
+      "epoch": 0.34813925570228094,
+      "grad_norm": 0.614689826965332,
+      "learning_rate": 2.9971290788975884e-05,
+      "loss": 0.9682,
+      "step": 145
+    },
+    {
+      "epoch": 0.36014405762304924,
+      "grad_norm": 0.6198326349258423,
+      "learning_rate": 2.9963460753897364e-05,
+      "loss": 0.9623,
+      "step": 150
+    },
+    {
+      "epoch": 0.37214885954381755,
+      "grad_norm": 0.5520050525665283,
+      "learning_rate": 2.9954688961194974e-05,
+      "loss": 0.9664,
+      "step": 155
+    },
+    {
+      "epoch": 0.38415366146458585,
+      "grad_norm": 0.5535730719566345,
+      "learning_rate": 2.994497596294037e-05,
+      "loss": 0.9632,
+      "step": 160
+    },
+    {
+      "epoch": 0.39615846338535415,
+      "grad_norm": 0.7371352314949036,
+      "learning_rate": 2.9934322370442022e-05,
+      "loss": 0.9166,
+      "step": 165
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.5894187688827515,
+      "learning_rate": 2.9922728854206704e-05,
+      "loss": 0.9436,
+      "step": 170
+    },
+    {
+      "epoch": 0.42016806722689076,
+      "grad_norm": 0.6794357299804688,
+      "learning_rate": 2.9910196143897334e-05,
+      "loss": 0.9549,
+      "step": 175
+    },
+    {
+      "epoch": 0.43217286914765907,
+      "grad_norm": 0.7796855568885803,
+      "learning_rate": 2.9896725028287017e-05,
+      "loss": 0.9387,
+      "step": 180
+    },
+    {
+      "epoch": 0.44417767106842737,
+      "grad_norm": 0.6924332976341248,
+      "learning_rate": 2.9882316355209423e-05,
+      "loss": 0.9127,
+      "step": 185
+    },
+    {
+      "epoch": 0.4561824729891957,
+      "grad_norm": 0.78730708360672,
+      "learning_rate": 2.986697103150542e-05,
+      "loss": 0.9377,
+      "step": 190
+    },
+    {
+      "epoch": 0.468187274909964,
+      "grad_norm": 0.6436963677406311,
+      "learning_rate": 2.9850690022965995e-05,
+      "loss": 0.9075,
+      "step": 195
+    },
+    {
+      "epoch": 0.4801920768307323,
+      "grad_norm": 0.6622775793075562,
+      "learning_rate": 2.9833474354271487e-05,
+      "loss": 0.8795,
+      "step": 200
+    },
+    {
+      "epoch": 0.4921968787515006,
+      "grad_norm": 0.6374281644821167,
+      "learning_rate": 2.981532510892707e-05,
+      "loss": 0.8786,
+      "step": 205
+    },
+    {
+      "epoch": 0.5042016806722689,
+      "grad_norm": 0.6543936133384705,
+      "learning_rate": 2.9796243429194578e-05,
+      "loss": 0.8963,
+      "step": 210
+    },
+    {
+      "epoch": 0.5162064825930373,
+      "grad_norm": 0.8162550330162048,
+      "learning_rate": 2.977623051602062e-05,
+      "loss": 0.8398,
+      "step": 215
+    },
+    {
+      "epoch": 0.5282112845138055,
+      "grad_norm": 0.6448330283164978,
+      "learning_rate": 2.9755287628960982e-05,
+      "loss": 0.8528,
+      "step": 220
+    },
+    {
+      "epoch": 0.5402160864345739,
+      "grad_norm": 4.620177268981934,
+      "learning_rate": 2.9733416086101356e-05,
+      "loss": 0.8638,
+      "step": 225
+    },
+    {
+      "epoch": 0.5522208883553421,
+      "grad_norm": 0.7428641319274902,
+      "learning_rate": 2.9710617263974385e-05,
+      "loss": 0.8236,
+      "step": 230
+    },
+    {
+      "epoch": 0.5642256902761105,
+      "grad_norm": 0.7523469924926758,
+      "learning_rate": 2.968689259747304e-05,
+      "loss": 0.8339,
+      "step": 235
+    },
+    {
+      "epoch": 0.5762304921968787,
+      "grad_norm": 0.7694314122200012,
+      "learning_rate": 2.966224357976029e-05,
+      "loss": 0.8341,
+      "step": 240
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 8.718358993530273,
+      "learning_rate": 2.9636671762175145e-05,
+      "loss": 0.8247,
+      "step": 245
+    },
+    {
+      "epoch": 0.6002400960384153,
+      "grad_norm": 0.8020578026771545,
+      "learning_rate": 2.9610178754135005e-05,
+      "loss": 0.7834,
+      "step": 250
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 0.7112622857093811,
+      "learning_rate": 2.958276622303438e-05,
+      "loss": 0.8256,
+      "step": 255
+    },
+    {
+      "epoch": 0.6242496998799519,
+      "grad_norm": 0.7433765530586243,
+      "learning_rate": 2.9554435894139945e-05,
+      "loss": 0.7668,
+      "step": 260
+    },
+    {
+      "epoch": 0.6362545018007203,
+      "grad_norm": 0.7709916234016418,
+      "learning_rate": 2.9525189550481955e-05,
+      "loss": 0.8346,
+      "step": 265
+    },
+    {
+      "epoch": 0.6482593037214885,
+      "grad_norm": 1.9189757108688354,
+      "learning_rate": 2.9495029032742025e-05,
+      "loss": 0.7604,
+      "step": 270
+    },
+    {
+      "epoch": 0.6602641056422569,
+      "grad_norm": 0.8425916433334351,
+      "learning_rate": 2.9463956239137287e-05,
+      "loss": 0.8273,
+      "step": 275
+    },
+    {
+      "epoch": 0.6722689075630253,
+      "grad_norm": 0.8174909353256226,
+      "learning_rate": 2.943197312530091e-05,
+      "loss": 0.7908,
+      "step": 280
+    },
+    {
+      "epoch": 0.6842737094837935,
+      "grad_norm": 0.6657884120941162,
+      "learning_rate": 2.9399081704159034e-05,
+      "loss": 0.7703,
+      "step": 285
+    },
+    {
+      "epoch": 0.6962785114045619,
+      "grad_norm": 0.8427141904830933,
+      "learning_rate": 2.936528404580408e-05,
+      "loss": 0.8011,
+      "step": 290
+    },
+    {
+      "epoch": 0.7082833133253301,
+      "grad_norm": 0.8150597214698792,
+      "learning_rate": 2.9330582277364453e-05,
+      "loss": 0.7654,
+      "step": 295
+    },
+    {
+      "epoch": 0.7202881152460985,
+      "grad_norm": 0.7505553364753723,
+      "learning_rate": 2.929497858287067e-05,
+      "loss": 0.7376,
+      "step": 300
+    },
+    {
+      "epoch": 0.7322929171668667,
+      "grad_norm": 0.7946950793266296,
+      "learning_rate": 2.925847520311791e-05,
+      "loss": 0.7239,
+      "step": 305
+    },
+    {
+      "epoch": 0.7442977190876351,
+      "grad_norm": 0.8390210866928101,
+      "learning_rate": 2.9221074435524995e-05,
+      "loss": 0.7742,
+      "step": 310
+    },
+    {
+      "epoch": 0.7563025210084033,
+      "grad_norm": 0.8262509107589722,
+      "learning_rate": 2.9182778633989756e-05,
+      "loss": 0.759,
+      "step": 315
+    },
+    {
+      "epoch": 0.7683073229291717,
+      "grad_norm": 0.9196308255195618,
+      "learning_rate": 2.9143590208740925e-05,
+      "loss": 0.7897,
+      "step": 320
+    },
+    {
+      "epoch": 0.78031212484994,
+      "grad_norm": 0.8700764775276184,
+      "learning_rate": 2.9103511626186444e-05,
+      "loss": 0.7395,
+      "step": 325
+    },
+    {
+      "epoch": 0.7923169267707083,
+      "grad_norm": 0.8262081742286682,
+      "learning_rate": 2.9062545408758193e-05,
+      "loss": 0.7038,
+      "step": 330
+    },
+    {
+      "epoch": 0.8043217286914766,
+      "grad_norm": 0.8056402206420898,
+      "learning_rate": 2.902069413475329e-05,
+      "loss": 0.7026,
+      "step": 335
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.8297358751296997,
+      "learning_rate": 2.8977960438171787e-05,
+      "loss": 0.7308,
+      "step": 340
+    },
+    {
+      "epoch": 0.8283313325330132,
+      "grad_norm": 0.8549435138702393,
+      "learning_rate": 2.8934347008550886e-05,
+      "loss": 0.6703,
+      "step": 345
+    },
+    {
+      "epoch": 0.8403361344537815,
+      "grad_norm": 0.9239634871482849,
+      "learning_rate": 2.8889856590795705e-05,
+      "loss": 0.7002,
+      "step": 350
+    },
+    {
+      "epoch": 0.8523409363745498,
+      "grad_norm": 0.8247022032737732,
+      "learning_rate": 2.8844491985006486e-05,
+      "loss": 0.7076,
+      "step": 355
+    },
+    {
+      "epoch": 0.8643457382953181,
+      "grad_norm": 0.9243834614753723,
+      "learning_rate": 2.8798256046302376e-05,
+      "loss": 0.6776,
+      "step": 360
+    },
+    {
+      "epoch": 0.8763505402160864,
+      "grad_norm": 0.9479813575744629,
+      "learning_rate": 2.875115168464174e-05,
+      "loss": 0.6979,
+      "step": 365
+    },
+    {
+      "epoch": 0.8883553421368547,
+      "grad_norm": 0.988172173500061,
+      "learning_rate": 2.8703181864639013e-05,
+      "loss": 0.7221,
+      "step": 370
+    },
+    {
+      "epoch": 0.9003601440576231,
+      "grad_norm": 1.0673149824142456,
+      "learning_rate": 2.865434960537811e-05,
+      "loss": 0.6326,
+      "step": 375
+    },
+    {
+      "epoch": 0.9123649459783914,
+      "grad_norm": 0.9364232420921326,
+      "learning_rate": 2.860465798022242e-05,
+      "loss": 0.6799,
+      "step": 380
+    },
+    {
+      "epoch": 0.9243697478991597,
+      "grad_norm": 0.9759359955787659,
+      "learning_rate": 2.8554110116621376e-05,
+      "loss": 0.6705,
+      "step": 385
+    },
+    {
+      "epoch": 0.936374549819928,
+      "grad_norm": 0.8587977290153503,
+      "learning_rate": 2.8502709195913617e-05,
+      "loss": 0.6405,
+      "step": 390
+    },
+    {
+      "epoch": 0.9483793517406963,
+      "grad_norm": 0.8970200419425964,
+      "learning_rate": 2.8450458453126773e-05,
+      "loss": 0.6404,
+      "step": 395
+    },
+    {
+      "epoch": 0.9603841536614646,
+      "grad_norm": 0.8413552045822144,
+      "learning_rate": 2.839736117677386e-05,
+      "loss": 0.6313,
+      "step": 400
+    },
+    {
+      "epoch": 0.9723889555822329,
+      "grad_norm": 0.9825181365013123,
+      "learning_rate": 2.8343420708646292e-05,
+      "loss": 0.5839,
+      "step": 405
+    },
+    {
+      "epoch": 0.9843937575030012,
+      "grad_norm": 0.8249124884605408,
+      "learning_rate": 2.8288640443603587e-05,
+      "loss": 0.6214,
+      "step": 410
+    },
+    {
+      "epoch": 0.9963985594237695,
+      "grad_norm": 1.0160948038101196,
+      "learning_rate": 2.823302382935968e-05,
+      "loss": 0.6417,
+      "step": 415
+    },
+    {
+      "epoch": 1.007202881152461,
+      "grad_norm": 0.9784197211265564,
+      "learning_rate": 2.8176574366265963e-05,
+      "loss": 0.5966,
+      "step": 420
+    },
+    {
+      "epoch": 1.0192076830732293,
+      "grad_norm": 0.9859529137611389,
+      "learning_rate": 2.8119295607090936e-05,
+      "loss": 0.5378,
+      "step": 425
+    },
+    {
+      "epoch": 1.0312124849939976,
+      "grad_norm": 1.0028948783874512,
+      "learning_rate": 2.8061191156796658e-05,
+      "loss": 0.5948,
+      "step": 430
+    },
+    {
+      "epoch": 1.043217286914766,
+      "grad_norm": 0.9926419854164124,
+      "learning_rate": 2.8002264672311822e-05,
+      "loss": 0.5835,
+      "step": 435
+    },
+    {
+      "epoch": 1.0552220888355341,
+      "grad_norm": 1.0303875207901,
+      "learning_rate": 2.7942519862301618e-05,
+      "loss": 0.4992,
+      "step": 440
+    },
+    {
+      "epoch": 1.0672268907563025,
+      "grad_norm": 0.8923658132553101,
+      "learning_rate": 2.7881960486934314e-05,
+      "loss": 0.5432,
+      "step": 445
+    },
+    {
+      "epoch": 1.0792316926770709,
+      "grad_norm": 0.8442468643188477,
+      "learning_rate": 2.7820590357644604e-05,
+      "loss": 0.5246,
+      "step": 450
+    },
+    {
+      "epoch": 1.0912364945978392,
+      "grad_norm": 0.9535374641418457,
+      "learning_rate": 2.7758413336893725e-05,
+      "loss": 0.5832,
+      "step": 455
+    },
+    {
+      "epoch": 1.1032412965186074,
+      "grad_norm": 1.013627290725708,
+      "learning_rate": 2.7695433337926362e-05,
+      "loss": 0.5381,
+      "step": 460
+    },
+    {
+      "epoch": 1.1152460984393757,
+      "grad_norm": 0.9319940805435181,
+      "learning_rate": 2.763165432452438e-05,
+      "loss": 0.5282,
+      "step": 465
+    },
+    {
+      "epoch": 1.127250900360144,
+      "grad_norm": 0.8916425704956055,
+      "learning_rate": 2.756708031075731e-05,
+      "loss": 0.5252,
+      "step": 470
+    },
+    {
+      "epoch": 1.1392557022809124,
+      "grad_norm": 0.9970868825912476,
+      "learning_rate": 2.7501715360729766e-05,
+      "loss": 0.5196,
+      "step": 475
+    },
+    {
+      "epoch": 1.1512605042016806,
+      "grad_norm": 0.9539552330970764,
+      "learning_rate": 2.7435563588325627e-05,
+      "loss": 0.4874,
+      "step": 480
+    },
+    {
+      "epoch": 1.163265306122449,
+      "grad_norm": 1.0665792226791382,
+      "learning_rate": 2.7368629156949137e-05,
+      "loss": 0.5093,
+      "step": 485
+    },
+    {
+      "epoch": 1.1752701080432173,
+      "grad_norm": 1.048183798789978,
+      "learning_rate": 2.7300916279262868e-05,
+      "loss": 0.522,
+      "step": 490
+    },
+    {
+      "epoch": 1.1872749099639857,
+      "grad_norm": 1.1342055797576904,
+      "learning_rate": 2.7232429216922583e-05,
+      "loss": 0.5293,
+      "step": 495
+    },
+    {
+      "epoch": 1.199279711884754,
+      "grad_norm": 0.9771571159362793,
+      "learning_rate": 2.7163172280309028e-05,
+      "loss": 0.5041,
+      "step": 500
+    },
+    {
+      "epoch": 1.2112845138055222,
+      "grad_norm": 1.2090368270874023,
+      "learning_rate": 2.7093149828256643e-05,
+      "loss": 0.5112,
+      "step": 505
+    },
+    {
+      "epoch": 1.2232893157262905,
+      "grad_norm": 1.1582231521606445,
+      "learning_rate": 2.702236626777923e-05,
+      "loss": 0.4962,
+      "step": 510
+    },
+    {
+      "epoch": 1.2352941176470589,
+      "grad_norm": 1.037467360496521,
+      "learning_rate": 2.6950826053792586e-05,
+      "loss": 0.4958,
+      "step": 515
+    },
+    {
+      "epoch": 1.247298919567827,
+      "grad_norm": 0.9874077439308167,
+      "learning_rate": 2.6878533688834125e-05,
+      "loss": 0.4696,
+      "step": 520
+    },
+    {
+      "epoch": 1.2593037214885954,
+      "grad_norm": 1.03226900100708,
+      "learning_rate": 2.6805493722779513e-05,
+      "loss": 0.4994,
+      "step": 525
+    },
+    {
+      "epoch": 1.2713085234093637,
+      "grad_norm": 0.986599326133728,
+      "learning_rate": 2.6731710752556293e-05,
+      "loss": 0.4679,
+      "step": 530
+    },
+    {
+      "epoch": 1.283313325330132,
+      "grad_norm": 0.8920894861221313,
+      "learning_rate": 2.6657189421854564e-05,
+      "loss": 0.5125,
+      "step": 535
+    },
+    {
+      "epoch": 1.2953181272509005,
+      "grad_norm": 1.0151013135910034,
+      "learning_rate": 2.6581934420834754e-05,
+      "loss": 0.508,
+      "step": 540
+    },
+    {
+      "epoch": 1.3073229291716686,
+      "grad_norm": 1.0022249221801758,
+      "learning_rate": 2.65059504858324e-05,
+      "loss": 0.4258,
+      "step": 545
+    },
+    {
+      "epoch": 1.319327731092437,
+      "grad_norm": 0.912742018699646,
+      "learning_rate": 2.6429242399060063e-05,
+      "loss": 0.5101,
+      "step": 550
+    },
+    {
+      "epoch": 1.3313325330132053,
+      "grad_norm": 1.016350507736206,
+      "learning_rate": 2.635181498830637e-05,
+      "loss": 0.4921,
+      "step": 555
+    },
+    {
+      "epoch": 1.3433373349339737,
+      "grad_norm": 0.8971051573753357,
+      "learning_rate": 2.627367312663214e-05,
+      "loss": 0.4477,
+      "step": 560
+    },
+    {
+      "epoch": 1.355342136854742,
+      "grad_norm": 0.9581533074378967,
+      "learning_rate": 2.6194821732063708e-05,
+      "loss": 0.4356,
+      "step": 565
+    },
+    {
+      "epoch": 1.3673469387755102,
+      "grad_norm": 1.00229012966156,
+      "learning_rate": 2.6115265767283377e-05,
+      "loss": 0.4632,
+      "step": 570
+    },
+    {
+      "epoch": 1.3793517406962785,
+      "grad_norm": 0.9300605654716492,
+      "learning_rate": 2.6035010239317106e-05,
+      "loss": 0.5101,
+      "step": 575
+    },
+    {
+      "epoch": 1.3913565426170469,
+      "grad_norm": 0.9333869814872742,
+      "learning_rate": 2.5954060199219364e-05,
+      "loss": 0.4389,
+      "step": 580
+    },
+    {
+      "epoch": 1.403361344537815,
+      "grad_norm": 1.121485710144043,
+      "learning_rate": 2.587242074175523e-05,
+      "loss": 0.4602,
+      "step": 585
+    },
+    {
+      "epoch": 1.4153661464585834,
+      "grad_norm": 0.9909285306930542,
+      "learning_rate": 2.5790097005079766e-05,
+      "loss": 0.4386,
+      "step": 590
+    },
+    {
+      "epoch": 1.4273709483793517,
+      "grad_norm": 1.1501377820968628,
+      "learning_rate": 2.5707094170414606e-05,
+      "loss": 0.4577,
+      "step": 595
+    },
+    {
+      "epoch": 1.43937575030012,
+      "grad_norm": 1.0231988430023193,
+      "learning_rate": 2.5623417461721887e-05,
+      "loss": 0.4762,
+      "step": 600
+    },
+    {
+      "epoch": 1.4513805522208885,
+      "grad_norm": 0.9353159070014954,
+      "learning_rate": 2.5539072145375452e-05,
+      "loss": 0.4456,
+      "step": 605
+    },
+    {
+      "epoch": 1.4633853541416566,
+      "grad_norm": 1.0465285778045654,
+      "learning_rate": 2.5454063529829405e-05,
+      "loss": 0.4379,
+      "step": 610
+    },
+    {
+      "epoch": 1.475390156062425,
+      "grad_norm": 1.1565396785736084,
+      "learning_rate": 2.5368396965284017e-05,
+      "loss": 0.4122,
+      "step": 615
+    },
+    {
+      "epoch": 1.4873949579831933,
+      "grad_norm": 1.128082513809204,
+      "learning_rate": 2.5282077843349e-05,
+      "loss": 0.4581,
+      "step": 620
+    },
+    {
+      "epoch": 1.4993997599039615,
+      "grad_norm": 0.9868873357772827,
+      "learning_rate": 2.519511159670417e-05,
+      "loss": 0.4402,
+      "step": 625
+    },
+    {
+      "epoch": 1.51140456182473,
+      "grad_norm": 1.0035743713378906,
+      "learning_rate": 2.510750369875752e-05,
+      "loss": 0.4489,
+      "step": 630
+    },
+    {
+      "epoch": 1.5234093637454982,
+      "grad_norm": 1.1779558658599854,
+      "learning_rate": 2.5019259663300758e-05,
+      "loss": 0.4474,
+      "step": 635
+    },
+    {
+      "epoch": 1.5354141656662665,
+      "grad_norm": 1.0303417444229126,
+      "learning_rate": 2.4930385044162282e-05,
+      "loss": 0.4338,
+      "step": 640
+    },
+    {
+      "epoch": 1.547418967587035,
+      "grad_norm": 1.0198581218719482,
+      "learning_rate": 2.4840885434857608e-05,
+      "loss": 0.399,
+      "step": 645
+    },
+    {
+      "epoch": 1.559423769507803,
+      "grad_norm": 0.9304978847503662,
+      "learning_rate": 2.4750766468237388e-05,
+      "loss": 0.4285,
+      "step": 650
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.9525741338729858,
+      "learning_rate": 2.466003381613283e-05,
+      "loss": 0.4253,
+      "step": 655
+    },
+    {
+      "epoch": 1.5834333733493398,
+      "grad_norm": 0.9774038791656494,
+      "learning_rate": 2.456869318899878e-05,
+      "loss": 0.4608,
+      "step": 660
+    },
+    {
+      "epoch": 1.595438175270108,
+      "grad_norm": 0.9870194792747498,
+      "learning_rate": 2.447675033555429e-05,
+      "loss": 0.3993,
+      "step": 665
+    },
+    {
+      "epoch": 1.6074429771908765,
+      "grad_norm": 0.9398269653320312,
+      "learning_rate": 2.4384211042420826e-05,
+      "loss": 0.4121,
+      "step": 670
+    },
+    {
+      "epoch": 1.6194477791116446,
+      "grad_norm": 0.9968813061714172,
+      "learning_rate": 2.4291081133758063e-05,
+      "loss": 0.4298,
+      "step": 675
+    },
+    {
+      "epoch": 1.631452581032413,
+      "grad_norm": 1.1225817203521729,
+      "learning_rate": 2.4197366470897352e-05,
+      "loss": 0.3736,
+      "step": 680
+    },
+    {
+      "epoch": 1.6434573829531813,
+      "grad_norm": 1.1227293014526367,
+      "learning_rate": 2.4103072951972794e-05,
+      "loss": 0.4418,
+      "step": 685
+    },
+    {
+      "epoch": 1.6554621848739495,
+      "grad_norm": 1.0404561758041382,
+      "learning_rate": 2.400820651155005e-05,
+      "loss": 0.3952,
+      "step": 690
+    },
+    {
+      "epoch": 1.667466986794718,
+      "grad_norm": 0.9955992102622986,
+      "learning_rate": 2.3912773120252822e-05,
+      "loss": 0.387,
+      "step": 695
+    },
+    {
+      "epoch": 1.6794717887154862,
+      "grad_norm": 1.0180872678756714,
+      "learning_rate": 2.3816778784387097e-05,
+      "loss": 0.3817,
+      "step": 700
+    },
+    {
+      "epoch": 1.6914765906362546,
+      "grad_norm": 0.9844180345535278,
+      "learning_rate": 2.372022954556311e-05,
+      "loss": 0.4103,
+      "step": 705
+    },
+    {
+      "epoch": 1.703481392557023,
+      "grad_norm": 1.05225670337677,
+      "learning_rate": 2.3623131480315107e-05,
+      "loss": 0.367,
+      "step": 710
+    },
+    {
+      "epoch": 1.715486194477791,
+      "grad_norm": 1.0176974534988403,
+      "learning_rate": 2.352549069971891e-05,
+      "loss": 0.3605,
+      "step": 715
+    },
+    {
+      "epoch": 1.7274909963985594,
+      "grad_norm": 1.0177741050720215,
+      "learning_rate": 2.3427313349007302e-05,
+      "loss": 0.3782,
+      "step": 720
+    },
+    {
+      "epoch": 1.7394957983193278,
+      "grad_norm": 0.9692652821540833,
+      "learning_rate": 2.3328605607183255e-05,
+      "loss": 0.3554,
+      "step": 725
+    },
+    {
+      "epoch": 1.751500600240096,
+      "grad_norm": 0.892565131187439,
+      "learning_rate": 2.322937368663105e-05,
+      "loss": 0.3714,
+      "step": 730
+    },
+    {
+      "epoch": 1.7635054021608645,
+      "grad_norm": 0.9849589467048645,
+      "learning_rate": 2.312962383272529e-05,
+      "loss": 0.3648,
+      "step": 735
+    },
+    {
+      "epoch": 1.7755102040816326,
+      "grad_norm": 1.1365302801132202,
+      "learning_rate": 2.3029362323437823e-05,
+      "loss": 0.3451,
+      "step": 740
+    },
+    {
+      "epoch": 1.787515006002401,
+      "grad_norm": 0.9451067447662354,
+      "learning_rate": 2.292859546894263e-05,
+      "loss": 0.3465,
+      "step": 745
+    },
+    {
+      "epoch": 1.7995198079231693,
+      "grad_norm": 1.0000107288360596,
+      "learning_rate": 2.282732961121869e-05,
+      "loss": 0.3297,
+      "step": 750
+    },
+    {
+      "epoch": 1.8115246098439375,
+      "grad_norm": 0.9930965900421143,
+      "learning_rate": 2.2725571123650814e-05,
+      "loss": 0.3413,
+      "step": 755
+    },
+    {
+      "epoch": 1.8235294117647058,
+      "grad_norm": 1.1068987846374512,
+      "learning_rate": 2.2623326410628537e-05,
+      "loss": 0.3501,
+      "step": 760
+    },
+    {
+      "epoch": 1.8355342136854742,
+      "grad_norm": 1.0239458084106445,
+      "learning_rate": 2.2520601907143045e-05,
+      "loss": 0.3738,
+      "step": 765
+    },
+    {
+      "epoch": 1.8475390156062423,
+      "grad_norm": 1.0029224157333374,
+      "learning_rate": 2.2417404078382153e-05,
+      "loss": 0.3683,
+      "step": 770
+    },
+    {
+      "epoch": 1.859543817527011,
+      "grad_norm": 1.0009812116622925,
+      "learning_rate": 2.2313739419323443e-05,
+      "loss": 0.3782,
+      "step": 775
+    },
+    {
+      "epoch": 1.871548619447779,
+      "grad_norm": 1.0867785215377808,
+      "learning_rate": 2.220961445432546e-05,
+      "loss": 0.3504,
+      "step": 780
+    },
+    {
+      "epoch": 1.8835534213685474,
+      "grad_norm": 1.0827844142913818,
+      "learning_rate": 2.2105035736717086e-05,
+      "loss": 0.3416,
+      "step": 785
+    },
+    {
+      "epoch": 1.8955582232893158,
+      "grad_norm": 0.9669961929321289,
+      "learning_rate": 2.2000009848385107e-05,
+      "loss": 0.3535,
+      "step": 790
+    },
+    {
+      "epoch": 1.907563025210084,
+      "grad_norm": 1.0839868783950806,
+      "learning_rate": 2.189454339935996e-05,
+      "loss": 0.3245,
+      "step": 795
+    },
+    {
+      "epoch": 1.9195678271308525,
+      "grad_norm": 1.1661854982376099,
+      "learning_rate": 2.1788643027399725e-05,
+      "loss": 0.3481,
+      "step": 800
+    },
+    {
+      "epoch": 1.9315726290516206,
+      "grad_norm": 1.0360976457595825,
+      "learning_rate": 2.1682315397572344e-05,
+      "loss": 0.3725,
+      "step": 805
+    },
+    {
+      "epoch": 1.943577430972389,
+      "grad_norm": 0.9784966707229614,
+      "learning_rate": 2.157556720183616e-05,
+      "loss": 0.318,
+      "step": 810
+    },
+    {
+      "epoch": 1.9555822328931574,
+      "grad_norm": 1.1009594202041626,
+      "learning_rate": 2.1468405158618744e-05,
+      "loss": 0.3223,
+      "step": 815
+    },
+    {
+      "epoch": 1.9675870348139255,
+      "grad_norm": 0.9810320734977722,
+      "learning_rate": 2.136083601239403e-05,
+      "loss": 0.3622,
+      "step": 820
+    },
+    {
+      "epoch": 1.9795918367346939,
+      "grad_norm": 1.039110779762268,
+      "learning_rate": 2.125286653325787e-05,
+      "loss": 0.3408,
+      "step": 825
+    },
+    {
+      "epoch": 1.9915966386554622,
+      "grad_norm": 1.1904340982437134,
+      "learning_rate": 2.114450351650193e-05,
+      "loss": 0.3264,
+      "step": 830
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 2085,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3201796128560906e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

16/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:285c11f66de8dfb56501536af272fbbd494c94d206e611f25981c1b5403c74e4
+size 7736

16/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)