Upload folder using huggingface_hub

base/27_128_e3_3e-5/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

base/27_128_e3_3e-5/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": [
+    "q_proj",
+    "k_proj",
+    "up_proj",
+    "o_proj",
+    "v_proj",
+    "gate_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

base/27_128_e3_3e-5/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20eb958840f9e4c97bd6d094068a2fbf75c68e57bb0b9840b6467f92fcb7ebc1
+size 791751704

base/27_128_e3_3e-5/latest

@@ -0,0 +1 @@
+global_step876

base/27_128_e3_3e-5/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2da6c43392ff074d2078d1f98446ee6f2c37985f6300138ba5d96317aa2d369c
+size 16389

base/27_128_e3_3e-5/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e720b7240c6ca3adc47f9995b4ecec1d629ebd1e670093a6c6711a3c5add95d
+size 16389

base/27_128_e3_3e-5/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73a96890ed44bd2838f8a5fdc04c822e7f40c28d8c0cabe099c4f568a462fc14
+size 16389

base/27_128_e3_3e-5/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb31f81f2f74c8fd3b62b1035736280705a7be02bf2be1fea1002c2ba23ffecf
+size 16389

base/27_128_e3_3e-5/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4dd141d526e6808072039465ba0e3515579022dcdc26184a3e9b0dde7d16efad
+size 16389

base/27_128_e3_3e-5/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6be138e2d5e22eece141d1aec9ae04ab2b60741d10e1ddc25a2746255826469
+size 16389

base/27_128_e3_3e-5/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1d197d90f426f4a8756f752d866ec473a450d6519a7501d420c89a35175a2d0
+size 16389

base/27_128_e3_3e-5/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0984a3dd2af4470e659ccd40cda96313ce1ceb404c9428d22ce6b5a3088871c3
+size 16389

base/27_128_e3_3e-5/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae17c1bf2f4a3b70966840d66947a882d1cecba3b18f9e6a4743b48808610073
+size 1401

base/27_128_e3_3e-5/special_tokens_map.json

@@ -0,0 +1,45 @@
+{
+  "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": "<reponame>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

base/27_128_e3_3e-5/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": "<reponame>",
+  "padding_side": "left",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

base/27_128_e3_3e-5/trainer_state.json

@@ -0,0 +1,1259 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 876,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.017123287671232876,
+      "grad_norm": 0.8992072939872742,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.2593,
+      "step": 5
+    },
+    {
+      "epoch": 0.03424657534246575,
+      "grad_norm": 0.686211109161377,
+      "learning_rate": 6.136363636363637e-06,
+      "loss": 1.3131,
+      "step": 10
+    },
+    {
+      "epoch": 0.05136986301369863,
+      "grad_norm": 0.609504222869873,
+      "learning_rate": 9.545454545454545e-06,
+      "loss": 1.262,
+      "step": 15
+    },
+    {
+      "epoch": 0.0684931506849315,
+      "grad_norm": 0.5175422430038452,
+      "learning_rate": 1.2954545454545455e-05,
+      "loss": 1.2124,
+      "step": 20
+    },
+    {
+      "epoch": 0.08561643835616438,
+      "grad_norm": 0.41778406500816345,
+      "learning_rate": 1.6363636363636363e-05,
+      "loss": 1.2041,
+      "step": 25
+    },
+    {
+      "epoch": 0.10273972602739725,
+      "grad_norm": 0.42357075214385986,
+      "learning_rate": 1.9772727272727274e-05,
+      "loss": 1.135,
+      "step": 30
+    },
+    {
+      "epoch": 0.11986301369863013,
+      "grad_norm": 0.4278438687324524,
+      "learning_rate": 2.318181818181818e-05,
+      "loss": 1.1928,
+      "step": 35
+    },
+    {
+      "epoch": 0.136986301369863,
+      "grad_norm": 0.4151443541049957,
+      "learning_rate": 2.6590909090909093e-05,
+      "loss": 1.133,
+      "step": 40
+    },
+    {
+      "epoch": 0.1541095890410959,
+      "grad_norm": 0.42306843400001526,
+      "learning_rate": 3e-05,
+      "loss": 1.1583,
+      "step": 45
+    },
+    {
+      "epoch": 0.17123287671232876,
+      "grad_norm": 0.6116857528686523,
+      "learning_rate": 2.999732673837156e-05,
+      "loss": 1.0129,
+      "step": 50
+    },
+    {
+      "epoch": 0.18835616438356165,
+      "grad_norm": 0.4719073176383972,
+      "learning_rate": 2.9989307906329936e-05,
+      "loss": 1.0218,
+      "step": 55
+    },
+    {
+      "epoch": 0.2054794520547945,
+      "grad_norm": 0.5087824463844299,
+      "learning_rate": 2.9975946362066596e-05,
+      "loss": 1.0357,
+      "step": 60
+    },
+    {
+      "epoch": 0.2226027397260274,
+      "grad_norm": 0.6392008066177368,
+      "learning_rate": 2.995724686810202e-05,
+      "loss": 1.0296,
+      "step": 65
+    },
+    {
+      "epoch": 0.23972602739726026,
+      "grad_norm": 0.6877881288528442,
+      "learning_rate": 2.9933216089588158e-05,
+      "loss": 0.979,
+      "step": 70
+    },
+    {
+      "epoch": 0.2568493150684932,
+      "grad_norm": 0.5768792629241943,
+      "learning_rate": 2.9903862591932762e-05,
+      "loss": 0.9785,
+      "step": 75
+    },
+    {
+      "epoch": 0.273972602739726,
+      "grad_norm": 0.6000942587852478,
+      "learning_rate": 2.986919683774636e-05,
+      "loss": 0.9618,
+      "step": 80
+    },
+    {
+      "epoch": 0.2910958904109589,
+      "grad_norm": 0.5740535259246826,
+      "learning_rate": 2.9829231183113013e-05,
+      "loss": 0.8941,
+      "step": 85
+    },
+    {
+      "epoch": 0.3082191780821918,
+      "grad_norm": 0.6583782434463501,
+      "learning_rate": 2.9783979873186188e-05,
+      "loss": 0.8784,
+      "step": 90
+    },
+    {
+      "epoch": 0.3253424657534247,
+      "grad_norm": 0.6332166194915771,
+      "learning_rate": 2.973345903711128e-05,
+      "loss": 0.8786,
+      "step": 95
+    },
+    {
+      "epoch": 0.3424657534246575,
+      "grad_norm": 0.7230481505393982,
+      "learning_rate": 2.9677686682276623e-05,
+      "loss": 0.8577,
+      "step": 100
+    },
+    {
+      "epoch": 0.3595890410958904,
+      "grad_norm": 0.7715655565261841,
+      "learning_rate": 2.9616682687895038e-05,
+      "loss": 0.8136,
+      "step": 105
+    },
+    {
+      "epoch": 0.3767123287671233,
+      "grad_norm": 0.861534833908081,
+      "learning_rate": 2.9550468797918162e-05,
+      "loss": 0.7541,
+      "step": 110
+    },
+    {
+      "epoch": 0.3938356164383562,
+      "grad_norm": 0.8687571883201599,
+      "learning_rate": 2.947906861328618e-05,
+      "loss": 0.7291,
+      "step": 115
+    },
+    {
+      "epoch": 0.410958904109589,
+      "grad_norm": 0.7383458614349365,
+      "learning_rate": 2.9402507583515604e-05,
+      "loss": 0.7574,
+      "step": 120
+    },
+    {
+      "epoch": 0.4280821917808219,
+      "grad_norm": 0.8834396004676819,
+      "learning_rate": 2.9320812997628184e-05,
+      "loss": 0.7187,
+      "step": 125
+    },
+    {
+      "epoch": 0.4452054794520548,
+      "grad_norm": 0.9517204165458679,
+      "learning_rate": 2.923401397442415e-05,
+      "loss": 0.6853,
+      "step": 130
+    },
+    {
+      "epoch": 0.4623287671232877,
+      "grad_norm": 0.8729910254478455,
+      "learning_rate": 2.914214145210324e-05,
+      "loss": 0.7112,
+      "step": 135
+    },
+    {
+      "epoch": 0.4794520547945205,
+      "grad_norm": 1.302199363708496,
+      "learning_rate": 2.9045228177237285e-05,
+      "loss": 0.6494,
+      "step": 140
+    },
+    {
+      "epoch": 0.4965753424657534,
+      "grad_norm": 0.8917989730834961,
+      "learning_rate": 2.894330869309814e-05,
+      "loss": 0.679,
+      "step": 145
+    },
+    {
+      "epoch": 0.5136986301369864,
+      "grad_norm": 1.0300097465515137,
+      "learning_rate": 2.8836419327345297e-05,
+      "loss": 0.6395,
+      "step": 150
+    },
+    {
+      "epoch": 0.5308219178082192,
+      "grad_norm": 0.9277153611183167,
+      "learning_rate": 2.8724598179077413e-05,
+      "loss": 0.6514,
+      "step": 155
+    },
+    {
+      "epoch": 0.547945205479452,
+      "grad_norm": 0.9896652698516846,
+      "learning_rate": 2.8607885105252473e-05,
+      "loss": 0.6081,
+      "step": 160
+    },
+    {
+      "epoch": 0.565068493150685,
+      "grad_norm": 0.9680744409561157,
+      "learning_rate": 2.848632170648139e-05,
+      "loss": 0.58,
+      "step": 165
+    },
+    {
+      "epoch": 0.5821917808219178,
+      "grad_norm": 0.998518705368042,
+      "learning_rate": 2.8359951312200077e-05,
+      "loss": 0.5747,
+      "step": 170
+    },
+    {
+      "epoch": 0.5993150684931506,
+      "grad_norm": 0.9801236391067505,
+      "learning_rate": 2.8228818965225325e-05,
+      "loss": 0.5791,
+      "step": 175
+    },
+    {
+      "epoch": 0.6164383561643836,
+      "grad_norm": 1.064215064048767,
+      "learning_rate": 2.8092971405700004e-05,
+      "loss": 0.5642,
+      "step": 180
+    },
+    {
+      "epoch": 0.6335616438356164,
+      "grad_norm": 1.0428051948547363,
+      "learning_rate": 2.7952457054433193e-05,
+      "loss": 0.55,
+      "step": 185
+    },
+    {
+      "epoch": 0.6506849315068494,
+      "grad_norm": 1.105124831199646,
+      "learning_rate": 2.780732599564137e-05,
+      "loss": 0.5283,
+      "step": 190
+    },
+    {
+      "epoch": 0.6678082191780822,
+      "grad_norm": 1.0974406003952026,
+      "learning_rate": 2.76576299590966e-05,
+      "loss": 0.5229,
+      "step": 195
+    },
+    {
+      "epoch": 0.684931506849315,
+      "grad_norm": 1.5371181964874268,
+      "learning_rate": 2.7503422301688276e-05,
+      "loss": 0.5339,
+      "step": 200
+    },
+    {
+      "epoch": 0.702054794520548,
+      "grad_norm": 0.9692041277885437,
+      "learning_rate": 2.7344757988404845e-05,
+      "loss": 0.4879,
+      "step": 205
+    },
+    {
+      "epoch": 0.7191780821917808,
+      "grad_norm": 1.2286427021026611,
+      "learning_rate": 2.718169357274238e-05,
+      "loss": 0.4744,
+      "step": 210
+    },
+    {
+      "epoch": 0.7363013698630136,
+      "grad_norm": 1.2261204719543457,
+      "learning_rate": 2.7014287176546922e-05,
+      "loss": 0.4662,
+      "step": 215
+    },
+    {
+      "epoch": 0.7534246575342466,
+      "grad_norm": 1.2392516136169434,
+      "learning_rate": 2.6842598469297846e-05,
+      "loss": 0.4905,
+      "step": 220
+    },
+    {
+      "epoch": 0.7705479452054794,
+      "grad_norm": 1.1272625923156738,
+      "learning_rate": 2.6666688646839574e-05,
+      "loss": 0.4704,
+      "step": 225
+    },
+    {
+      "epoch": 0.7876712328767124,
+      "grad_norm": 1.1429407596588135,
+      "learning_rate": 2.6486620409569222e-05,
+      "loss": 0.442,
+      "step": 230
+    },
+    {
+      "epoch": 0.8047945205479452,
+      "grad_norm": 1.1834821701049805,
+      "learning_rate": 2.6302457940088024e-05,
+      "loss": 0.4392,
+      "step": 235
+    },
+    {
+      "epoch": 0.821917808219178,
+      "grad_norm": 1.1873235702514648,
+      "learning_rate": 2.611426688032439e-05,
+      "loss": 0.4514,
+      "step": 240
+    },
+    {
+      "epoch": 0.839041095890411,
+      "grad_norm": 1.210160732269287,
+      "learning_rate": 2.5922114308136826e-05,
+      "loss": 0.4705,
+      "step": 245
+    },
+    {
+      "epoch": 0.8561643835616438,
+      "grad_norm": 1.2718369960784912,
+      "learning_rate": 2.5726068713405084e-05,
+      "loss": 0.3893,
+      "step": 250
+    },
+    {
+      "epoch": 0.8732876712328768,
+      "grad_norm": 1.2908711433410645,
+      "learning_rate": 2.5526199973617932e-05,
+      "loss": 0.4112,
+      "step": 255
+    },
+    {
+      "epoch": 0.8904109589041096,
+      "grad_norm": 1.1591285467147827,
+      "learning_rate": 2.532257932896641e-05,
+      "loss": 0.4073,
+      "step": 260
+    },
+    {
+      "epoch": 0.9075342465753424,
+      "grad_norm": 1.0960865020751953,
+      "learning_rate": 2.511527935695133e-05,
+      "loss": 0.3853,
+      "step": 265
+    },
+    {
+      "epoch": 0.9246575342465754,
+      "grad_norm": 1.0500437021255493,
+      "learning_rate": 2.4904373946514136e-05,
+      "loss": 0.4019,
+      "step": 270
+    },
+    {
+      "epoch": 0.9417808219178082,
+      "grad_norm": 1.0678784847259521,
+      "learning_rate": 2.468993827170028e-05,
+      "loss": 0.3461,
+      "step": 275
+    },
+    {
+      "epoch": 0.958904109589041,
+      "grad_norm": 1.1523617506027222,
+      "learning_rate": 2.4472048764864602e-05,
+      "loss": 0.3523,
+      "step": 280
+    },
+    {
+      "epoch": 0.976027397260274,
+      "grad_norm": 1.2985448837280273,
+      "learning_rate": 2.425078308942815e-05,
+      "loss": 0.3768,
+      "step": 285
+    },
+    {
+      "epoch": 0.9931506849315068,
+      "grad_norm": 1.0816726684570312,
+      "learning_rate": 2.402622011219622e-05,
+      "loss": 0.3778,
+      "step": 290
+    },
+    {
+      "epoch": 1.0102739726027397,
+      "grad_norm": 1.247622013092041,
+      "learning_rate": 2.379843987524753e-05,
+      "loss": 0.2685,
+      "step": 295
+    },
+    {
+      "epoch": 1.0273972602739727,
+      "grad_norm": 1.1854517459869385,
+      "learning_rate": 2.3567523567404346e-05,
+      "loss": 0.3149,
+      "step": 300
+    },
+    {
+      "epoch": 1.0445205479452055,
+      "grad_norm": 1.4153354167938232,
+      "learning_rate": 2.3333553495294033e-05,
+      "loss": 0.3364,
+      "step": 305
+    },
+    {
+      "epoch": 1.0616438356164384,
+      "grad_norm": 1.265658974647522,
+      "learning_rate": 2.309661305401205e-05,
+      "loss": 0.3206,
+      "step": 310
+    },
+    {
+      "epoch": 1.0787671232876712,
+      "grad_norm": 1.4305566549301147,
+      "learning_rate": 2.285678669739705e-05,
+      "loss": 0.3169,
+      "step": 315
+    },
+    {
+      "epoch": 1.095890410958904,
+      "grad_norm": 1.1760265827178955,
+      "learning_rate": 2.2614159907928588e-05,
+      "loss": 0.2646,
+      "step": 320
+    },
+    {
+      "epoch": 1.1130136986301369,
+      "grad_norm": 1.2012296915054321,
+      "learning_rate": 2.236881916625816e-05,
+      "loss": 0.2822,
+      "step": 325
+    },
+    {
+      "epoch": 1.13013698630137,
+      "grad_norm": 1.2347042560577393,
+      "learning_rate": 2.212085192038453e-05,
+      "loss": 0.2647,
+      "step": 330
+    },
+    {
+      "epoch": 1.1472602739726028,
+      "grad_norm": 1.1436082124710083,
+      "learning_rate": 2.1870346554484154e-05,
+      "loss": 0.2705,
+      "step": 335
+    },
+    {
+      "epoch": 1.1643835616438356,
+      "grad_norm": 1.1666163206100464,
+      "learning_rate": 2.161739235740802e-05,
+      "loss": 0.2819,
+      "step": 340
+    },
+    {
+      "epoch": 1.1815068493150684,
+      "grad_norm": 1.696736216545105,
+      "learning_rate": 2.1362079490855968e-05,
+      "loss": 0.2371,
+      "step": 345
+    },
+    {
+      "epoch": 1.1986301369863013,
+      "grad_norm": 1.3252989053726196,
+      "learning_rate": 2.110449895723991e-05,
+      "loss": 0.3085,
+      "step": 350
+    },
+    {
+      "epoch": 1.2157534246575343,
+      "grad_norm": 1.2029995918273926,
+      "learning_rate": 2.084474256724743e-05,
+      "loss": 0.2801,
+      "step": 355
+    },
+    {
+      "epoch": 1.2328767123287672,
+      "grad_norm": 1.111670732498169,
+      "learning_rate": 2.0582902907117193e-05,
+      "loss": 0.2363,
+      "step": 360
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.2142661809921265,
+      "learning_rate": 2.0319073305638035e-05,
+      "loss": 0.2182,
+      "step": 365
+    },
+    {
+      "epoch": 1.2671232876712328,
+      "grad_norm": 1.2321888208389282,
+      "learning_rate": 2.00533478008833e-05,
+      "loss": 0.2399,
+      "step": 370
+    },
+    {
+      "epoch": 1.2842465753424657,
+      "grad_norm": 1.1036635637283325,
+      "learning_rate": 1.97858211066924e-05,
+      "loss": 0.2191,
+      "step": 375
+    },
+    {
+      "epoch": 1.3013698630136985,
+      "grad_norm": 1.0933023691177368,
+      "learning_rate": 1.9516588578911484e-05,
+      "loss": 0.2082,
+      "step": 380
+    },
+    {
+      "epoch": 1.3184931506849316,
+      "grad_norm": 1.0782479047775269,
+      "learning_rate": 1.9245746181405306e-05,
+      "loss": 0.2077,
+      "step": 385
+    },
+    {
+      "epoch": 1.3356164383561644,
+      "grad_norm": 1.0349798202514648,
+      "learning_rate": 1.8973390451852348e-05,
+      "loss": 0.2284,
+      "step": 390
+    },
+    {
+      "epoch": 1.3527397260273972,
+      "grad_norm": 1.1387205123901367,
+      "learning_rate": 1.8699618467335428e-05,
+      "loss": 0.2147,
+      "step": 395
+    },
+    {
+      "epoch": 1.36986301369863,
+      "grad_norm": 1.1331334114074707,
+      "learning_rate": 1.8424527809740028e-05,
+      "loss": 0.2295,
+      "step": 400
+    },
+    {
+      "epoch": 1.3869863013698631,
+      "grad_norm": 1.1048861742019653,
+      "learning_rate": 1.8148216530972714e-05,
+      "loss": 0.2174,
+      "step": 405
+    },
+    {
+      "epoch": 1.404109589041096,
+      "grad_norm": 1.2603099346160889,
+      "learning_rate": 1.7870783118012034e-05,
+      "loss": 0.2067,
+      "step": 410
+    },
+    {
+      "epoch": 1.4212328767123288,
+      "grad_norm": 1.1757439374923706,
+      "learning_rate": 1.7592326457804295e-05,
+      "loss": 0.1849,
+      "step": 415
+    },
+    {
+      "epoch": 1.4383561643835616,
+      "grad_norm": 1.2938828468322754,
+      "learning_rate": 1.7312945802016817e-05,
+      "loss": 0.1823,
+      "step": 420
+    },
+    {
+      "epoch": 1.4554794520547945,
+      "grad_norm": 1.0625684261322021,
+      "learning_rate": 1.7032740731661178e-05,
+      "loss": 0.1763,
+      "step": 425
+    },
+    {
+      "epoch": 1.4726027397260273,
+      "grad_norm": 1.1918516159057617,
+      "learning_rate": 1.675181112159907e-05,
+      "loss": 0.2128,
+      "step": 430
+    },
+    {
+      "epoch": 1.4897260273972603,
+      "grad_norm": 1.1850873231887817,
+      "learning_rate": 1.6470257104943414e-05,
+      "loss": 0.1854,
+      "step": 435
+    },
+    {
+      "epoch": 1.5068493150684932,
+      "grad_norm": 1.1940261125564575,
+      "learning_rate": 1.618817903736741e-05,
+      "loss": 0.1876,
+      "step": 440
+    },
+    {
+      "epoch": 1.523972602739726,
+      "grad_norm": 1.21336030960083,
+      "learning_rate": 1.5905677461334292e-05,
+      "loss": 0.1807,
+      "step": 445
+    },
+    {
+      "epoch": 1.541095890410959,
+      "grad_norm": 1.1555956602096558,
+      "learning_rate": 1.5622853070260492e-05,
+      "loss": 0.181,
+      "step": 450
+    },
+    {
+      "epoch": 1.558219178082192,
+      "grad_norm": 1.279056429862976,
+      "learning_rate": 1.5339806672624982e-05,
+      "loss": 0.1568,
+      "step": 455
+    },
+    {
+      "epoch": 1.5753424657534247,
+      "grad_norm": 1.2448151111602783,
+      "learning_rate": 1.5056639156037597e-05,
+      "loss": 0.1724,
+      "step": 460
+    },
+    {
+      "epoch": 1.5924657534246576,
+      "grad_norm": 1.1050242185592651,
+      "learning_rate": 1.4773451451279213e-05,
+      "loss": 0.1894,
+      "step": 465
+    },
+    {
+      "epoch": 1.6095890410958904,
+      "grad_norm": 1.309141993522644,
+      "learning_rate": 1.4490344496326463e-05,
+      "loss": 0.1547,
+      "step": 470
+    },
+    {
+      "epoch": 1.6267123287671232,
+      "grad_norm": 1.1085047721862793,
+      "learning_rate": 1.4207419200373942e-05,
+      "loss": 0.1365,
+      "step": 475
+    },
+    {
+      "epoch": 1.643835616438356,
+      "grad_norm": 1.0569626092910767,
+      "learning_rate": 1.3924776407866634e-05,
+      "loss": 0.1455,
+      "step": 480
+    },
+    {
+      "epoch": 1.660958904109589,
+      "grad_norm": 1.2008103132247925,
+      "learning_rate": 1.3642516862555433e-05,
+      "loss": 0.155,
+      "step": 485
+    },
+    {
+      "epoch": 1.678082191780822,
+      "grad_norm": 1.1394250392913818,
+      "learning_rate": 1.3360741171588578e-05,
+      "loss": 0.138,
+      "step": 490
+    },
+    {
+      "epoch": 1.6952054794520548,
+      "grad_norm": 1.2802852392196655,
+      "learning_rate": 1.3079549769651737e-05,
+      "loss": 0.1875,
+      "step": 495
+    },
+    {
+      "epoch": 1.7123287671232876,
+      "grad_norm": 1.1892162561416626,
+      "learning_rate": 1.2799042883169576e-05,
+      "loss": 0.1714,
+      "step": 500
+    },
+    {
+      "epoch": 1.7294520547945207,
+      "grad_norm": 1.1918212175369263,
+      "learning_rate": 1.2519320494581581e-05,
+      "loss": 0.1418,
+      "step": 505
+    },
+    {
+      "epoch": 1.7465753424657535,
+      "grad_norm": 1.1229795217514038,
+      "learning_rate": 1.2240482306704831e-05,
+      "loss": 0.1482,
+      "step": 510
+    },
+    {
+      "epoch": 1.7636986301369864,
+      "grad_norm": 1.0794895887374878,
+      "learning_rate": 1.1962627707196407e-05,
+      "loss": 0.1459,
+      "step": 515
+    },
+    {
+      "epoch": 1.7808219178082192,
+      "grad_norm": 1.0708686113357544,
+      "learning_rate": 1.1685855733128203e-05,
+      "loss": 0.1485,
+      "step": 520
+    },
+    {
+      "epoch": 1.797945205479452,
+      "grad_norm": 1.2428174018859863,
+      "learning_rate": 1.1410265035686639e-05,
+      "loss": 0.1315,
+      "step": 525
+    },
+    {
+      "epoch": 1.8150684931506849,
+      "grad_norm": 1.053895115852356,
+      "learning_rate": 1.1135953845009914e-05,
+      "loss": 0.1204,
+      "step": 530
+    },
+    {
+      "epoch": 1.8321917808219177,
+      "grad_norm": 1.1570361852645874,
+      "learning_rate": 1.0863019935175415e-05,
+      "loss": 0.1284,
+      "step": 535
+    },
+    {
+      "epoch": 1.8493150684931505,
+      "grad_norm": 1.2741470336914062,
+      "learning_rate": 1.0591560589349568e-05,
+      "loss": 0.1278,
+      "step": 540
+    },
+    {
+      "epoch": 1.8664383561643836,
+      "grad_norm": 1.00489342212677,
+      "learning_rate": 1.0321672565112767e-05,
+      "loss": 0.1327,
+      "step": 545
+    },
+    {
+      "epoch": 1.8835616438356164,
+      "grad_norm": 1.2728906869888306,
+      "learning_rate": 1.0053452059971555e-05,
+      "loss": 0.1129,
+      "step": 550
+    },
+    {
+      "epoch": 1.9006849315068495,
+      "grad_norm": 1.1312276124954224,
+      "learning_rate": 9.786994677070523e-06,
+      "loss": 0.1136,
+      "step": 555
+    },
+    {
+      "epoch": 1.9178082191780823,
+      "grad_norm": 1.1324329376220703,
+      "learning_rate": 9.52239539111598e-06,
+      "loss": 0.1221,
+      "step": 560
+    },
+    {
+      "epoch": 1.9349315068493151,
+      "grad_norm": 0.9352611303329468,
+      "learning_rate": 9.259748514523654e-06,
+      "loss": 0.1076,
+      "step": 565
+    },
+    {
+      "epoch": 1.952054794520548,
+      "grad_norm": 1.131257176399231,
+      "learning_rate": 8.999147663802494e-06,
+      "loss": 0.1142,
+      "step": 570
+    },
+    {
+      "epoch": 1.9691780821917808,
+      "grad_norm": 1.1916224956512451,
+      "learning_rate": 8.740685726186445e-06,
+      "loss": 0.1131,
+      "step": 575
+    },
+    {
+      "epoch": 1.9863013698630136,
+      "grad_norm": 1.1361243724822998,
+      "learning_rate": 8.484454826526199e-06,
+      "loss": 0.1242,
+      "step": 580
+    },
+    {
+      "epoch": 2.0034246575342465,
+      "grad_norm": 0.956641674041748,
+      "learning_rate": 8.2305462944527e-06,
+      "loss": 0.1195,
+      "step": 585
+    },
+    {
+      "epoch": 2.0205479452054793,
+      "grad_norm": 0.8145921230316162,
+      "learning_rate": 7.979050631824074e-06,
+      "loss": 0.0951,
+      "step": 590
+    },
+    {
+      "epoch": 2.037671232876712,
+      "grad_norm": 0.8547300696372986,
+      "learning_rate": 7.730057480467604e-06,
+      "loss": 0.0903,
+      "step": 595
+    },
+    {
+      "epoch": 2.0547945205479454,
+      "grad_norm": 0.7731233835220337,
+      "learning_rate": 7.4836555902282534e-06,
+      "loss": 0.086,
+      "step": 600
+    },
+    {
+      "epoch": 2.0719178082191783,
+      "grad_norm": 0.975101888179779,
+      "learning_rate": 7.239932787335147e-06,
+      "loss": 0.0982,
+      "step": 605
+    },
+    {
+      "epoch": 2.089041095890411,
+      "grad_norm": 0.8168078660964966,
+      "learning_rate": 6.9989759430972105e-06,
+      "loss": 0.075,
+      "step": 610
+    },
+    {
+      "epoch": 2.106164383561644,
+      "grad_norm": 0.9285591244697571,
+      "learning_rate": 6.760870942939202e-06,
+      "loss": 0.0854,
+      "step": 615
+    },
+    {
+      "epoch": 2.1232876712328768,
+      "grad_norm": 1.3084709644317627,
+      "learning_rate": 6.525702655789201e-06,
+      "loss": 0.0754,
+      "step": 620
+    },
+    {
+      "epoch": 2.1404109589041096,
+      "grad_norm": 0.9635082483291626,
+      "learning_rate": 6.293554903828302e-06,
+      "loss": 0.0713,
+      "step": 625
+    },
+    {
+      "epoch": 2.1575342465753424,
+      "grad_norm": 0.9480213522911072,
+      "learning_rate": 6.0645104326135e-06,
+      "loss": 0.0763,
+      "step": 630
+    },
+    {
+      "epoch": 2.1746575342465753,
+      "grad_norm": 1.314867377281189,
+      "learning_rate": 5.8386508815842746e-06,
+      "loss": 0.0797,
+      "step": 635
+    },
+    {
+      "epoch": 2.191780821917808,
+      "grad_norm": 0.9061669111251831,
+      "learning_rate": 5.61605675496345e-06,
+      "loss": 0.0701,
+      "step": 640
+    },
+    {
+      "epoch": 2.208904109589041,
+      "grad_norm": 0.8163189888000488,
+      "learning_rate": 5.396807393062681e-06,
+      "loss": 0.072,
+      "step": 645
+    },
+    {
+      "epoch": 2.2260273972602738,
+      "grad_norm": 0.8815593719482422,
+      "learning_rate": 5.180980944002799e-06,
+      "loss": 0.0845,
+      "step": 650
+    },
+    {
+      "epoch": 2.243150684931507,
+      "grad_norm": 1.3875430822372437,
+      "learning_rate": 4.9686543358590934e-06,
+      "loss": 0.071,
+      "step": 655
+    },
+    {
+      "epoch": 2.26027397260274,
+      "grad_norm": 0.9008333086967468,
+      "learning_rate": 4.759903249241464e-06,
+      "loss": 0.0732,
+      "step": 660
+    },
+    {
+      "epoch": 2.2773972602739727,
+      "grad_norm": 1.1322684288024902,
+      "learning_rate": 4.554802090319209e-06,
+      "loss": 0.0744,
+      "step": 665
+    },
+    {
+      "epoch": 2.2945205479452055,
+      "grad_norm": 0.8206696510314941,
+      "learning_rate": 4.353423964300074e-06,
+      "loss": 0.0733,
+      "step": 670
+    },
+    {
+      "epoch": 2.3116438356164384,
+      "grad_norm": 0.8082016110420227,
+      "learning_rate": 4.155840649373015e-06,
+      "loss": 0.0821,
+      "step": 675
+    },
+    {
+      "epoch": 2.328767123287671,
+      "grad_norm": 0.6733844876289368,
+      "learning_rate": 3.96212257112391e-06,
+      "loss": 0.0604,
+      "step": 680
+    },
+    {
+      "epoch": 2.345890410958904,
+      "grad_norm": 0.8901552557945251,
+      "learning_rate": 3.772338777433482e-06,
+      "loss": 0.0711,
+      "step": 685
+    },
+    {
+      "epoch": 2.363013698630137,
+      "grad_norm": 0.9542776346206665,
+      "learning_rate": 3.5865569138661814e-06,
+      "loss": 0.0666,
+      "step": 690
+    },
+    {
+      "epoch": 2.3801369863013697,
+      "grad_norm": 0.6933631896972656,
+      "learning_rate": 3.4048431995589453e-06,
+      "loss": 0.0692,
+      "step": 695
+    },
+    {
+      "epoch": 2.3972602739726026,
+      "grad_norm": 0.7117693424224854,
+      "learning_rate": 3.22726240361843e-06,
+      "loss": 0.0674,
+      "step": 700
+    },
+    {
+      "epoch": 2.4143835616438354,
+      "grad_norm": 0.8828039169311523,
+      "learning_rate": 3.053877822034995e-06,
+      "loss": 0.0603,
+      "step": 705
+    },
+    {
+      "epoch": 2.4315068493150687,
+      "grad_norm": 0.8636316657066345,
+      "learning_rate": 2.884751255121827e-06,
+      "loss": 0.0641,
+      "step": 710
+    },
+    {
+      "epoch": 2.4486301369863015,
+      "grad_norm": 0.7074395418167114,
+      "learning_rate": 2.7199429854871544e-06,
+      "loss": 0.0594,
+      "step": 715
+    },
+    {
+      "epoch": 2.4657534246575343,
+      "grad_norm": 0.7124409675598145,
+      "learning_rate": 2.559511756547407e-06,
+      "loss": 0.0657,
+      "step": 720
+    },
+    {
+      "epoch": 2.482876712328767,
+      "grad_norm": 0.957909345626831,
+      "learning_rate": 2.403514751589032e-06,
+      "loss": 0.0673,
+      "step": 725
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8058770298957825,
+      "learning_rate": 2.252007573386365e-06,
+      "loss": 0.0753,
+      "step": 730
+    },
+    {
+      "epoch": 2.517123287671233,
+      "grad_norm": 0.6695918440818787,
+      "learning_rate": 2.105044224382854e-06,
+      "loss": 0.0597,
+      "step": 735
+    },
+    {
+      "epoch": 2.5342465753424657,
+      "grad_norm": 0.7419140934944153,
+      "learning_rate": 1.9626770874427368e-06,
+      "loss": 0.0639,
+      "step": 740
+    },
+    {
+      "epoch": 2.5513698630136985,
+      "grad_norm": 0.7338897585868835,
+      "learning_rate": 1.8249569071799134e-06,
+      "loss": 0.0681,
+      "step": 745
+    },
+    {
+      "epoch": 2.5684931506849313,
+      "grad_norm": 0.7281093597412109,
+      "learning_rate": 1.69193277187083e-06,
+      "loss": 0.0526,
+      "step": 750
+    },
+    {
+      "epoch": 2.5856164383561646,
+      "grad_norm": 0.7843554615974426,
+      "learning_rate": 1.5636520959577094e-06,
+      "loss": 0.0701,
+      "step": 755
+    },
+    {
+      "epoch": 2.602739726027397,
+      "grad_norm": 0.7026539444923401,
+      "learning_rate": 1.44016060314835e-06,
+      "loss": 0.0507,
+      "step": 760
+    },
+    {
+      "epoch": 2.6198630136986303,
+      "grad_norm": 0.5951750874519348,
+      "learning_rate": 1.321502310118649e-06,
+      "loss": 0.052,
+      "step": 765
+    },
+    {
+      "epoch": 2.636986301369863,
+      "grad_norm": 0.7210755944252014,
+      "learning_rate": 1.2077195108234934e-06,
+      "loss": 0.0635,
+      "step": 770
+    },
+    {
+      "epoch": 2.654109589041096,
+      "grad_norm": 0.7983770370483398,
+      "learning_rate": 1.098852761421719e-06,
+      "loss": 0.0603,
+      "step": 775
+    },
+    {
+      "epoch": 2.671232876712329,
+      "grad_norm": 0.6377550363540649,
+      "learning_rate": 9.949408658205072e-07,
+      "loss": 0.0586,
+      "step": 780
+    },
+    {
+      "epoch": 2.6883561643835616,
+      "grad_norm": 0.6245830059051514,
+      "learning_rate": 8.960208618442883e-07,
+      "loss": 0.0496,
+      "step": 785
+    },
+    {
+      "epoch": 2.7054794520547945,
+      "grad_norm": 0.8159735798835754,
+      "learning_rate": 8.021280080331816e-07,
+      "loss": 0.0689,
+      "step": 790
+    },
+    {
+      "epoch": 2.7226027397260273,
+      "grad_norm": 0.6015287637710571,
+      "learning_rate": 7.132957710756277e-07,
+      "loss": 0.0634,
+      "step": 795
+    },
+    {
+      "epoch": 2.73972602739726,
+      "grad_norm": 0.5720701813697815,
+      "learning_rate": 6.295558138796803e-07,
+      "loss": 0.046,
+      "step": 800
+    },
+    {
+      "epoch": 2.756849315068493,
+      "grad_norm": 0.6111004948616028,
+      "learning_rate": 5.509379842872558e-07,
+      "loss": 0.058,
+      "step": 805
+    },
+    {
+      "epoch": 2.7739726027397262,
+      "grad_norm": 0.6543359160423279,
+      "learning_rate": 4.774703044353051e-07,
+      "loss": 0.049,
+      "step": 810
+    },
+    {
+      "epoch": 2.791095890410959,
+      "grad_norm": 0.6640830039978027,
+      "learning_rate": 4.091789607677582e-07,
+      "loss": 0.055,
+      "step": 815
+    },
+    {
+      "epoch": 2.808219178082192,
+      "grad_norm": 0.7844074368476868,
+      "learning_rate": 3.460882947017635e-07,
+      "loss": 0.0577,
+      "step": 820
+    },
+    {
+      "epoch": 2.8253424657534247,
+      "grad_norm": 0.700030505657196,
+      "learning_rate": 2.8822079395154357e-07,
+      "loss": 0.051,
+      "step": 825
+    },
+    {
+      "epoch": 2.8424657534246576,
+      "grad_norm": 0.5808168053627014,
+      "learning_rate": 2.3559708451300622e-07,
+      "loss": 0.0525,
+      "step": 830
+    },
+    {
+      "epoch": 2.8595890410958904,
+      "grad_norm": 0.5799618363380432,
+      "learning_rate": 1.8823592331191242e-07,
+      "loss": 0.0613,
+      "step": 835
+    },
+    {
+      "epoch": 2.8767123287671232,
+      "grad_norm": 0.5387691259384155,
+      "learning_rate": 1.4615419151824406e-07,
+      "loss": 0.0502,
+      "step": 840
+    },
+    {
+      "epoch": 2.893835616438356,
+      "grad_norm": 0.7428579926490784,
+      "learning_rate": 1.0936688852919042e-07,
+      "loss": 0.0618,
+      "step": 845
+    },
+    {
+      "epoch": 2.910958904109589,
+      "grad_norm": 0.8076790571212769,
+      "learning_rate": 7.788712662281317e-08,
+      "loss": 0.0614,
+      "step": 850
+    },
+    {
+      "epoch": 2.928082191780822,
+      "grad_norm": 0.8102460503578186,
+      "learning_rate": 5.1726126284389886e-08,
+      "loss": 0.0615,
+      "step": 855
+    },
+    {
+      "epoch": 2.9452054794520546,
+      "grad_norm": 0.5508604049682617,
+      "learning_rate": 3.0893212207036556e-08,
+      "loss": 0.0614,
+      "step": 860
+    },
+    {
+      "epoch": 2.962328767123288,
+      "grad_norm": 0.6640682220458984,
+      "learning_rate": 1.5395809968061226e-08,
+      "loss": 0.0608,
+      "step": 865
+    },
+    {
+      "epoch": 2.9794520547945207,
+      "grad_norm": 0.7033427953720093,
+      "learning_rate": 5.239443382229481e-09,
+      "loss": 0.056,
+      "step": 870
+    },
+    {
+      "epoch": 2.9965753424657535,
+      "grad_norm": 0.6203268766403198,
+      "learning_rate": 4.277325328860826e-10,
+      "loss": 0.0644,
+      "step": 875
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 876,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.43138759768644e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

base/27_128_e3_3e-5/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08855170a206759d03b3719ad076ff3e524891a7141f179501090097be8ec04a
+size 8273

base/27_128_e3_3e-5/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)