Upload folder using huggingface_hub

9/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

9/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": [
+    "v_proj",
+    "gate_proj",
+    "down_proj",
+    "o_proj",
+    "k_proj",
+    "up_proj",
+    "q_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

9/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96fed5fd5f0d5e028956b99fa162582f9d18ec5fb15efc09ac86d05c1b7a9143
+size 791751704

9/latest

@@ -0,0 +1 @@
+global_step912

9/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e31e8412ec8ae597420e2807b4e63743460f599ff8867a0f93354d241031b9dd
+size 14960

9/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7539949654454dba83587f03b01e2bf2e56e6bd18153ca635b2c1e0ce06fbcea
+size 14960

9/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b690c98aa5a8e8ef673235c0e36671d17f34eebaea788f66bc9473d447c0bd4
+size 14960

9/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acb01bc992c5367ca479c8b0339fedbfc00af93067b999bc5278b9df78a46bba
+size 14960

9/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16d9c75ca2fe7f938ea560035db6cf26e7a29373280cbdc3b053557e8e0f975d
+size 1064

9/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
+  }
+}

9/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
+}

9/trainer_state.json

@@ -0,0 +1,1308 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 912,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.010964912280701754,
+      "grad_norm": 1.1064090728759766,
+      "learning_rate": 1.0526315789473683e-06,
+      "loss": 1.2509,
+      "step": 5
+    },
+    {
+      "epoch": 0.021929824561403508,
+      "grad_norm": 0.8702306747436523,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 1.2373,
+      "step": 10
+    },
+    {
+      "epoch": 0.03289473684210526,
+      "grad_norm": 0.7855818271636963,
+      "learning_rate": 3.6842105263157892e-06,
+      "loss": 1.2409,
+      "step": 15
+    },
+    {
+      "epoch": 0.043859649122807015,
+      "grad_norm": 0.6700618267059326,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 1.2426,
+      "step": 20
+    },
+    {
+      "epoch": 0.05482456140350877,
+      "grad_norm": 0.6601688861846924,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 1.2052,
+      "step": 25
+    },
+    {
+      "epoch": 0.06578947368421052,
+      "grad_norm": 0.5013778805732727,
+      "learning_rate": 7.631578947368421e-06,
+      "loss": 1.1919,
+      "step": 30
+    },
+    {
+      "epoch": 0.07675438596491228,
+      "grad_norm": 0.547326385974884,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 1.1842,
+      "step": 35
+    },
+    {
+      "epoch": 0.08771929824561403,
+      "grad_norm": 0.6308258175849915,
+      "learning_rate": 1.0263157894736843e-05,
+      "loss": 1.2126,
+      "step": 40
+    },
+    {
+      "epoch": 0.09868421052631579,
+      "grad_norm": 0.5670621395111084,
+      "learning_rate": 1.1578947368421052e-05,
+      "loss": 1.1301,
+      "step": 45
+    },
+    {
+      "epoch": 0.10964912280701754,
+      "grad_norm": 0.6730433702468872,
+      "learning_rate": 1.2894736842105264e-05,
+      "loss": 1.1245,
+      "step": 50
+    },
+    {
+      "epoch": 0.1206140350877193,
+      "grad_norm": 0.45176073908805847,
+      "learning_rate": 1.4210526315789473e-05,
+      "loss": 1.1493,
+      "step": 55
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 0.43592914938926697,
+      "learning_rate": 1.5526315789473686e-05,
+      "loss": 1.1641,
+      "step": 60
+    },
+    {
+      "epoch": 0.1425438596491228,
+      "grad_norm": 0.594582200050354,
+      "learning_rate": 1.6842105263157893e-05,
+      "loss": 1.1486,
+      "step": 65
+    },
+    {
+      "epoch": 0.15350877192982457,
+      "grad_norm": 0.46377938985824585,
+      "learning_rate": 1.8157894736842107e-05,
+      "loss": 1.0793,
+      "step": 70
+    },
+    {
+      "epoch": 0.16447368421052633,
+      "grad_norm": 0.6227210164070129,
+      "learning_rate": 1.9473684210526318e-05,
+      "loss": 1.1772,
+      "step": 75
+    },
+    {
+      "epoch": 0.17543859649122806,
+      "grad_norm": 0.5389492511749268,
+      "learning_rate": 2.078947368421053e-05,
+      "loss": 1.1091,
+      "step": 80
+    },
+    {
+      "epoch": 0.18640350877192982,
+      "grad_norm": 0.5326628684997559,
+      "learning_rate": 2.2105263157894736e-05,
+      "loss": 1.0536,
+      "step": 85
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "grad_norm": 0.6689491868019104,
+      "learning_rate": 2.3421052631578947e-05,
+      "loss": 1.1105,
+      "step": 90
+    },
+    {
+      "epoch": 0.20833333333333334,
+      "grad_norm": 0.5751927495002747,
+      "learning_rate": 2.4736842105263158e-05,
+      "loss": 1.0089,
+      "step": 95
+    },
+    {
+      "epoch": 0.21929824561403508,
+      "grad_norm": 0.5814879536628723,
+      "learning_rate": 2.605263157894737e-05,
+      "loss": 1.0118,
+      "step": 100
+    },
+    {
+      "epoch": 0.23026315789473684,
+      "grad_norm": 0.5261718034744263,
+      "learning_rate": 2.736842105263158e-05,
+      "loss": 1.0419,
+      "step": 105
+    },
+    {
+      "epoch": 0.2412280701754386,
+      "grad_norm": 0.5454193949699402,
+      "learning_rate": 2.868421052631579e-05,
+      "loss": 1.052,
+      "step": 110
+    },
+    {
+      "epoch": 0.25219298245614036,
+      "grad_norm": 0.5560798645019531,
+      "learning_rate": 3e-05,
+      "loss": 1.0284,
+      "step": 115
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.5735645294189453,
+      "learning_rate": 2.9999605558856208e-05,
+      "loss": 0.9681,
+      "step": 120
+    },
+    {
+      "epoch": 0.2741228070175439,
+      "grad_norm": 0.6058073043823242,
+      "learning_rate": 2.9998422256169334e-05,
+      "loss": 0.999,
+      "step": 125
+    },
+    {
+      "epoch": 0.2850877192982456,
+      "grad_norm": 0.6186266541481018,
+      "learning_rate": 2.999645015417182e-05,
+      "loss": 1.0081,
+      "step": 130
+    },
+    {
+      "epoch": 0.29605263157894735,
+      "grad_norm": 0.658362627029419,
+      "learning_rate": 2.999368935658075e-05,
+      "loss": 1.0287,
+      "step": 135
+    },
+    {
+      "epoch": 0.30701754385964913,
+      "grad_norm": 0.6877792477607727,
+      "learning_rate": 2.9990140008592416e-05,
+      "loss": 0.9967,
+      "step": 140
+    },
+    {
+      "epoch": 0.31798245614035087,
+      "grad_norm": 0.6147474050521851,
+      "learning_rate": 2.9985802296874666e-05,
+      "loss": 0.9413,
+      "step": 145
+    },
+    {
+      "epoch": 0.32894736842105265,
+      "grad_norm": 0.5925890803337097,
+      "learning_rate": 2.9980676449557092e-05,
+      "loss": 0.9951,
+      "step": 150
+    },
+    {
+      "epoch": 0.3399122807017544,
+      "grad_norm": 0.6306092143058777,
+      "learning_rate": 2.9974762736219048e-05,
+      "loss": 0.9636,
+      "step": 155
+    },
+    {
+      "epoch": 0.3508771929824561,
+      "grad_norm": 0.6569765210151672,
+      "learning_rate": 2.9968061467875442e-05,
+      "loss": 0.9573,
+      "step": 160
+    },
+    {
+      "epoch": 0.3618421052631579,
+      "grad_norm": 0.7791516780853271,
+      "learning_rate": 2.9960572996960397e-05,
+      "loss": 0.948,
+      "step": 165
+    },
+    {
+      "epoch": 0.37280701754385964,
+      "grad_norm": 0.7109338045120239,
+      "learning_rate": 2.9952297717308726e-05,
+      "loss": 0.9343,
+      "step": 170
+    },
+    {
+      "epoch": 0.38377192982456143,
+      "grad_norm": 0.7839727401733398,
+      "learning_rate": 2.9943236064135187e-05,
+      "loss": 1.0018,
+      "step": 175
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.6728869080543518,
+      "learning_rate": 2.9933388514011638e-05,
+      "loss": 0.9086,
+      "step": 180
+    },
+    {
+      "epoch": 0.4057017543859649,
+      "grad_norm": 0.6420652270317078,
+      "learning_rate": 2.992275558484193e-05,
+      "loss": 0.9154,
+      "step": 185
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.84992915391922,
+      "learning_rate": 2.9911337835834698e-05,
+      "loss": 0.9551,
+      "step": 190
+    },
+    {
+      "epoch": 0.4276315789473684,
+      "grad_norm": 0.7592620253562927,
+      "learning_rate": 2.9899135867473938e-05,
+      "loss": 0.8403,
+      "step": 195
+    },
+    {
+      "epoch": 0.43859649122807015,
+      "grad_norm": 0.8052639961242676,
+      "learning_rate": 2.988615032148743e-05,
+      "loss": 0.8782,
+      "step": 200
+    },
+    {
+      "epoch": 0.44956140350877194,
+      "grad_norm": 0.6697429418563843,
+      "learning_rate": 2.987238188081299e-05,
+      "loss": 0.8509,
+      "step": 205
+    },
+    {
+      "epoch": 0.4605263157894737,
+      "grad_norm": 0.6837584376335144,
+      "learning_rate": 2.985783126956255e-05,
+      "loss": 0.8474,
+      "step": 210
+    },
+    {
+      "epoch": 0.47149122807017546,
+      "grad_norm": 0.6947823762893677,
+      "learning_rate": 2.9842499252984074e-05,
+      "loss": 0.8486,
+      "step": 215
+    },
+    {
+      "epoch": 0.4824561403508772,
+      "grad_norm": 0.7756646275520325,
+      "learning_rate": 2.982638663742132e-05,
+      "loss": 0.8482,
+      "step": 220
+    },
+    {
+      "epoch": 0.4934210526315789,
+      "grad_norm": 0.772576093673706,
+      "learning_rate": 2.980949427027142e-05,
+      "loss": 0.8263,
+      "step": 225
+    },
+    {
+      "epoch": 0.5043859649122807,
+      "grad_norm": 0.7955979704856873,
+      "learning_rate": 2.9791823039940322e-05,
+      "loss": 0.8094,
+      "step": 230
+    },
+    {
+      "epoch": 0.5153508771929824,
+      "grad_norm": 0.8120087385177612,
+      "learning_rate": 2.977337387579607e-05,
+      "loss": 0.8461,
+      "step": 235
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.8410856127738953,
+      "learning_rate": 2.9754147748119915e-05,
+      "loss": 0.8086,
+      "step": 240
+    },
+    {
+      "epoch": 0.5372807017543859,
+      "grad_norm": 0.7663746476173401,
+      "learning_rate": 2.9734145668055296e-05,
+      "loss": 0.8518,
+      "step": 245
+    },
+    {
+      "epoch": 0.5482456140350878,
+      "grad_norm": 0.7535541653633118,
+      "learning_rate": 2.9713368687554663e-05,
+      "loss": 0.8012,
+      "step": 250
+    },
+    {
+      "epoch": 0.5592105263157895,
+      "grad_norm": 0.7745866179466248,
+      "learning_rate": 2.9691817899324136e-05,
+      "loss": 0.8491,
+      "step": 255
+    },
+    {
+      "epoch": 0.5701754385964912,
+      "grad_norm": 0.8415214419364929,
+      "learning_rate": 2.9669494436766066e-05,
+      "loss": 0.7938,
+      "step": 260
+    },
+    {
+      "epoch": 0.581140350877193,
+      "grad_norm": 0.7968622446060181,
+      "learning_rate": 2.964639947391939e-05,
+      "loss": 0.8084,
+      "step": 265
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "grad_norm": 0.7041695713996887,
+      "learning_rate": 2.9622534225397923e-05,
+      "loss": 0.7618,
+      "step": 270
+    },
+    {
+      "epoch": 0.6030701754385965,
+      "grad_norm": 0.8253146409988403,
+      "learning_rate": 2.959789994632645e-05,
+      "loss": 0.7927,
+      "step": 275
+    },
+    {
+      "epoch": 0.6140350877192983,
+      "grad_norm": 0.8634894490242004,
+      "learning_rate": 2.9572497932274735e-05,
+      "loss": 0.7702,
+      "step": 280
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.7633554339408875,
+      "learning_rate": 2.954632951918937e-05,
+      "loss": 0.7368,
+      "step": 285
+    },
+    {
+      "epoch": 0.6359649122807017,
+      "grad_norm": 0.9200448393821716,
+      "learning_rate": 2.951939608332354e-05,
+      "loss": 0.757,
+      "step": 290
+    },
+    {
+      "epoch": 0.6469298245614035,
+      "grad_norm": 0.8447757959365845,
+      "learning_rate": 2.9491699041164603e-05,
+      "loss": 0.753,
+      "step": 295
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 0.8062968254089355,
+      "learning_rate": 2.9463239849359624e-05,
+      "loss": 0.7379,
+      "step": 300
+    },
+    {
+      "epoch": 0.668859649122807,
+      "grad_norm": 0.8329290151596069,
+      "learning_rate": 2.9434020004638757e-05,
+      "loss": 0.6848,
+      "step": 305
+    },
+    {
+      "epoch": 0.6798245614035088,
+      "grad_norm": 0.9136500358581543,
+      "learning_rate": 2.9404041043736537e-05,
+      "loss": 0.75,
+      "step": 310
+    },
+    {
+      "epoch": 0.6907894736842105,
+      "grad_norm": 0.844235897064209,
+      "learning_rate": 2.937330454331104e-05,
+      "loss": 0.7197,
+      "step": 315
+    },
+    {
+      "epoch": 0.7017543859649122,
+      "grad_norm": 1.1511467695236206,
+      "learning_rate": 2.9341812119860993e-05,
+      "loss": 0.7365,
+      "step": 320
+    },
+    {
+      "epoch": 0.7127192982456141,
+      "grad_norm": 0.9086587429046631,
+      "learning_rate": 2.9309565429640724e-05,
+      "loss": 0.7242,
+      "step": 325
+    },
+    {
+      "epoch": 0.7236842105263158,
+      "grad_norm": 0.8911153078079224,
+      "learning_rate": 2.9276566168573093e-05,
+      "loss": 0.7308,
+      "step": 330
+    },
+    {
+      "epoch": 0.7346491228070176,
+      "grad_norm": 0.9036852717399597,
+      "learning_rate": 2.924281607216026e-05,
+      "loss": 0.7327,
+      "step": 335
+    },
+    {
+      "epoch": 0.7456140350877193,
+      "grad_norm": 0.899604856967926,
+      "learning_rate": 2.920831691539245e-05,
+      "loss": 0.6947,
+      "step": 340
+    },
+    {
+      "epoch": 0.756578947368421,
+      "grad_norm": 0.942329466342926,
+      "learning_rate": 2.9173070512654574e-05,
+      "loss": 0.674,
+      "step": 345
+    },
+    {
+      "epoch": 0.7675438596491229,
+      "grad_norm": 0.8732370138168335,
+      "learning_rate": 2.913707871763082e-05,
+      "loss": 0.6779,
+      "step": 350
+    },
+    {
+      "epoch": 0.7785087719298246,
+      "grad_norm": 0.9765074849128723,
+      "learning_rate": 2.9100343423207162e-05,
+      "loss": 0.6519,
+      "step": 355
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 1.0394716262817383,
+      "learning_rate": 2.9062866561371807e-05,
+      "loss": 0.6543,
+      "step": 360
+    },
+    {
+      "epoch": 0.8004385964912281,
+      "grad_norm": 1.048437237739563,
+      "learning_rate": 2.9024650103113586e-05,
+      "loss": 0.6694,
+      "step": 365
+    },
+    {
+      "epoch": 0.8114035087719298,
+      "grad_norm": 0.9271934628486633,
+      "learning_rate": 2.89856960583183e-05,
+      "loss": 0.6436,
+      "step": 370
+    },
+    {
+      "epoch": 0.8223684210526315,
+      "grad_norm": 0.8801912665367126,
+      "learning_rate": 2.894600647566302e-05,
+      "loss": 0.6592,
+      "step": 375
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.0545217990875244,
+      "learning_rate": 2.890558344250832e-05,
+      "loss": 0.6386,
+      "step": 380
+    },
+    {
+      "epoch": 0.8442982456140351,
+      "grad_norm": 0.8045287728309631,
+      "learning_rate": 2.8864429084788534e-05,
+      "loss": 0.6628,
+      "step": 385
+    },
+    {
+      "epoch": 0.8552631578947368,
+      "grad_norm": 0.9419070482254028,
+      "learning_rate": 2.882254556689991e-05,
+      "loss": 0.638,
+      "step": 390
+    },
+    {
+      "epoch": 0.8662280701754386,
+      "grad_norm": 0.9611677527427673,
+      "learning_rate": 2.8779935091586814e-05,
+      "loss": 0.6133,
+      "step": 395
+    },
+    {
+      "epoch": 0.8771929824561403,
+      "grad_norm": 0.9035685062408447,
+      "learning_rate": 2.873659989982586e-05,
+      "loss": 0.6229,
+      "step": 400
+    },
+    {
+      "epoch": 0.8881578947368421,
+      "grad_norm": 0.8449916839599609,
+      "learning_rate": 2.869254227070807e-05,
+      "loss": 0.6222,
+      "step": 405
+    },
+    {
+      "epoch": 0.8991228070175439,
+      "grad_norm": 1.004002332687378,
+      "learning_rate": 2.8647764521318976e-05,
+      "loss": 0.6473,
+      "step": 410
+    },
+    {
+      "epoch": 0.9100877192982456,
+      "grad_norm": 0.8791800141334534,
+      "learning_rate": 2.860226900661682e-05,
+      "loss": 0.6005,
+      "step": 415
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 0.9364210367202759,
+      "learning_rate": 2.8556058119308644e-05,
+      "loss": 0.6124,
+      "step": 420
+    },
+    {
+      "epoch": 0.9320175438596491,
+      "grad_norm": 0.8335201144218445,
+      "learning_rate": 2.8509134289724473e-05,
+      "loss": 0.6555,
+      "step": 425
+    },
+    {
+      "epoch": 0.9429824561403509,
+      "grad_norm": 0.9103546738624573,
+      "learning_rate": 2.8461499985689523e-05,
+      "loss": 0.6054,
+      "step": 430
+    },
+    {
+      "epoch": 0.9539473684210527,
+      "grad_norm": 0.9063127040863037,
+      "learning_rate": 2.841315771239436e-05,
+      "loss": 0.6164,
+      "step": 435
+    },
+    {
+      "epoch": 0.9649122807017544,
+      "grad_norm": 0.9455368518829346,
+      "learning_rate": 2.8364110012263207e-05,
+      "loss": 0.6237,
+      "step": 440
+    },
+    {
+      "epoch": 0.9758771929824561,
+      "grad_norm": 0.9800766706466675,
+      "learning_rate": 2.8314359464820184e-05,
+      "loss": 0.5732,
+      "step": 445
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "grad_norm": 0.8358300924301147,
+      "learning_rate": 2.826390868655367e-05,
+      "loss": 0.5788,
+      "step": 450
+    },
+    {
+      "epoch": 0.9978070175438597,
+      "grad_norm": 0.9222908616065979,
+      "learning_rate": 2.8212760330778688e-05,
+      "loss": 0.611,
+      "step": 455
+    },
+    {
+      "epoch": 1.0087719298245614,
+      "grad_norm": 1.228837490081787,
+      "learning_rate": 2.8160917087497368e-05,
+      "loss": 0.6056,
+      "step": 460
+    },
+    {
+      "epoch": 1.019736842105263,
+      "grad_norm": 0.966229259967804,
+      "learning_rate": 2.8108381683257468e-05,
+      "loss": 0.5213,
+      "step": 465
+    },
+    {
+      "epoch": 1.030701754385965,
+      "grad_norm": 0.9428152441978455,
+      "learning_rate": 2.8055156881008975e-05,
+      "loss": 0.5024,
+      "step": 470
+    },
+    {
+      "epoch": 1.0416666666666667,
+      "grad_norm": 0.849367082118988,
+      "learning_rate": 2.8001245479958807e-05,
+      "loss": 0.5336,
+      "step": 475
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 1.2180472612380981,
+      "learning_rate": 2.7946650315423595e-05,
+      "loss": 0.506,
+      "step": 480
+    },
+    {
+      "epoch": 1.0635964912280702,
+      "grad_norm": 0.8793379068374634,
+      "learning_rate": 2.7891374258680553e-05,
+      "loss": 0.4775,
+      "step": 485
+    },
+    {
+      "epoch": 1.0745614035087718,
+      "grad_norm": 0.9625962376594543,
+      "learning_rate": 2.783542021681649e-05,
+      "loss": 0.4986,
+      "step": 490
+    },
+    {
+      "epoch": 1.0855263157894737,
+      "grad_norm": 1.000110149383545,
+      "learning_rate": 2.7778791132574908e-05,
+      "loss": 0.5079,
+      "step": 495
+    },
+    {
+      "epoch": 1.0964912280701755,
+      "grad_norm": 0.8772032260894775,
+      "learning_rate": 2.7721489984201238e-05,
+      "loss": 0.5494,
+      "step": 500
+    },
+    {
+      "epoch": 1.1074561403508771,
+      "grad_norm": 0.9418705701828003,
+      "learning_rate": 2.766351978528622e-05,
+      "loss": 0.4887,
+      "step": 505
+    },
+    {
+      "epoch": 1.118421052631579,
+      "grad_norm": 1.0216141939163208,
+      "learning_rate": 2.7604883584607396e-05,
+      "loss": 0.5073,
+      "step": 510
+    },
+    {
+      "epoch": 1.1293859649122808,
+      "grad_norm": 1.127656102180481,
+      "learning_rate": 2.7545584465968768e-05,
+      "loss": 0.5022,
+      "step": 515
+    },
+    {
+      "epoch": 1.1403508771929824,
+      "grad_norm": 1.1076064109802246,
+      "learning_rate": 2.7485625548038636e-05,
+      "loss": 0.4659,
+      "step": 520
+    },
+    {
+      "epoch": 1.1513157894736843,
+      "grad_norm": 0.9588743448257446,
+      "learning_rate": 2.742500998418555e-05,
+      "loss": 0.4965,
+      "step": 525
+    },
+    {
+      "epoch": 1.162280701754386,
+      "grad_norm": 1.0297181606292725,
+      "learning_rate": 2.7363740962312484e-05,
+      "loss": 0.4545,
+      "step": 530
+    },
+    {
+      "epoch": 1.1732456140350878,
+      "grad_norm": 0.9798672795295715,
+      "learning_rate": 2.730182170468919e-05,
+      "loss": 0.4714,
+      "step": 535
+    },
+    {
+      "epoch": 1.1842105263157894,
+      "grad_norm": 1.0046989917755127,
+      "learning_rate": 2.7239255467782702e-05,
+      "loss": 0.4818,
+      "step": 540
+    },
+    {
+      "epoch": 1.1951754385964912,
+      "grad_norm": 1.0530850887298584,
+      "learning_rate": 2.7176045542086096e-05,
+      "loss": 0.495,
+      "step": 545
+    },
+    {
+      "epoch": 1.206140350877193,
+      "grad_norm": 1.084905743598938,
+      "learning_rate": 2.7112195251945424e-05,
+      "loss": 0.4923,
+      "step": 550
+    },
+    {
+      "epoch": 1.2171052631578947,
+      "grad_norm": 0.9365867376327515,
+      "learning_rate": 2.7047707955384878e-05,
+      "loss": 0.4198,
+      "step": 555
+    },
+    {
+      "epoch": 1.2280701754385965,
+      "grad_norm": 0.9768854379653931,
+      "learning_rate": 2.69825870439302e-05,
+      "loss": 0.4522,
+      "step": 560
+    },
+    {
+      "epoch": 1.2390350877192982,
+      "grad_norm": 1.1862937211990356,
+      "learning_rate": 2.6916835942430292e-05,
+      "loss": 0.4644,
+      "step": 565
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9556865096092224,
+      "learning_rate": 2.6850458108877112e-05,
+      "loss": 0.4495,
+      "step": 570
+    },
+    {
+      "epoch": 1.2609649122807016,
+      "grad_norm": 1.3105326890945435,
+      "learning_rate": 2.6783457034223808e-05,
+      "loss": 0.4667,
+      "step": 575
+    },
+    {
+      "epoch": 1.2719298245614035,
+      "grad_norm": 0.967395007610321,
+      "learning_rate": 2.6715836242201107e-05,
+      "loss": 0.4352,
+      "step": 580
+    },
+    {
+      "epoch": 1.2828947368421053,
+      "grad_norm": 0.9425144195556641,
+      "learning_rate": 2.6647599289132033e-05,
+      "loss": 0.5043,
+      "step": 585
+    },
+    {
+      "epoch": 1.293859649122807,
+      "grad_norm": 0.9878008365631104,
+      "learning_rate": 2.6578749763744815e-05,
+      "loss": 0.4339,
+      "step": 590
+    },
+    {
+      "epoch": 1.3048245614035088,
+      "grad_norm": 1.0370670557022095,
+      "learning_rate": 2.6509291286984194e-05,
+      "loss": 0.404,
+      "step": 595
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 1.032353401184082,
+      "learning_rate": 2.6439227511820976e-05,
+      "loss": 0.4718,
+      "step": 600
+    },
+    {
+      "epoch": 1.3267543859649122,
+      "grad_norm": 1.0497735738754272,
+      "learning_rate": 2.6368562123059907e-05,
+      "loss": 0.4717,
+      "step": 605
+    },
+    {
+      "epoch": 1.337719298245614,
+      "grad_norm": 0.9888692498207092,
+      "learning_rate": 2.629729883714589e-05,
+      "loss": 0.478,
+      "step": 610
+    },
+    {
+      "epoch": 1.3486842105263157,
+      "grad_norm": 0.9506012797355652,
+      "learning_rate": 2.622544140196852e-05,
+      "loss": 0.4658,
+      "step": 615
+    },
+    {
+      "epoch": 1.3596491228070176,
+      "grad_norm": 1.129773497581482,
+      "learning_rate": 2.6152993596665003e-05,
+      "loss": 0.4058,
+      "step": 620
+    },
+    {
+      "epoch": 1.3706140350877192,
+      "grad_norm": 1.0846400260925293,
+      "learning_rate": 2.6079959231421347e-05,
+      "loss": 0.4221,
+      "step": 625
+    },
+    {
+      "epoch": 1.381578947368421,
+      "grad_norm": 1.0114156007766724,
+      "learning_rate": 2.6006342147272037e-05,
+      "loss": 0.4096,
+      "step": 630
+    },
+    {
+      "epoch": 1.3925438596491229,
+      "grad_norm": 0.9736616611480713,
+      "learning_rate": 2.5932146215897983e-05,
+      "loss": 0.4501,
+      "step": 635
+    },
+    {
+      "epoch": 1.4035087719298245,
+      "grad_norm": 1.0868955850601196,
+      "learning_rate": 2.5857375339422926e-05,
+      "loss": 0.4208,
+      "step": 640
+    },
+    {
+      "epoch": 1.4144736842105263,
+      "grad_norm": 0.9600351452827454,
+      "learning_rate": 2.5782033450208208e-05,
+      "loss": 0.4301,
+      "step": 645
+    },
+    {
+      "epoch": 1.4254385964912282,
+      "grad_norm": 1.0037248134613037,
+      "learning_rate": 2.5706124510645955e-05,
+      "loss": 0.416,
+      "step": 650
+    },
+    {
+      "epoch": 1.4364035087719298,
+      "grad_norm": 1.1140856742858887,
+      "learning_rate": 2.5629652512950686e-05,
+      "loss": 0.4043,
+      "step": 655
+    },
+    {
+      "epoch": 1.4473684210526316,
+      "grad_norm": 1.184102177619934,
+      "learning_rate": 2.555262147894938e-05,
+      "loss": 0.3945,
+      "step": 660
+    },
+    {
+      "epoch": 1.4583333333333333,
+      "grad_norm": 1.0726951360702515,
+      "learning_rate": 2.5475035459869913e-05,
+      "loss": 0.4407,
+      "step": 665
+    },
+    {
+      "epoch": 1.469298245614035,
+      "grad_norm": 1.0315403938293457,
+      "learning_rate": 2.5396898536128035e-05,
+      "loss": 0.4129,
+      "step": 670
+    },
+    {
+      "epoch": 1.4802631578947367,
+      "grad_norm": 1.006007432937622,
+      "learning_rate": 2.531821481711276e-05,
+      "loss": 0.4063,
+      "step": 675
+    },
+    {
+      "epoch": 1.4912280701754386,
+      "grad_norm": 0.9738907217979431,
+      "learning_rate": 2.5238988440970235e-05,
+      "loss": 0.4014,
+      "step": 680
+    },
+    {
+      "epoch": 1.5021929824561404,
+      "grad_norm": 0.9589958786964417,
+      "learning_rate": 2.5159223574386117e-05,
+      "loss": 0.3696,
+      "step": 685
+    },
+    {
+      "epoch": 1.513157894736842,
+      "grad_norm": 1.093380093574524,
+      "learning_rate": 2.507892441236643e-05,
+      "loss": 0.3691,
+      "step": 690
+    },
+    {
+      "epoch": 1.5241228070175439,
+      "grad_norm": 1.1060693264007568,
+      "learning_rate": 2.499809517801696e-05,
+      "loss": 0.4062,
+      "step": 695
+    },
+    {
+      "epoch": 1.5350877192982457,
+      "grad_norm": 1.0412400960922241,
+      "learning_rate": 2.4916740122321116e-05,
+      "loss": 0.348,
+      "step": 700
+    },
+    {
+      "epoch": 1.5460526315789473,
+      "grad_norm": 1.1559436321258545,
+      "learning_rate": 2.4834863523916404e-05,
+      "loss": 0.3408,
+      "step": 705
+    },
+    {
+      "epoch": 1.557017543859649,
+      "grad_norm": 1.0723360776901245,
+      "learning_rate": 2.4752469688869364e-05,
+      "loss": 0.3901,
+      "step": 710
+    },
+    {
+      "epoch": 1.567982456140351,
+      "grad_norm": 0.9203932285308838,
+      "learning_rate": 2.4669562950449136e-05,
+      "loss": 0.3711,
+      "step": 715
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.9626147150993347,
+      "learning_rate": 2.4586147668899557e-05,
+      "loss": 0.3773,
+      "step": 720
+    },
+    {
+      "epoch": 1.5899122807017543,
+      "grad_norm": 0.969040036201477,
+      "learning_rate": 2.450222823120983e-05,
+      "loss": 0.3898,
+      "step": 725
+    },
+    {
+      "epoch": 1.6008771929824561,
+      "grad_norm": 0.9797758460044861,
+      "learning_rate": 2.4417809050883824e-05,
+      "loss": 0.396,
+      "step": 730
+    },
+    {
+      "epoch": 1.611842105263158,
+      "grad_norm": 0.9515657424926758,
+      "learning_rate": 2.433289456770794e-05,
+      "loss": 0.3743,
+      "step": 735
+    },
+    {
+      "epoch": 1.6228070175438596,
+      "grad_norm": 0.9654616117477417,
+      "learning_rate": 2.4247489247517635e-05,
+      "loss": 0.3971,
+      "step": 740
+    },
+    {
+      "epoch": 1.6337719298245614,
+      "grad_norm": 0.9589864015579224,
+      "learning_rate": 2.4161597581962526e-05,
+      "loss": 0.3406,
+      "step": 745
+    },
+    {
+      "epoch": 1.6447368421052633,
+      "grad_norm": 1.0553503036499023,
+      "learning_rate": 2.4075224088270186e-05,
+      "loss": 0.3449,
+      "step": 750
+    },
+    {
+      "epoch": 1.655701754385965,
+      "grad_norm": 0.9229257106781006,
+      "learning_rate": 2.3988373309008575e-05,
+      "loss": 0.3879,
+      "step": 755
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 1.0924843549728394,
+      "learning_rate": 2.390104981184712e-05,
+      "loss": 0.3531,
+      "step": 760
+    },
+    {
+      "epoch": 1.6776315789473686,
+      "grad_norm": 1.0531260967254639,
+      "learning_rate": 2.3813258189316488e-05,
+      "loss": 0.3315,
+      "step": 765
+    },
+    {
+      "epoch": 1.6885964912280702,
+      "grad_norm": 1.1287767887115479,
+      "learning_rate": 2.3725003058567103e-05,
+      "loss": 0.3269,
+      "step": 770
+    },
+    {
+      "epoch": 1.6995614035087718,
+      "grad_norm": 1.0668278932571411,
+      "learning_rate": 2.3636289061126244e-05,
+      "loss": 0.3821,
+      "step": 775
+    },
+    {
+      "epoch": 1.7105263157894737,
+      "grad_norm": 1.0022592544555664,
+      "learning_rate": 2.3547120862653995e-05,
+      "loss": 0.3232,
+      "step": 780
+    },
+    {
+      "epoch": 1.7214912280701755,
+      "grad_norm": 0.9137780070304871,
+      "learning_rate": 2.3457503152697853e-05,
+      "loss": 0.2921,
+      "step": 785
+    },
+    {
+      "epoch": 1.7324561403508771,
+      "grad_norm": 0.9505869150161743,
+      "learning_rate": 2.3367440644446084e-05,
+      "loss": 0.3401,
+      "step": 790
+    },
+    {
+      "epoch": 1.743421052631579,
+      "grad_norm": 0.9819185137748718,
+      "learning_rate": 2.3276938074479862e-05,
+      "loss": 0.3571,
+      "step": 795
+    },
+    {
+      "epoch": 1.7543859649122808,
+      "grad_norm": 1.0384228229522705,
+      "learning_rate": 2.3186000202524146e-05,
+      "loss": 0.3348,
+      "step": 800
+    },
+    {
+      "epoch": 1.7653508771929824,
+      "grad_norm": 1.0425379276275635,
+      "learning_rate": 2.309463181119736e-05,
+      "loss": 0.2937,
+      "step": 805
+    },
+    {
+      "epoch": 1.776315789473684,
+      "grad_norm": 1.0403761863708496,
+      "learning_rate": 2.300283770575989e-05,
+      "loss": 0.3164,
+      "step": 810
+    },
+    {
+      "epoch": 1.787280701754386,
+      "grad_norm": 0.9930003881454468,
+      "learning_rate": 2.2910622713861313e-05,
+      "loss": 0.3322,
+      "step": 815
+    },
+    {
+      "epoch": 1.7982456140350878,
+      "grad_norm": 0.8970758318901062,
+      "learning_rate": 2.2817991685286557e-05,
+      "loss": 0.3245,
+      "step": 820
+    },
+    {
+      "epoch": 1.8092105263157894,
+      "grad_norm": 1.084914207458496,
+      "learning_rate": 2.2724949491700797e-05,
+      "loss": 0.3044,
+      "step": 825
+    },
+    {
+      "epoch": 1.8201754385964912,
+      "grad_norm": 0.9980108141899109,
+      "learning_rate": 2.2631501026393267e-05,
+      "loss": 0.2933,
+      "step": 830
+    },
+    {
+      "epoch": 1.831140350877193,
+      "grad_norm": 1.011869192123413,
+      "learning_rate": 2.2537651204019915e-05,
+      "loss": 0.3133,
+      "step": 835
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 1.1241363286972046,
+      "learning_rate": 2.2443404960344902e-05,
+      "loss": 0.3483,
+      "step": 840
+    },
+    {
+      "epoch": 1.8530701754385965,
+      "grad_norm": 0.9130306839942932,
+      "learning_rate": 2.2348767251981048e-05,
+      "loss": 0.3212,
+      "step": 845
+    },
+    {
+      "epoch": 1.8640350877192984,
+      "grad_norm": 1.0724334716796875,
+      "learning_rate": 2.2253743056129154e-05,
+      "loss": 0.339,
+      "step": 850
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 1.1117773056030273,
+      "learning_rate": 2.2158337370316213e-05,
+      "loss": 0.3165,
+      "step": 855
+    },
+    {
+      "epoch": 1.8859649122807016,
+      "grad_norm": 1.1159254312515259,
+      "learning_rate": 2.2062555212132603e-05,
+      "loss": 0.3226,
+      "step": 860
+    },
+    {
+      "epoch": 1.8969298245614035,
+      "grad_norm": 1.09803307056427,
+      "learning_rate": 2.1966401618968194e-05,
+      "loss": 0.2926,
+      "step": 865
+    },
+    {
+      "epoch": 1.9078947368421053,
+      "grad_norm": 1.00844407081604,
+      "learning_rate": 2.1869881647747427e-05,
+      "loss": 0.295,
+      "step": 870
+    },
+    {
+      "epoch": 1.918859649122807,
+      "grad_norm": 0.9908561110496521,
+      "learning_rate": 2.177300037466334e-05,
+      "loss": 0.2741,
+      "step": 875
+    },
+    {
+      "epoch": 1.9298245614035088,
+      "grad_norm": 0.9940894246101379,
+      "learning_rate": 2.167576289491063e-05,
+      "loss": 0.314,
+      "step": 880
+    },
+    {
+      "epoch": 1.9407894736842106,
+      "grad_norm": 0.9110603332519531,
+      "learning_rate": 2.157817432241766e-05,
+      "loss": 0.2681,
+      "step": 885
+    },
+    {
+      "epoch": 1.9517543859649122,
+      "grad_norm": 0.9817435145378113,
+      "learning_rate": 2.148023978957751e-05,
+      "loss": 0.3066,
+      "step": 890
+    },
+    {
+      "epoch": 1.962719298245614,
+      "grad_norm": 1.0423473119735718,
+      "learning_rate": 2.1381964446978078e-05,
+      "loss": 0.2846,
+      "step": 895
+    },
+    {
+      "epoch": 1.973684210526316,
+      "grad_norm": 0.9325022101402283,
+      "learning_rate": 2.1283353463131163e-05,
+      "loss": 0.2777,
+      "step": 900
+    },
+    {
+      "epoch": 1.9846491228070176,
+      "grad_norm": 0.9839675426483154,
+      "learning_rate": 2.1184412024200663e-05,
+      "loss": 0.3071,
+      "step": 905
+    },
+    {
+      "epoch": 1.9956140350877192,
+      "grad_norm": 1.064647912979126,
+      "learning_rate": 2.108514533372983e-05,
+      "loss": 0.3028,
+      "step": 910
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 2280,
+  "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": 1.3328166782196777e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

9/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3137701cca554e162b236c231363e5a59d8882d8bb19f4b66f407c6a042003eb
+size 7736

9/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)