Upload folder using huggingface_hub

13/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

13/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": [
+    "gate_proj",
+    "down_proj",
+    "v_proj",
+    "q_proj",
+    "o_proj",
+    "k_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

13/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0215e43a1cf134a77607460937b5d697ac1baede78950506891cceb32fc5a0a
+size 791751704

13/latest

@@ -0,0 +1 @@
+global_step620

13/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba1543d6669c759ca015d4496e66f78ed88f7b666bb66529626022e8533a37d5
+size 15920

13/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6ec36f42343dcfe20a575016052da7d80898996e4890fbcb473378f26eda1c5
+size 15920

13/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12b7b3a98a839c3570e3a04270d73e8b9dedc5531f35f9b4e2632aefc1f40d26
+size 15920

13/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0295bcd547c759acc3bad4a781d85c06c3210ee6156cf8678cc3949c565f363d
+size 15920

13/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:151e5911095bd8e5b97d57e6ff64fcba48682419aee8e9f2bde33ddc0b88b4ad
+size 15920

13/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a290ee5afd78b898c056c8eeec0c703066c0cdca54621d429e8244bc2e36271d
+size 15920

13/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d61b29bc14c2459e8787abbdc60da095ab3fa03ab8fddd2d3d6b042dbf25f80f
+size 15920

13/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d27d083cf211f3cf4b02cb09d20b308a414300efe44ba45f6c5d1bea547c58e2
+size 15920

13/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7709717aa7d3b3fcb114631ce9a46c3450879a8efd243c325e7f45d19fc4748
+size 1064

13/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
+  }
+}

13/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
+}

13/trainer_state.json

@@ -0,0 +1,902 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 620,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01615508885298869,
+      "grad_norm": 1.271867275238037,
+      "learning_rate": 1.5384615384615385e-06,
+      "loss": 1.3063,
+      "step": 5
+    },
+    {
+      "epoch": 0.03231017770597738,
+      "grad_norm": 0.926250696182251,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 1.3332,
+      "step": 10
+    },
+    {
+      "epoch": 0.048465266558966075,
+      "grad_norm": 0.7140916585922241,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.2903,
+      "step": 15
+    },
+    {
+      "epoch": 0.06462035541195477,
+      "grad_norm": 0.9422391653060913,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.2598,
+      "step": 20
+    },
+    {
+      "epoch": 0.08077544426494346,
+      "grad_norm": 0.5254983305931091,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.2614,
+      "step": 25
+    },
+    {
+      "epoch": 0.09693053311793215,
+      "grad_norm": 0.6005647778511047,
+      "learning_rate": 1.1153846153846154e-05,
+      "loss": 1.2564,
+      "step": 30
+    },
+    {
+      "epoch": 0.11308562197092084,
+      "grad_norm": 0.5752702355384827,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 1.2446,
+      "step": 35
+    },
+    {
+      "epoch": 0.12924071082390953,
+      "grad_norm": 0.4668022096157074,
+      "learning_rate": 1.5e-05,
+      "loss": 1.2149,
+      "step": 40
+    },
+    {
+      "epoch": 0.14539579967689822,
+      "grad_norm": 0.530459463596344,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 1.2028,
+      "step": 45
+    },
+    {
+      "epoch": 0.16155088852988692,
+      "grad_norm": 0.4735831916332245,
+      "learning_rate": 1.8846153846153846e-05,
+      "loss": 1.2197,
+      "step": 50
+    },
+    {
+      "epoch": 0.1777059773828756,
+      "grad_norm": 0.5452806949615479,
+      "learning_rate": 2.076923076923077e-05,
+      "loss": 1.1691,
+      "step": 55
+    },
+    {
+      "epoch": 0.1938610662358643,
+      "grad_norm": 0.4458446800708771,
+      "learning_rate": 2.269230769230769e-05,
+      "loss": 1.17,
+      "step": 60
+    },
+    {
+      "epoch": 0.210016155088853,
+      "grad_norm": 0.566159725189209,
+      "learning_rate": 2.4615384615384616e-05,
+      "loss": 1.1737,
+      "step": 65
+    },
+    {
+      "epoch": 0.22617124394184168,
+      "grad_norm": 0.4490140974521637,
+      "learning_rate": 2.6538461538461538e-05,
+      "loss": 1.1417,
+      "step": 70
+    },
+    {
+      "epoch": 0.24232633279483037,
+      "grad_norm": 0.45892786979675293,
+      "learning_rate": 2.846153846153846e-05,
+      "loss": 1.1651,
+      "step": 75
+    },
+    {
+      "epoch": 0.25848142164781907,
+      "grad_norm": 0.5318244099617004,
+      "learning_rate": 2.9999965837847747e-05,
+      "loss": 1.1468,
+      "step": 80
+    },
+    {
+      "epoch": 0.27463651050080773,
+      "grad_norm": 0.5207120776176453,
+      "learning_rate": 2.999877017885749e-05,
+      "loss": 1.1483,
+      "step": 85
+    },
+    {
+      "epoch": 0.29079159935379645,
+      "grad_norm": 0.5204576849937439,
+      "learning_rate": 2.9995866567858337e-05,
+      "loss": 1.0666,
+      "step": 90
+    },
+    {
+      "epoch": 0.3069466882067851,
+      "grad_norm": 0.5433797836303711,
+      "learning_rate": 2.9991255335492612e-05,
+      "loss": 1.0818,
+      "step": 95
+    },
+    {
+      "epoch": 0.32310177705977383,
+      "grad_norm": 0.5017735958099365,
+      "learning_rate": 2.9984937006854268e-05,
+      "loss": 1.089,
+      "step": 100
+    },
+    {
+      "epoch": 0.3392568659127625,
+      "grad_norm": 0.6111322045326233,
+      "learning_rate": 2.9976912301429102e-05,
+      "loss": 1.0468,
+      "step": 105
+    },
+    {
+      "epoch": 0.3554119547657512,
+      "grad_norm": 0.5563653707504272,
+      "learning_rate": 2.996718213301282e-05,
+      "loss": 1.0671,
+      "step": 110
+    },
+    {
+      "epoch": 0.3715670436187399,
+      "grad_norm": 0.7332136034965515,
+      "learning_rate": 2.9955747609606973e-05,
+      "loss": 1.0187,
+      "step": 115
+    },
+    {
+      "epoch": 0.3877221324717286,
+      "grad_norm": 0.5513479709625244,
+      "learning_rate": 2.9942610033292804e-05,
+      "loss": 1.0218,
+      "step": 120
+    },
+    {
+      "epoch": 0.40387722132471726,
+      "grad_norm": 0.5962786674499512,
+      "learning_rate": 2.9927770900082956e-05,
+      "loss": 1.0027,
+      "step": 125
+    },
+    {
+      "epoch": 0.420032310177706,
+      "grad_norm": 0.5897378325462341,
+      "learning_rate": 2.9911231899751127e-05,
+      "loss": 0.9925,
+      "step": 130
+    },
+    {
+      "epoch": 0.43618739903069464,
+      "grad_norm": 0.6131622791290283,
+      "learning_rate": 2.9892994915639648e-05,
+      "loss": 1.0063,
+      "step": 135
+    },
+    {
+      "epoch": 0.45234248788368336,
+      "grad_norm": 0.6106935143470764,
+      "learning_rate": 2.9873062024445035e-05,
+      "loss": 0.9924,
+      "step": 140
+    },
+    {
+      "epoch": 0.46849757673667203,
+      "grad_norm": 0.7261186838150024,
+      "learning_rate": 2.9851435495981487e-05,
+      "loss": 0.9812,
+      "step": 145
+    },
+    {
+      "epoch": 0.48465266558966075,
+      "grad_norm": 0.6918673515319824,
+      "learning_rate": 2.982811779292243e-05,
+      "loss": 0.9703,
+      "step": 150
+    },
+    {
+      "epoch": 0.5008077544426495,
+      "grad_norm": 0.6846053004264832,
+      "learning_rate": 2.980311157052007e-05,
+      "loss": 0.9455,
+      "step": 155
+    },
+    {
+      "epoch": 0.5169628432956381,
+      "grad_norm": 0.7054921984672546,
+      "learning_rate": 2.977641967630308e-05,
+      "loss": 0.9684,
+      "step": 160
+    },
+    {
+      "epoch": 0.5331179321486268,
+      "grad_norm": 0.8386259078979492,
+      "learning_rate": 2.974804514975226e-05,
+      "loss": 0.9483,
+      "step": 165
+    },
+    {
+      "epoch": 0.5492730210016155,
+      "grad_norm": 0.8338026404380798,
+      "learning_rate": 2.9717991221954516e-05,
+      "loss": 0.8942,
+      "step": 170
+    },
+    {
+      "epoch": 0.5654281098546042,
+      "grad_norm": 0.8031806349754333,
+      "learning_rate": 2.9686261315234845e-05,
+      "loss": 0.8801,
+      "step": 175
+    },
+    {
+      "epoch": 0.5815831987075929,
+      "grad_norm": 0.8551956415176392,
+      "learning_rate": 2.9652859042766687e-05,
+      "loss": 0.9189,
+      "step": 180
+    },
+    {
+      "epoch": 0.5977382875605816,
+      "grad_norm": 0.781044065952301,
+      "learning_rate": 2.961778820816045e-05,
+      "loss": 0.9013,
+      "step": 185
+    },
+    {
+      "epoch": 0.6138933764135702,
+      "grad_norm": 0.8323260545730591,
+      "learning_rate": 2.958105280503039e-05,
+      "loss": 0.8815,
+      "step": 190
+    },
+    {
+      "epoch": 0.630048465266559,
+      "grad_norm": 1.069571614265442,
+      "learning_rate": 2.954265701653985e-05,
+      "loss": 0.887,
+      "step": 195
+    },
+    {
+      "epoch": 0.6462035541195477,
+      "grad_norm": 0.8414632678031921,
+      "learning_rate": 2.9502605214924905e-05,
+      "loss": 0.867,
+      "step": 200
+    },
+    {
+      "epoch": 0.6623586429725363,
+      "grad_norm": 0.9376639127731323,
+      "learning_rate": 2.9460901960996484e-05,
+      "loss": 0.8239,
+      "step": 205
+    },
+    {
+      "epoch": 0.678513731825525,
+      "grad_norm": 0.82874596118927,
+      "learning_rate": 2.941755200362104e-05,
+      "loss": 0.8348,
+      "step": 210
+    },
+    {
+      "epoch": 0.6946688206785138,
+      "grad_norm": 0.8920372128486633,
+      "learning_rate": 2.9372560279179733e-05,
+      "loss": 0.8882,
+      "step": 215
+    },
+    {
+      "epoch": 0.7108239095315024,
+      "grad_norm": 0.8290151357650757,
+      "learning_rate": 2.932593191100637e-05,
+      "loss": 0.783,
+      "step": 220
+    },
+    {
+      "epoch": 0.7269789983844911,
+      "grad_norm": 0.9877433180809021,
+      "learning_rate": 2.9277672208803948e-05,
+      "loss": 0.8075,
+      "step": 225
+    },
+    {
+      "epoch": 0.7431340872374798,
+      "grad_norm": 0.9012677073478699,
+      "learning_rate": 2.922778666804006e-05,
+      "loss": 0.8248,
+      "step": 230
+    },
+    {
+      "epoch": 0.7592891760904685,
+      "grad_norm": 0.8435679078102112,
+      "learning_rate": 2.917628096932108e-05,
+      "loss": 0.7821,
+      "step": 235
+    },
+    {
+      "epoch": 0.7754442649434572,
+      "grad_norm": 0.9440993070602417,
+      "learning_rate": 2.912316097774531e-05,
+      "loss": 0.8035,
+      "step": 240
+    },
+    {
+      "epoch": 0.7915993537964459,
+      "grad_norm": 0.9287444353103638,
+      "learning_rate": 2.9068432742235126e-05,
+      "loss": 0.7929,
+      "step": 245
+    },
+    {
+      "epoch": 0.8077544426494345,
+      "grad_norm": 0.9541422724723816,
+      "learning_rate": 2.9012102494848125e-05,
+      "loss": 0.7871,
+      "step": 250
+    },
+    {
+      "epoch": 0.8239095315024233,
+      "grad_norm": 0.9758172631263733,
+      "learning_rate": 2.8954176650067496e-05,
+      "loss": 0.7192,
+      "step": 255
+    },
+    {
+      "epoch": 0.840064620355412,
+      "grad_norm": 0.8767186999320984,
+      "learning_rate": 2.8894661804071588e-05,
+      "loss": 0.7646,
+      "step": 260
+    },
+    {
+      "epoch": 0.8562197092084006,
+      "grad_norm": 1.0081284046173096,
+      "learning_rate": 2.8833564733982746e-05,
+      "loss": 0.7274,
+      "step": 265
+    },
+    {
+      "epoch": 0.8723747980613893,
+      "grad_norm": 1.123024582862854,
+      "learning_rate": 2.877089239709564e-05,
+      "loss": 0.7234,
+      "step": 270
+    },
+    {
+      "epoch": 0.8885298869143781,
+      "grad_norm": 1.1195224523544312,
+      "learning_rate": 2.8706651930084965e-05,
+      "loss": 0.721,
+      "step": 275
+    },
+    {
+      "epoch": 0.9046849757673667,
+      "grad_norm": 1.0532171726226807,
+      "learning_rate": 2.8640850648192795e-05,
+      "loss": 0.7322,
+      "step": 280
+    },
+    {
+      "epoch": 0.9208400646203554,
+      "grad_norm": 1.0443161725997925,
+      "learning_rate": 2.8573496044395574e-05,
+      "loss": 0.7272,
+      "step": 285
+    },
+    {
+      "epoch": 0.9369951534733441,
+      "grad_norm": 0.9977967739105225,
+      "learning_rate": 2.850459578855086e-05,
+      "loss": 0.7261,
+      "step": 290
+    },
+    {
+      "epoch": 0.9531502423263328,
+      "grad_norm": 0.9388869404792786,
+      "learning_rate": 2.843415772652395e-05,
+      "loss": 0.7133,
+      "step": 295
+    },
+    {
+      "epoch": 0.9693053311793215,
+      "grad_norm": 0.9346816539764404,
+      "learning_rate": 2.8362189879294434e-05,
+      "loss": 0.6597,
+      "step": 300
+    },
+    {
+      "epoch": 0.9854604200323102,
+      "grad_norm": 1.0388387441635132,
+      "learning_rate": 2.828870044204284e-05,
+      "loss": 0.6581,
+      "step": 305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5073713064193726,
+      "learning_rate": 2.8213697783217413e-05,
+      "loss": 0.6607,
+      "step": 310
+    },
+    {
+      "epoch": 1.0161550888529887,
+      "grad_norm": 1.178043007850647,
+      "learning_rate": 2.8137190443581175e-05,
+      "loss": 0.6305,
+      "step": 315
+    },
+    {
+      "epoch": 1.0323101777059773,
+      "grad_norm": 0.9829104542732239,
+      "learning_rate": 2.8059187135239383e-05,
+      "loss": 0.6241,
+      "step": 320
+    },
+    {
+      "epoch": 1.048465266558966,
+      "grad_norm": 1.065422534942627,
+      "learning_rate": 2.7979696740647433e-05,
+      "loss": 0.6547,
+      "step": 325
+    },
+    {
+      "epoch": 1.0646203554119547,
+      "grad_norm": 1.1355617046356201,
+      "learning_rate": 2.7898728311599414e-05,
+      "loss": 0.5667,
+      "step": 330
+    },
+    {
+      "epoch": 1.0807754442649435,
+      "grad_norm": 0.9137387275695801,
+      "learning_rate": 2.781629106819733e-05,
+      "loss": 0.5735,
+      "step": 335
+    },
+    {
+      "epoch": 1.0969305331179322,
+      "grad_norm": 1.1564706563949585,
+      "learning_rate": 2.7732394397801196e-05,
+      "loss": 0.5792,
+      "step": 340
+    },
+    {
+      "epoch": 1.1130856219709209,
+      "grad_norm": 1.0879970788955688,
+      "learning_rate": 2.764704785396007e-05,
+      "loss": 0.5671,
+      "step": 345
+    },
+    {
+      "epoch": 1.1292407108239095,
+      "grad_norm": 1.134347677230835,
+      "learning_rate": 2.7560261155324176e-05,
+      "loss": 0.5761,
+      "step": 350
+    },
+    {
+      "epoch": 1.1453957996768982,
+      "grad_norm": 1.189737319946289,
+      "learning_rate": 2.747204418453818e-05,
+      "loss": 0.5255,
+      "step": 355
+    },
+    {
+      "epoch": 1.1615508885298869,
+      "grad_norm": 1.0731488466262817,
+      "learning_rate": 2.7382406987115863e-05,
+      "loss": 0.5373,
+      "step": 360
+    },
+    {
+      "epoch": 1.1777059773828755,
+      "grad_norm": 1.045630693435669,
+      "learning_rate": 2.7291359770296196e-05,
+      "loss": 0.5756,
+      "step": 365
+    },
+    {
+      "epoch": 1.1938610662358644,
+      "grad_norm": 1.1830902099609375,
+      "learning_rate": 2.7198912901881e-05,
+      "loss": 0.4824,
+      "step": 370
+    },
+    {
+      "epoch": 1.210016155088853,
+      "grad_norm": 1.1311571598052979,
+      "learning_rate": 2.7105076909054357e-05,
+      "loss": 0.5156,
+      "step": 375
+    },
+    {
+      "epoch": 1.2261712439418417,
+      "grad_norm": 1.1190736293792725,
+      "learning_rate": 2.7009862477183837e-05,
+      "loss": 0.479,
+      "step": 380
+    },
+    {
+      "epoch": 1.2423263327948304,
+      "grad_norm": 1.1775109767913818,
+      "learning_rate": 2.6913280448603727e-05,
+      "loss": 0.5447,
+      "step": 385
+    },
+    {
+      "epoch": 1.258481421647819,
+      "grad_norm": 0.9826534986495972,
+      "learning_rate": 2.6815341821380384e-05,
+      "loss": 0.5142,
+      "step": 390
+    },
+    {
+      "epoch": 1.2746365105008077,
+      "grad_norm": 1.1323788166046143,
+      "learning_rate": 2.6716057748059867e-05,
+      "loss": 0.5244,
+      "step": 395
+    },
+    {
+      "epoch": 1.2907915993537964,
+      "grad_norm": 1.423813819885254,
+      "learning_rate": 2.6615439534397934e-05,
+      "loss": 0.5275,
+      "step": 400
+    },
+    {
+      "epoch": 1.306946688206785,
+      "grad_norm": 1.2067975997924805,
+      "learning_rate": 2.6513498638072664e-05,
+      "loss": 0.5486,
+      "step": 405
+    },
+    {
+      "epoch": 1.3231017770597737,
+      "grad_norm": 1.089034080505371,
+      "learning_rate": 2.6410246667379698e-05,
+      "loss": 0.5377,
+      "step": 410
+    },
+    {
+      "epoch": 1.3392568659127626,
+      "grad_norm": 1.1521645784378052,
+      "learning_rate": 2.630569537991042e-05,
+      "loss": 0.5249,
+      "step": 415
+    },
+    {
+      "epoch": 1.3554119547657513,
+      "grad_norm": 1.172168254852295,
+      "learning_rate": 2.6199856681213023e-05,
+      "loss": 0.4809,
+      "step": 420
+    },
+    {
+      "epoch": 1.37156704361874,
+      "grad_norm": 1.086441159248352,
+      "learning_rate": 2.6092742623436856e-05,
+      "loss": 0.4989,
+      "step": 425
+    },
+    {
+      "epoch": 1.3877221324717286,
+      "grad_norm": 1.279645323753357,
+      "learning_rate": 2.5984365403959966e-05,
+      "loss": 0.4776,
+      "step": 430
+    },
+    {
+      "epoch": 1.4038772213247173,
+      "grad_norm": 1.2726421356201172,
+      "learning_rate": 2.5874737364000167e-05,
+      "loss": 0.4759,
+      "step": 435
+    },
+    {
+      "epoch": 1.420032310177706,
+      "grad_norm": 1.104711890220642,
+      "learning_rate": 2.5763870987209704e-05,
+      "loss": 0.4877,
+      "step": 440
+    },
+    {
+      "epoch": 1.4361873990306946,
+      "grad_norm": 1.4142299890518188,
+      "learning_rate": 2.5651778898253718e-05,
+      "loss": 0.494,
+      "step": 445
+    },
+    {
+      "epoch": 1.4523424878836835,
+      "grad_norm": 1.147704839706421,
+      "learning_rate": 2.5538473861372628e-05,
+      "loss": 0.4801,
+      "step": 450
+    },
+    {
+      "epoch": 1.468497576736672,
+      "grad_norm": 1.1201279163360596,
+      "learning_rate": 2.542396877892862e-05,
+      "loss": 0.4552,
+      "step": 455
+    },
+    {
+      "epoch": 1.4846526655896608,
+      "grad_norm": 1.317685842514038,
+      "learning_rate": 2.5308276689936454e-05,
+      "loss": 0.4721,
+      "step": 460
+    },
+    {
+      "epoch": 1.5008077544426495,
+      "grad_norm": 1.111350417137146,
+      "learning_rate": 2.5191410768578643e-05,
+      "loss": 0.462,
+      "step": 465
+    },
+    {
+      "epoch": 1.5169628432956381,
+      "grad_norm": 1.1139695644378662,
+      "learning_rate": 2.5073384322705278e-05,
+      "loss": 0.4189,
+      "step": 470
+    },
+    {
+      "epoch": 1.5331179321486268,
+      "grad_norm": 1.1013826131820679,
+      "learning_rate": 2.4954210792318637e-05,
+      "loss": 0.4015,
+      "step": 475
+    },
+    {
+      "epoch": 1.5492730210016155,
+      "grad_norm": 1.0391738414764404,
+      "learning_rate": 2.483390374804272e-05,
+      "loss": 0.5059,
+      "step": 480
+    },
+    {
+      "epoch": 1.5654281098546043,
+      "grad_norm": 1.1611058712005615,
+      "learning_rate": 2.4712476889577944e-05,
+      "loss": 0.4572,
+      "step": 485
+    },
+    {
+      "epoch": 1.5815831987075928,
+      "grad_norm": 1.1287028789520264,
+      "learning_rate": 2.458994404414109e-05,
+      "loss": 0.4706,
+      "step": 490
+    },
+    {
+      "epoch": 1.5977382875605817,
+      "grad_norm": 1.2190521955490112,
+      "learning_rate": 2.4466319164890794e-05,
+      "loss": 0.4592,
+      "step": 495
+    },
+    {
+      "epoch": 1.6138933764135701,
+      "grad_norm": 1.0761996507644653,
+      "learning_rate": 2.434161632933863e-05,
+      "loss": 0.4168,
+      "step": 500
+    },
+    {
+      "epoch": 1.630048465266559,
+      "grad_norm": 1.2677689790725708,
+      "learning_rate": 2.4215849737746087e-05,
+      "loss": 0.4239,
+      "step": 505
+    },
+    {
+      "epoch": 1.6462035541195477,
+      "grad_norm": 1.077335238456726,
+      "learning_rate": 2.4089033711507523e-05,
+      "loss": 0.4223,
+      "step": 510
+    },
+    {
+      "epoch": 1.6623586429725363,
+      "grad_norm": 1.092629075050354,
+      "learning_rate": 2.3961182691519386e-05,
+      "loss": 0.396,
+      "step": 515
+    },
+    {
+      "epoch": 1.678513731825525,
+      "grad_norm": 1.0773464441299438,
+      "learning_rate": 2.383231123653574e-05,
+      "loss": 0.3559,
+      "step": 520
+    },
+    {
+      "epoch": 1.6946688206785137,
+      "grad_norm": 1.2922345399856567,
+      "learning_rate": 2.370243402151045e-05,
+      "loss": 0.4258,
+      "step": 525
+    },
+    {
+      "epoch": 1.7108239095315025,
+      "grad_norm": 1.1194257736206055,
+      "learning_rate": 2.35715658359261e-05,
+      "loss": 0.4128,
+      "step": 530
+    },
+    {
+      "epoch": 1.726978998384491,
+      "grad_norm": 1.2243081331253052,
+      "learning_rate": 2.3439721582109874e-05,
+      "loss": 0.4641,
+      "step": 535
+    },
+    {
+      "epoch": 1.7431340872374799,
+      "grad_norm": 1.1439075469970703,
+      "learning_rate": 2.3306916273536564e-05,
+      "loss": 0.3985,
+      "step": 540
+    },
+    {
+      "epoch": 1.7592891760904685,
+      "grad_norm": 1.09407639503479,
+      "learning_rate": 2.3173165033118983e-05,
+      "loss": 0.3708,
+      "step": 545
+    },
+    {
+      "epoch": 1.7754442649434572,
+      "grad_norm": 1.1844106912612915,
+      "learning_rate": 2.303848309148584e-05,
+      "loss": 0.4523,
+      "step": 550
+    },
+    {
+      "epoch": 1.7915993537964459,
+      "grad_norm": 1.1165605783462524,
+      "learning_rate": 2.2902885785247406e-05,
+      "loss": 0.3978,
+      "step": 555
+    },
+    {
+      "epoch": 1.8077544426494345,
+      "grad_norm": 1.1502916812896729,
+      "learning_rate": 2.2766388555249087e-05,
+      "loss": 0.3759,
+      "step": 560
+    },
+    {
+      "epoch": 1.8239095315024234,
+      "grad_norm": 1.003591537475586,
+      "learning_rate": 2.262900694481314e-05,
+      "loss": 0.3749,
+      "step": 565
+    },
+    {
+      "epoch": 1.8400646203554119,
+      "grad_norm": 1.1967657804489136,
+      "learning_rate": 2.2490756597968663e-05,
+      "loss": 0.375,
+      "step": 570
+    },
+    {
+      "epoch": 1.8562197092084007,
+      "grad_norm": 1.1565355062484741,
+      "learning_rate": 2.235165325767026e-05,
+      "loss": 0.3611,
+      "step": 575
+    },
+    {
+      "epoch": 1.8723747980613892,
+      "grad_norm": 1.1527678966522217,
+      "learning_rate": 2.2211712764005245e-05,
+      "loss": 0.4101,
+      "step": 580
+    },
+    {
+      "epoch": 1.888529886914378,
+      "grad_norm": 1.2227520942687988,
+      "learning_rate": 2.207095105238997e-05,
+      "loss": 0.3426,
+      "step": 585
+    },
+    {
+      "epoch": 1.9046849757673667,
+      "grad_norm": 1.1511895656585693,
+      "learning_rate": 2.1929384151755138e-05,
+      "loss": 0.3967,
+      "step": 590
+    },
+    {
+      "epoch": 1.9208400646203554,
+      "grad_norm": 1.1515477895736694,
+      "learning_rate": 2.1787028182720606e-05,
+      "loss": 0.3717,
+      "step": 595
+    },
+    {
+      "epoch": 1.936995153473344,
+      "grad_norm": 1.1370980739593506,
+      "learning_rate": 2.1643899355759674e-05,
+      "loss": 0.3588,
+      "step": 600
+    },
+    {
+      "epoch": 1.9531502423263327,
+      "grad_norm": 1.209424376487732,
+      "learning_rate": 2.1500013969353107e-05,
+      "loss": 0.329,
+      "step": 605
+    },
+    {
+      "epoch": 1.9693053311793216,
+      "grad_norm": 1.08406662940979,
+      "learning_rate": 2.1355388408133254e-05,
+      "loss": 0.3333,
+      "step": 610
+    },
+    {
+      "epoch": 1.98546042003231,
+      "grad_norm": 1.0804470777511597,
+      "learning_rate": 2.1210039141018204e-05,
+      "loss": 0.3432,
+      "step": 615
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6371866464614868,
+      "learning_rate": 2.1063982719336497e-05,
+      "loss": 0.3595,
+      "step": 620
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1550,
+  "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": 9.190029403650785e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

13/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8eb054144fcb38fbe96cf3375334d0fd0faf77b0413bc06783881a4111681914
+size 7736

13/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)