Upload folder using huggingface_hub

klora/2/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

klora/2/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",
+    "up_proj",
+    "o_proj",
+    "gate_proj",
+    "q_proj",
+    "down_proj",
+    "k_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

klora/2/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b919f00f81686ec666bebc3f0e4ae984fd0e6c046ea4030d0a770bdfc8171672
+size 791751704

klora/2/latest

@@ -0,0 +1 @@
+global_step820

klora/2/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47321622bc7faee6badd51b5dc05ed4e5c7763bd448b702ceed7397470771ae2
+size 15920

klora/2/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:369266aeb2e7f5d05788859de911eb5bd4d7098edb4f96c62f081634e46566a0
+size 15920

klora/2/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98dc83c1f0e403edba41382ae2c3b4d8d793da2845f79fe3cfba08235b9df86a
+size 15920

klora/2/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22c3172692b7158b91bffe55abfc0d9dd28d35799afe36b9d624f3b2d8a845ed
+size 15920

klora/2/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfd8cc2be6961cca00e95302145070c4822302f966f46d1a057d76b42209cc46
+size 15920

klora/2/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:254ed4cbe00387a422e1d45a6ff53bfee83ad65f2c36a62b9befb889e24af025
+size 15920

klora/2/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:631f9e140fce08c47f763906215c80072c0e336ac54ebd9e3988d00f71c7575a
+size 15920

klora/2/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9c8df773da2116cd545fab5b3316dbdd4963547be8ec8cfb3cf4e53781851db
+size 15920

klora/2/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef7f2c8815012e644c0672748bc2fd3f3c879b12b709a596a65c97a6e53c0b91
+size 1064

klora/2/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
+  }
+}

klora/2/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
+}

klora/2/trainer_state.json

@@ -0,0 +1,1182 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 820,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.012195121951219513,
+      "grad_norm": 1.184765100479126,
+      "learning_rate": 1.1650485436893204e-06,
+      "loss": 1.2972,
+      "step": 5
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 0.9755042195320129,
+      "learning_rate": 2.621359223300971e-06,
+      "loss": 1.3303,
+      "step": 10
+    },
+    {
+      "epoch": 0.036585365853658534,
+      "grad_norm": 0.8106493949890137,
+      "learning_rate": 4.0776699029126215e-06,
+      "loss": 1.2908,
+      "step": 15
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 0.6635167002677917,
+      "learning_rate": 5.533980582524272e-06,
+      "loss": 1.2413,
+      "step": 20
+    },
+    {
+      "epoch": 0.06097560975609756,
+      "grad_norm": 0.5763493180274963,
+      "learning_rate": 6.990291262135923e-06,
+      "loss": 1.2209,
+      "step": 25
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 0.49019569158554077,
+      "learning_rate": 8.446601941747573e-06,
+      "loss": 1.265,
+      "step": 30
+    },
+    {
+      "epoch": 0.08536585365853659,
+      "grad_norm": 0.5669918060302734,
+      "learning_rate": 9.902912621359224e-06,
+      "loss": 1.2391,
+      "step": 35
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 0.5410548448562622,
+      "learning_rate": 1.1359223300970873e-05,
+      "loss": 1.2694,
+      "step": 40
+    },
+    {
+      "epoch": 0.10975609756097561,
+      "grad_norm": 0.489359587430954,
+      "learning_rate": 1.2815533980582524e-05,
+      "loss": 1.2039,
+      "step": 45
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 0.6161925792694092,
+      "learning_rate": 1.4271844660194176e-05,
+      "loss": 1.1988,
+      "step": 50
+    },
+    {
+      "epoch": 0.13414634146341464,
+      "grad_norm": 0.5076792240142822,
+      "learning_rate": 1.5728155339805827e-05,
+      "loss": 1.128,
+      "step": 55
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 0.46671441197395325,
+      "learning_rate": 1.7184466019417476e-05,
+      "loss": 1.1529,
+      "step": 60
+    },
+    {
+      "epoch": 0.15853658536585366,
+      "grad_norm": 0.511127233505249,
+      "learning_rate": 1.8640776699029126e-05,
+      "loss": 1.1286,
+      "step": 65
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 0.5400530695915222,
+      "learning_rate": 2.009708737864078e-05,
+      "loss": 1.1555,
+      "step": 70
+    },
+    {
+      "epoch": 0.18292682926829268,
+      "grad_norm": 0.5800668597221375,
+      "learning_rate": 2.1553398058252428e-05,
+      "loss": 1.1579,
+      "step": 75
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 0.5733080506324768,
+      "learning_rate": 2.3009708737864078e-05,
+      "loss": 1.2164,
+      "step": 80
+    },
+    {
+      "epoch": 0.2073170731707317,
+      "grad_norm": 0.5242522954940796,
+      "learning_rate": 2.4466019417475727e-05,
+      "loss": 1.1548,
+      "step": 85
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 0.5192080140113831,
+      "learning_rate": 2.5922330097087377e-05,
+      "loss": 1.1127,
+      "step": 90
+    },
+    {
+      "epoch": 0.23170731707317074,
+      "grad_norm": 0.6592035889625549,
+      "learning_rate": 2.737864077669903e-05,
+      "loss": 1.1187,
+      "step": 95
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 0.5072050094604492,
+      "learning_rate": 2.8834951456310683e-05,
+      "loss": 1.1057,
+      "step": 100
+    },
+    {
+      "epoch": 0.25609756097560976,
+      "grad_norm": 0.530845582485199,
+      "learning_rate": 2.9999980473292916e-05,
+      "loss": 1.0943,
+      "step": 105
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 0.5525540113449097,
+      "learning_rate": 2.999929704388311e-05,
+      "loss": 1.0717,
+      "step": 110
+    },
+    {
+      "epoch": 0.2804878048780488,
+      "grad_norm": 0.5346278548240662,
+      "learning_rate": 2.999763732995757e-05,
+      "loss": 1.0477,
+      "step": 115
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 0.634670615196228,
+      "learning_rate": 2.999500143954488e-05,
+      "loss": 1.0227,
+      "step": 120
+    },
+    {
+      "epoch": 0.3048780487804878,
+      "grad_norm": 0.6343799829483032,
+      "learning_rate": 2.9991389544211688e-05,
+      "loss": 1.0359,
+      "step": 125
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 0.6552476286888123,
+      "learning_rate": 2.998680187905151e-05,
+      "loss": 1.0654,
+      "step": 130
+    },
+    {
+      "epoch": 0.32926829268292684,
+      "grad_norm": 0.7905762195587158,
+      "learning_rate": 2.9981238742669432e-05,
+      "loss": 1.0706,
+      "step": 135
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 0.6723063588142395,
+      "learning_rate": 2.997470049716269e-05,
+      "loss": 1.0488,
+      "step": 140
+    },
+    {
+      "epoch": 0.35365853658536583,
+      "grad_norm": 0.6034112572669983,
+      "learning_rate": 2.9967187568097084e-05,
+      "loss": 1.0492,
+      "step": 145
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 0.5780961513519287,
+      "learning_rate": 2.995870044447928e-05,
+      "loss": 0.9935,
+      "step": 150
+    },
+    {
+      "epoch": 0.3780487804878049,
+      "grad_norm": 0.6658617854118347,
+      "learning_rate": 2.9949239678724995e-05,
+      "loss": 1.046,
+      "step": 155
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 0.6619867086410522,
+      "learning_rate": 2.993880588662303e-05,
+      "loss": 0.9965,
+      "step": 160
+    },
+    {
+      "epoch": 0.4024390243902439,
+      "grad_norm": 0.7782027721405029,
+      "learning_rate": 2.9927399747295178e-05,
+      "loss": 0.9886,
+      "step": 165
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 0.6822337508201599,
+      "learning_rate": 2.9915022003152058e-05,
+      "loss": 0.9674,
+      "step": 170
+    },
+    {
+      "epoch": 0.4268292682926829,
+      "grad_norm": 0.5955122113227844,
+      "learning_rate": 2.990167345984475e-05,
+      "loss": 1.0358,
+      "step": 175
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 0.7010889649391174,
+      "learning_rate": 2.988735498621238e-05,
+      "loss": 0.9443,
+      "step": 180
+    },
+    {
+      "epoch": 0.45121951219512196,
+      "grad_norm": 0.651681661605835,
+      "learning_rate": 2.9872067514225565e-05,
+      "loss": 0.9144,
+      "step": 185
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 0.7048653364181519,
+      "learning_rate": 2.985581203892575e-05,
+      "loss": 0.9196,
+      "step": 190
+    },
+    {
+      "epoch": 0.47560975609756095,
+      "grad_norm": 0.7128203511238098,
+      "learning_rate": 2.9838589618360434e-05,
+      "loss": 0.9225,
+      "step": 195
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 0.7625283002853394,
+      "learning_rate": 2.9820401373514333e-05,
+      "loss": 0.9858,
+      "step": 200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7980352640151978,
+      "learning_rate": 2.9801248488236364e-05,
+      "loss": 0.9252,
+      "step": 205
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 0.6355564594268799,
+      "learning_rate": 2.978113220916265e-05,
+      "loss": 0.9262,
+      "step": 210
+    },
+    {
+      "epoch": 0.524390243902439,
+      "grad_norm": 0.7836682200431824,
+      "learning_rate": 2.9760053845635324e-05,
+      "loss": 0.8971,
+      "step": 215
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 0.7409992218017578,
+      "learning_rate": 2.973801476961736e-05,
+      "loss": 0.8825,
+      "step": 220
+    },
+    {
+      "epoch": 0.5487804878048781,
+      "grad_norm": 0.6563221216201782,
+      "learning_rate": 2.9715016415603217e-05,
+      "loss": 0.8866,
+      "step": 225
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 0.6461041569709778,
+      "learning_rate": 2.9691060280525506e-05,
+      "loss": 0.8716,
+      "step": 230
+    },
+    {
+      "epoch": 0.573170731707317,
+      "grad_norm": 0.6985337734222412,
+      "learning_rate": 2.9666147923657552e-05,
+      "loss": 0.8699,
+      "step": 235
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 0.8816692233085632,
+      "learning_rate": 2.9640280966511898e-05,
+      "loss": 0.8506,
+      "step": 240
+    },
+    {
+      "epoch": 0.5975609756097561,
+      "grad_norm": 0.7365345358848572,
+      "learning_rate": 2.961346109273476e-05,
+      "loss": 0.9093,
+      "step": 245
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 0.8134493827819824,
+      "learning_rate": 2.9585690047996444e-05,
+      "loss": 0.7961,
+      "step": 250
+    },
+    {
+      "epoch": 0.6219512195121951,
+      "grad_norm": 0.7763360738754272,
+      "learning_rate": 2.955696963987773e-05,
+      "loss": 0.8697,
+      "step": 255
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 0.7731444239616394,
+      "learning_rate": 2.9527301737752196e-05,
+      "loss": 0.8777,
+      "step": 260
+    },
+    {
+      "epoch": 0.6463414634146342,
+      "grad_norm": 0.9481649398803711,
+      "learning_rate": 2.9496688272664577e-05,
+      "loss": 0.8514,
+      "step": 265
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 0.827903151512146,
+      "learning_rate": 2.9465131237205043e-05,
+      "loss": 0.908,
+      "step": 270
+    },
+    {
+      "epoch": 0.6707317073170732,
+      "grad_norm": 0.8994118571281433,
+      "learning_rate": 2.943263268537953e-05,
+      "loss": 0.7373,
+      "step": 275
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 0.8990964889526367,
+      "learning_rate": 2.9399194732476032e-05,
+      "loss": 0.82,
+      "step": 280
+    },
+    {
+      "epoch": 0.6951219512195121,
+      "grad_norm": 0.8939409255981445,
+      "learning_rate": 2.9364819554926912e-05,
+      "loss": 0.8102,
+      "step": 285
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 0.9013278484344482,
+      "learning_rate": 2.932950939016727e-05,
+      "loss": 0.7655,
+      "step": 290
+    },
+    {
+      "epoch": 0.7195121951219512,
+      "grad_norm": 0.7653906941413879,
+      "learning_rate": 2.929326653648929e-05,
+      "loss": 0.7864,
+      "step": 295
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 0.9510161876678467,
+      "learning_rate": 2.9256093352892635e-05,
+      "loss": 0.7845,
+      "step": 300
+    },
+    {
+      "epoch": 0.7439024390243902,
+      "grad_norm": 0.9910411834716797,
+      "learning_rate": 2.921799225893094e-05,
+      "loss": 0.8002,
+      "step": 305
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 0.8927392959594727,
+      "learning_rate": 2.917896573455429e-05,
+      "loss": 0.7693,
+      "step": 310
+    },
+    {
+      "epoch": 0.7682926829268293,
+      "grad_norm": 0.8335655927658081,
+      "learning_rate": 2.913901631994783e-05,
+      "loss": 0.7593,
+      "step": 315
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 0.9775434136390686,
+      "learning_rate": 2.909814661536641e-05,
+      "loss": 0.7371,
+      "step": 320
+    },
+    {
+      "epoch": 0.7926829268292683,
+      "grad_norm": 0.8624247908592224,
+      "learning_rate": 2.9056359280965345e-05,
+      "loss": 0.8004,
+      "step": 325
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 1.1061346530914307,
+      "learning_rate": 2.901365703662726e-05,
+      "loss": 0.7445,
+      "step": 330
+    },
+    {
+      "epoch": 0.8170731707317073,
+      "grad_norm": 0.8500701189041138,
+      "learning_rate": 2.897004266178508e-05,
+      "loss": 0.7437,
+      "step": 335
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 0.9865604639053345,
+      "learning_rate": 2.892551899524109e-05,
+      "loss": 0.701,
+      "step": 340
+    },
+    {
+      "epoch": 0.8414634146341463,
+      "grad_norm": 0.9660606384277344,
+      "learning_rate": 2.888008893498219e-05,
+      "loss": 0.7524,
+      "step": 345
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 0.9367020726203918,
+      "learning_rate": 2.8833755437991242e-05,
+      "loss": 0.7132,
+      "step": 350
+    },
+    {
+      "epoch": 0.8658536585365854,
+      "grad_norm": 1.0071465969085693,
+      "learning_rate": 2.878652152005462e-05,
+      "loss": 0.7564,
+      "step": 355
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 1.0385185480117798,
+      "learning_rate": 2.8738390255565925e-05,
+      "loss": 0.6634,
+      "step": 360
+    },
+    {
+      "epoch": 0.8902439024390244,
+      "grad_norm": 0.9005842208862305,
+      "learning_rate": 2.8689364777325847e-05,
+      "loss": 0.6951,
+      "step": 365
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 0.9195963144302368,
+      "learning_rate": 2.863944827633828e-05,
+      "loss": 0.7013,
+      "step": 370
+    },
+    {
+      "epoch": 0.9146341463414634,
+      "grad_norm": 0.9449793100357056,
+      "learning_rate": 2.8588644001602623e-05,
+      "loss": 0.6901,
+      "step": 375
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 0.9308488368988037,
+      "learning_rate": 2.8536955259902283e-05,
+      "loss": 0.7064,
+      "step": 380
+    },
+    {
+      "epoch": 0.9390243902439024,
+      "grad_norm": 0.8601620197296143,
+      "learning_rate": 2.8484385415589485e-05,
+      "loss": 0.6913,
+      "step": 385
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 0.8974642157554626,
+      "learning_rate": 2.843093789036624e-05,
+      "loss": 0.6692,
+      "step": 390
+    },
+    {
+      "epoch": 0.9634146341463414,
+      "grad_norm": 0.890214741230011,
+      "learning_rate": 2.837661616306166e-05,
+      "loss": 0.6839,
+      "step": 395
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 0.8833136558532715,
+      "learning_rate": 2.8321423769405538e-05,
+      "loss": 0.742,
+      "step": 400
+    },
+    {
+      "epoch": 0.9878048780487805,
+      "grad_norm": 1.1129741668701172,
+      "learning_rate": 2.8265364301798175e-05,
+      "loss": 0.6696,
+      "step": 405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.957071840763092,
+      "learning_rate": 2.8208441409076576e-05,
+      "loss": 0.6266,
+      "step": 410
+    },
+    {
+      "epoch": 1.0121951219512195,
+      "grad_norm": 1.0043803453445435,
+      "learning_rate": 2.8150658796276964e-05,
+      "loss": 0.6141,
+      "step": 415
+    },
+    {
+      "epoch": 1.024390243902439,
+      "grad_norm": 1.0592577457427979,
+      "learning_rate": 2.8092020224393603e-05,
+      "loss": 0.5861,
+      "step": 420
+    },
+    {
+      "epoch": 1.0365853658536586,
+      "grad_norm": 1.1829588413238525,
+      "learning_rate": 2.8032529510134008e-05,
+      "loss": 0.5592,
+      "step": 425
+    },
+    {
+      "epoch": 1.048780487804878,
+      "grad_norm": 1.0533852577209473,
+      "learning_rate": 2.797219052567052e-05,
+      "loss": 0.6068,
+      "step": 430
+    },
+    {
+      "epoch": 1.0609756097560976,
+      "grad_norm": 0.9526946544647217,
+      "learning_rate": 2.791100719838827e-05,
+      "loss": 0.5667,
+      "step": 435
+    },
+    {
+      "epoch": 1.0731707317073171,
+      "grad_norm": 1.1012799739837646,
+      "learning_rate": 2.7848983510629554e-05,
+      "loss": 0.6413,
+      "step": 440
+    },
+    {
+      "epoch": 1.0853658536585367,
+      "grad_norm": 1.601194143295288,
+      "learning_rate": 2.7786123499434642e-05,
+      "loss": 0.5612,
+      "step": 445
+    },
+    {
+      "epoch": 1.0975609756097562,
+      "grad_norm": 0.9721382856369019,
+      "learning_rate": 2.772243125627896e-05,
+      "loss": 0.5407,
+      "step": 450
+    },
+    {
+      "epoch": 1.1097560975609757,
+      "grad_norm": 1.0757757425308228,
+      "learning_rate": 2.7657910926806858e-05,
+      "loss": 0.505,
+      "step": 455
+    },
+    {
+      "epoch": 1.1219512195121952,
+      "grad_norm": 1.2296000719070435,
+      "learning_rate": 2.7592566710561704e-05,
+      "loss": 0.559,
+      "step": 460
+    },
+    {
+      "epoch": 1.1341463414634148,
+      "grad_norm": 0.9386566877365112,
+      "learning_rate": 2.7526402860712597e-05,
+      "loss": 0.5553,
+      "step": 465
+    },
+    {
+      "epoch": 1.146341463414634,
+      "grad_norm": 1.1901946067810059,
+      "learning_rate": 2.7459423683777484e-05,
+      "loss": 0.5844,
+      "step": 470
+    },
+    {
+      "epoch": 1.1585365853658536,
+      "grad_norm": 1.1464612483978271,
+      "learning_rate": 2.73916335393429e-05,
+      "loss": 0.5492,
+      "step": 475
+    },
+    {
+      "epoch": 1.170731707317073,
+      "grad_norm": 1.005790114402771,
+      "learning_rate": 2.732303683978018e-05,
+      "loss": 0.553,
+      "step": 480
+    },
+    {
+      "epoch": 1.1829268292682926,
+      "grad_norm": 1.2415597438812256,
+      "learning_rate": 2.725363804995827e-05,
+      "loss": 0.5452,
+      "step": 485
+    },
+    {
+      "epoch": 1.1951219512195121,
+      "grad_norm": 1.0382238626480103,
+      "learning_rate": 2.7183441686953122e-05,
+      "loss": 0.5413,
+      "step": 490
+    },
+    {
+      "epoch": 1.2073170731707317,
+      "grad_norm": 1.0460742712020874,
+      "learning_rate": 2.711245231975367e-05,
+      "loss": 0.5109,
+      "step": 495
+    },
+    {
+      "epoch": 1.2195121951219512,
+      "grad_norm": 0.9935807585716248,
+      "learning_rate": 2.7040674568964454e-05,
+      "loss": 0.5572,
+      "step": 500
+    },
+    {
+      "epoch": 1.2317073170731707,
+      "grad_norm": 1.0190331935882568,
+      "learning_rate": 2.6968113106504865e-05,
+      "loss": 0.5075,
+      "step": 505
+    },
+    {
+      "epoch": 1.2439024390243902,
+      "grad_norm": 0.9684451818466187,
+      "learning_rate": 2.689477265530504e-05,
+      "loss": 0.5295,
+      "step": 510
+    },
+    {
+      "epoch": 1.2560975609756098,
+      "grad_norm": 1.0213184356689453,
+      "learning_rate": 2.68206579889985e-05,
+      "loss": 0.5119,
+      "step": 515
+    },
+    {
+      "epoch": 1.2682926829268293,
+      "grad_norm": 0.9365750551223755,
+      "learning_rate": 2.674577393161137e-05,
+      "loss": 0.5111,
+      "step": 520
+    },
+    {
+      "epoch": 1.2804878048780488,
+      "grad_norm": 1.0988898277282715,
+      "learning_rate": 2.667012535724847e-05,
+      "loss": 0.4914,
+      "step": 525
+    },
+    {
+      "epoch": 1.2926829268292683,
+      "grad_norm": 1.0466548204421997,
+      "learning_rate": 2.6593717189775986e-05,
+      "loss": 0.4429,
+      "step": 530
+    },
+    {
+      "epoch": 1.3048780487804879,
+      "grad_norm": 1.141400694847107,
+      "learning_rate": 2.6516554402501048e-05,
+      "loss": 0.4968,
+      "step": 535
+    },
+    {
+      "epoch": 1.3170731707317074,
+      "grad_norm": 0.9664748311042786,
+      "learning_rate": 2.6438642017847983e-05,
+      "loss": 0.4933,
+      "step": 540
+    },
+    {
+      "epoch": 1.329268292682927,
+      "grad_norm": 1.1237977743148804,
+      "learning_rate": 2.635998510703143e-05,
+      "loss": 0.4822,
+      "step": 545
+    },
+    {
+      "epoch": 1.3414634146341464,
+      "grad_norm": 1.1644905805587769,
+      "learning_rate": 2.6280588789726257e-05,
+      "loss": 0.5107,
+      "step": 550
+    },
+    {
+      "epoch": 1.3536585365853657,
+      "grad_norm": 1.103337287902832,
+      "learning_rate": 2.6200458233734316e-05,
+      "loss": 0.4696,
+      "step": 555
+    },
+    {
+      "epoch": 1.3658536585365852,
+      "grad_norm": 1.017683744430542,
+      "learning_rate": 2.611959865464811e-05,
+      "loss": 0.4719,
+      "step": 560
+    },
+    {
+      "epoch": 1.3780487804878048,
+      "grad_norm": 1.2212704420089722,
+      "learning_rate": 2.6038015315511284e-05,
+      "loss": 0.4932,
+      "step": 565
+    },
+    {
+      "epoch": 1.3902439024390243,
+      "grad_norm": 1.2683626413345337,
+      "learning_rate": 2.5955713526476068e-05,
+      "loss": 0.4822,
+      "step": 570
+    },
+    {
+      "epoch": 1.4024390243902438,
+      "grad_norm": 0.9710960984230042,
+      "learning_rate": 2.587269864445766e-05,
+      "loss": 0.4526,
+      "step": 575
+    },
+    {
+      "epoch": 1.4146341463414633,
+      "grad_norm": 0.9223999977111816,
+      "learning_rate": 2.578897607278555e-05,
+      "loss": 0.473,
+      "step": 580
+    },
+    {
+      "epoch": 1.4268292682926829,
+      "grad_norm": 1.176310658454895,
+      "learning_rate": 2.5704551260851786e-05,
+      "loss": 0.4636,
+      "step": 585
+    },
+    {
+      "epoch": 1.4390243902439024,
+      "grad_norm": 1.1532803773880005,
+      "learning_rate": 2.5619429703756355e-05,
+      "loss": 0.4549,
+      "step": 590
+    },
+    {
+      "epoch": 1.451219512195122,
+      "grad_norm": 0.9891874194145203,
+      "learning_rate": 2.5533616941949426e-05,
+      "loss": 0.4909,
+      "step": 595
+    },
+    {
+      "epoch": 1.4634146341463414,
+      "grad_norm": 1.1279683113098145,
+      "learning_rate": 2.5447118560870803e-05,
+      "loss": 0.4825,
+      "step": 600
+    },
+    {
+      "epoch": 1.475609756097561,
+      "grad_norm": 0.9699839353561401,
+      "learning_rate": 2.5359940190586337e-05,
+      "loss": 0.4478,
+      "step": 605
+    },
+    {
+      "epoch": 1.4878048780487805,
+      "grad_norm": 0.970900297164917,
+      "learning_rate": 2.5272087505421482e-05,
+      "loss": 0.5015,
+      "step": 610
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0708060264587402,
+      "learning_rate": 2.5183566223591966e-05,
+      "loss": 0.4883,
+      "step": 615
+    },
+    {
+      "epoch": 1.5121951219512195,
+      "grad_norm": 1.0761215686798096,
+      "learning_rate": 2.509438210683158e-05,
+      "loss": 0.4213,
+      "step": 620
+    },
+    {
+      "epoch": 1.524390243902439,
+      "grad_norm": 1.0248342752456665,
+      "learning_rate": 2.5004540960017185e-05,
+      "loss": 0.4311,
+      "step": 625
+    },
+    {
+      "epoch": 1.5365853658536586,
+      "grad_norm": 1.1032685041427612,
+      "learning_rate": 2.4914048630790857e-05,
+      "loss": 0.4608,
+      "step": 630
+    },
+    {
+      "epoch": 1.548780487804878,
+      "grad_norm": 1.1839253902435303,
+      "learning_rate": 2.482291100917928e-05,
+      "loss": 0.4604,
+      "step": 635
+    },
+    {
+      "epoch": 1.5609756097560976,
+      "grad_norm": 1.2353657484054565,
+      "learning_rate": 2.473113402721036e-05,
+      "loss": 0.4499,
+      "step": 640
+    },
+    {
+      "epoch": 1.5731707317073171,
+      "grad_norm": 1.177707552909851,
+      "learning_rate": 2.4638723658527152e-05,
+      "loss": 0.4172,
+      "step": 645
+    },
+    {
+      "epoch": 1.5853658536585367,
+      "grad_norm": 1.0010558366775513,
+      "learning_rate": 2.4545685917998994e-05,
+      "loss": 0.4417,
+      "step": 650
+    },
+    {
+      "epoch": 1.5975609756097562,
+      "grad_norm": 1.1996454000473022,
+      "learning_rate": 2.4452026861330054e-05,
+      "loss": 0.4586,
+      "step": 655
+    },
+    {
+      "epoch": 1.6097560975609757,
+      "grad_norm": 1.1821777820587158,
+      "learning_rate": 2.4357752584665123e-05,
+      "loss": 0.4828,
+      "step": 660
+    },
+    {
+      "epoch": 1.6219512195121952,
+      "grad_norm": 1.1734565496444702,
+      "learning_rate": 2.426286922419288e-05,
+      "loss": 0.4167,
+      "step": 665
+    },
+    {
+      "epoch": 1.6341463414634148,
+      "grad_norm": 0.9794202446937561,
+      "learning_rate": 2.416738295574645e-05,
+      "loss": 0.3879,
+      "step": 670
+    },
+    {
+      "epoch": 1.6463414634146343,
+      "grad_norm": 1.191879391670227,
+      "learning_rate": 2.4071299994401462e-05,
+      "loss": 0.4004,
+      "step": 675
+    },
+    {
+      "epoch": 1.6585365853658538,
+      "grad_norm": 1.0724924802780151,
+      "learning_rate": 2.397462659407148e-05,
+      "loss": 0.4385,
+      "step": 680
+    },
+    {
+      "epoch": 1.6707317073170733,
+      "grad_norm": 1.1953388452529907,
+      "learning_rate": 2.387736904710098e-05,
+      "loss": 0.4674,
+      "step": 685
+    },
+    {
+      "epoch": 1.6829268292682928,
+      "grad_norm": 1.089429497718811,
+      "learning_rate": 2.377953368385577e-05,
+      "loss": 0.4316,
+      "step": 690
+    },
+    {
+      "epoch": 1.6951219512195121,
+      "grad_norm": 1.1765036582946777,
+      "learning_rate": 2.368112687231095e-05,
+      "loss": 0.4237,
+      "step": 695
+    },
+    {
+      "epoch": 1.7073170731707317,
+      "grad_norm": 1.0054248571395874,
+      "learning_rate": 2.3582155017636455e-05,
+      "loss": 0.4639,
+      "step": 700
+    },
+    {
+      "epoch": 1.7195121951219512,
+      "grad_norm": 1.0205795764923096,
+      "learning_rate": 2.3482624561780128e-05,
+      "loss": 0.3966,
+      "step": 705
+    },
+    {
+      "epoch": 1.7317073170731707,
+      "grad_norm": 1.1560744047164917,
+      "learning_rate": 2.3382541983048425e-05,
+      "loss": 0.3839,
+      "step": 710
+    },
+    {
+      "epoch": 1.7439024390243902,
+      "grad_norm": 1.079033613204956,
+      "learning_rate": 2.3281913795684757e-05,
+      "loss": 0.3966,
+      "step": 715
+    },
+    {
+      "epoch": 1.7560975609756098,
+      "grad_norm": 1.0057902336120605,
+      "learning_rate": 2.318074654944547e-05,
+      "loss": 0.4309,
+      "step": 720
+    },
+    {
+      "epoch": 1.7682926829268293,
+      "grad_norm": 1.2559707164764404,
+      "learning_rate": 2.307904682917357e-05,
+      "loss": 0.4111,
+      "step": 725
+    },
+    {
+      "epoch": 1.7804878048780488,
+      "grad_norm": 1.2369558811187744,
+      "learning_rate": 2.2976821254370077e-05,
+      "loss": 0.4051,
+      "step": 730
+    },
+    {
+      "epoch": 1.7926829268292683,
+      "grad_norm": 0.968970000743866,
+      "learning_rate": 2.2874076478763187e-05,
+      "loss": 0.3857,
+      "step": 735
+    },
+    {
+      "epoch": 1.8048780487804879,
+      "grad_norm": 1.0662639141082764,
+      "learning_rate": 2.277081918987522e-05,
+      "loss": 0.3677,
+      "step": 740
+    },
+    {
+      "epoch": 1.8170731707317072,
+      "grad_norm": 1.0119315385818481,
+      "learning_rate": 2.266705610858727e-05,
+      "loss": 0.3887,
+      "step": 745
+    },
+    {
+      "epoch": 1.8292682926829267,
+      "grad_norm": 1.0961999893188477,
+      "learning_rate": 2.2562793988701838e-05,
+      "loss": 0.4059,
+      "step": 750
+    },
+    {
+      "epoch": 1.8414634146341462,
+      "grad_norm": 1.0293822288513184,
+      "learning_rate": 2.245803961650316e-05,
+      "loss": 0.3473,
+      "step": 755
+    },
+    {
+      "epoch": 1.8536585365853657,
+      "grad_norm": 1.076990008354187,
+      "learning_rate": 2.2352799810315534e-05,
+      "loss": 0.3493,
+      "step": 760
+    },
+    {
+      "epoch": 1.8658536585365852,
+      "grad_norm": 1.1946849822998047,
+      "learning_rate": 2.2247081420059526e-05,
+      "loss": 0.3746,
+      "step": 765
+    },
+    {
+      "epoch": 1.8780487804878048,
+      "grad_norm": 1.3098491430282593,
+      "learning_rate": 2.2140891326806106e-05,
+      "loss": 0.3596,
+      "step": 770
+    },
+    {
+      "epoch": 1.8902439024390243,
+      "grad_norm": 1.1211118698120117,
+      "learning_rate": 2.203423644232877e-05,
+      "loss": 0.3514,
+      "step": 775
+    },
+    {
+      "epoch": 1.9024390243902438,
+      "grad_norm": 1.046966314315796,
+      "learning_rate": 2.1927123708653658e-05,
+      "loss": 0.3954,
+      "step": 780
+    },
+    {
+      "epoch": 1.9146341463414633,
+      "grad_norm": 0.9153071045875549,
+      "learning_rate": 2.1819560097607717e-05,
+      "loss": 0.3866,
+      "step": 785
+    },
+    {
+      "epoch": 1.9268292682926829,
+      "grad_norm": 1.07267427444458,
+      "learning_rate": 2.1711552610364914e-05,
+      "loss": 0.39,
+      "step": 790
+    },
+    {
+      "epoch": 1.9390243902439024,
+      "grad_norm": 1.155359148979187,
+      "learning_rate": 2.1603108276990536e-05,
+      "loss": 0.3738,
+      "step": 795
+    },
+    {
+      "epoch": 1.951219512195122,
+      "grad_norm": 1.2188688516616821,
+      "learning_rate": 2.1494234155983595e-05,
+      "loss": 0.3394,
+      "step": 800
+    },
+    {
+      "epoch": 1.9634146341463414,
+      "grad_norm": 1.0880542993545532,
+      "learning_rate": 2.1384937333817413e-05,
+      "loss": 0.3525,
+      "step": 805
+    },
+    {
+      "epoch": 1.975609756097561,
+      "grad_norm": 1.1321754455566406,
+      "learning_rate": 2.1275224924478408e-05,
+      "loss": 0.324,
+      "step": 810
+    },
+    {
+      "epoch": 1.9878048780487805,
+      "grad_norm": 0.9325311779975891,
+      "learning_rate": 2.1165104069002983e-05,
+      "loss": 0.3555,
+      "step": 815
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9841716289520264,
+      "learning_rate": 2.1054581935012777e-05,
+      "loss": 0.3838,
+      "step": 820
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 2050,
+  "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": 1.1934379878624788e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

klora/2/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6306c1f7abc4c6d9bb36e9b1abe5fbdb88eb9a230852bea4593589b0967666a
+size 7736

klora/2/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)