Upload folder using huggingface_hub

8/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

8/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": [
+    "o_proj",
+    "up_proj",
+    "v_proj",
+    "k_proj",
+    "gate_proj",
+    "q_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

8/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0b507ef16d3fded6884b9cbdf3f9808d9d1b7a5e7363df0ced105d82e66aa56
+size 791751704

8/latest

@@ -0,0 +1 @@
+global_step558

8/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:179d4ab53838324387824b06d20029dc88ebb7efd0eaac9d436857412446e7f0
+size 14960

8/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec0fe3597c75e0b683bc14f52e28675a84ff3bb8de71bf4327337f4a6da3e791
+size 14960

8/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14ca87e90b860193fb6719142097f879398933bab862a92f6aeab348c962012c
+size 14960

8/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a009f88ada0f4abb3b02e544c0529459eb60837d39e8e22c16e4ecc44f77a3d
+size 14960

8/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7c7c84a4af7cfcdea2d7506b84740479520cd8ec090512e9cfc725b5fcfece5
+size 1064

8/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
+  }
+}

8/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
+}

8/trainer_state.json

@@ -0,0 +1,811 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 558,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.017937219730941704,
+      "grad_norm": 1.4301773309707642,
+      "learning_rate": 1.7142857142857143e-06,
+      "loss": 1.2446,
+      "step": 5
+    },
+    {
+      "epoch": 0.03587443946188341,
+      "grad_norm": 0.7787048816680908,
+      "learning_rate": 3.857142857142857e-06,
+      "loss": 1.317,
+      "step": 10
+    },
+    {
+      "epoch": 0.053811659192825115,
+      "grad_norm": 0.5886793732643127,
+      "learning_rate": 6e-06,
+      "loss": 1.2483,
+      "step": 15
+    },
+    {
+      "epoch": 0.07174887892376682,
+      "grad_norm": 0.558383047580719,
+      "learning_rate": 8.142857142857142e-06,
+      "loss": 1.2155,
+      "step": 20
+    },
+    {
+      "epoch": 0.08968609865470852,
+      "grad_norm": 0.6142293214797974,
+      "learning_rate": 1.0285714285714286e-05,
+      "loss": 1.2447,
+      "step": 25
+    },
+    {
+      "epoch": 0.10762331838565023,
+      "grad_norm": 0.5331266522407532,
+      "learning_rate": 1.242857142857143e-05,
+      "loss": 1.1889,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556053811659193,
+      "grad_norm": 1.1998621225357056,
+      "learning_rate": 1.4571428571428571e-05,
+      "loss": 1.2313,
+      "step": 35
+    },
+    {
+      "epoch": 0.14349775784753363,
+      "grad_norm": 0.41463860869407654,
+      "learning_rate": 1.6714285714285716e-05,
+      "loss": 1.171,
+      "step": 40
+    },
+    {
+      "epoch": 0.16143497757847533,
+      "grad_norm": 0.4496464431285858,
+      "learning_rate": 1.8857142857142856e-05,
+      "loss": 1.162,
+      "step": 45
+    },
+    {
+      "epoch": 0.17937219730941703,
+      "grad_norm": 0.41104015707969666,
+      "learning_rate": 2.1e-05,
+      "loss": 1.1688,
+      "step": 50
+    },
+    {
+      "epoch": 0.19730941704035873,
+      "grad_norm": 0.49826157093048096,
+      "learning_rate": 2.3142857142857145e-05,
+      "loss": 1.1011,
+      "step": 55
+    },
+    {
+      "epoch": 0.21524663677130046,
+      "grad_norm": 0.4497723877429962,
+      "learning_rate": 2.5285714285714285e-05,
+      "loss": 1.102,
+      "step": 60
+    },
+    {
+      "epoch": 0.23318385650224216,
+      "grad_norm": 0.44472333788871765,
+      "learning_rate": 2.7428571428571428e-05,
+      "loss": 1.1289,
+      "step": 65
+    },
+    {
+      "epoch": 0.25112107623318386,
+      "grad_norm": 0.49345752596855164,
+      "learning_rate": 2.9571428571428575e-05,
+      "loss": 1.0881,
+      "step": 70
+    },
+    {
+      "epoch": 0.26905829596412556,
+      "grad_norm": 0.47529274225234985,
+      "learning_rate": 2.999932540055495e-05,
+      "loss": 1.1045,
+      "step": 75
+    },
+    {
+      "epoch": 0.28699551569506726,
+      "grad_norm": 0.6426385045051575,
+      "learning_rate": 2.9996584944302995e-05,
+      "loss": 1.0835,
+      "step": 80
+    },
+    {
+      "epoch": 0.30493273542600896,
+      "grad_norm": 0.4679868221282959,
+      "learning_rate": 2.9991736853628892e-05,
+      "loss": 1.043,
+      "step": 85
+    },
+    {
+      "epoch": 0.32286995515695066,
+      "grad_norm": 0.6055300831794739,
+      "learning_rate": 2.9984781809887943e-05,
+      "loss": 1.0091,
+      "step": 90
+    },
+    {
+      "epoch": 0.34080717488789236,
+      "grad_norm": 0.575788676738739,
+      "learning_rate": 2.997572079054867e-05,
+      "loss": 1.0445,
+      "step": 95
+    },
+    {
+      "epoch": 0.35874439461883406,
+      "grad_norm": 0.610512375831604,
+      "learning_rate": 2.9964555069055416e-05,
+      "loss": 0.9675,
+      "step": 100
+    },
+    {
+      "epoch": 0.37668161434977576,
+      "grad_norm": 0.6060314774513245,
+      "learning_rate": 2.9951286214649374e-05,
+      "loss": 1.0158,
+      "step": 105
+    },
+    {
+      "epoch": 0.39461883408071746,
+      "grad_norm": 0.6149461269378662,
+      "learning_rate": 2.993591609214806e-05,
+      "loss": 1.009,
+      "step": 110
+    },
+    {
+      "epoch": 0.4125560538116592,
+      "grad_norm": 0.6386514902114868,
+      "learning_rate": 2.9918446861683233e-05,
+      "loss": 0.9174,
+      "step": 115
+    },
+    {
+      "epoch": 0.4304932735426009,
+      "grad_norm": 0.591651976108551,
+      "learning_rate": 2.989888097839728e-05,
+      "loss": 0.9227,
+      "step": 120
+    },
+    {
+      "epoch": 0.4484304932735426,
+      "grad_norm": 0.7576113939285278,
+      "learning_rate": 2.9877221192098192e-05,
+      "loss": 0.9412,
+      "step": 125
+    },
+    {
+      "epoch": 0.4663677130044843,
+      "grad_norm": 0.7285865545272827,
+      "learning_rate": 2.9853470546873093e-05,
+      "loss": 0.9057,
+      "step": 130
+    },
+    {
+      "epoch": 0.484304932735426,
+      "grad_norm": 0.7394405603408813,
+      "learning_rate": 2.9827632380660435e-05,
+      "loss": 0.9103,
+      "step": 135
+    },
+    {
+      "epoch": 0.5022421524663677,
+      "grad_norm": 0.6587325930595398,
+      "learning_rate": 2.9799710324780854e-05,
+      "loss": 0.8788,
+      "step": 140
+    },
+    {
+      "epoch": 0.5201793721973094,
+      "grad_norm": 0.7074690461158752,
+      "learning_rate": 2.9769708303426855e-05,
+      "loss": 0.9147,
+      "step": 145
+    },
+    {
+      "epoch": 0.5381165919282511,
+      "grad_norm": 0.8340660929679871,
+      "learning_rate": 2.973763053311127e-05,
+      "loss": 0.8934,
+      "step": 150
+    },
+    {
+      "epoch": 0.5560538116591929,
+      "grad_norm": 0.7627359628677368,
+      "learning_rate": 2.9703481522074692e-05,
+      "loss": 0.8488,
+      "step": 155
+    },
+    {
+      "epoch": 0.5739910313901345,
+      "grad_norm": 0.7680438756942749,
+      "learning_rate": 2.9667266069651855e-05,
+      "loss": 0.8333,
+      "step": 160
+    },
+    {
+      "epoch": 0.5919282511210763,
+      "grad_norm": 0.8545948266983032,
+      "learning_rate": 2.9628989265597162e-05,
+      "loss": 0.8482,
+      "step": 165
+    },
+    {
+      "epoch": 0.6098654708520179,
+      "grad_norm": 0.8448952436447144,
+      "learning_rate": 2.9588656489369338e-05,
+      "loss": 0.8114,
+      "step": 170
+    },
+    {
+      "epoch": 0.6278026905829597,
+      "grad_norm": 0.9756552577018738,
+      "learning_rate": 2.9546273409375413e-05,
+      "loss": 0.8041,
+      "step": 175
+    },
+    {
+      "epoch": 0.6457399103139013,
+      "grad_norm": 0.9768025279045105,
+      "learning_rate": 2.9501845982174075e-05,
+      "loss": 0.7813,
+      "step": 180
+    },
+    {
+      "epoch": 0.6636771300448431,
+      "grad_norm": 0.7772969603538513,
+      "learning_rate": 2.9455380451638528e-05,
+      "loss": 0.7701,
+      "step": 185
+    },
+    {
+      "epoch": 0.6816143497757847,
+      "grad_norm": 0.8379116654396057,
+      "learning_rate": 2.9406883348078965e-05,
+      "loss": 0.7999,
+      "step": 190
+    },
+    {
+      "epoch": 0.6995515695067265,
+      "grad_norm": 1.05807626247406,
+      "learning_rate": 2.935636148732481e-05,
+      "loss": 0.8285,
+      "step": 195
+    },
+    {
+      "epoch": 0.7174887892376681,
+      "grad_norm": 0.8303896188735962,
+      "learning_rate": 2.930382196976679e-05,
+      "loss": 0.8002,
+      "step": 200
+    },
+    {
+      "epoch": 0.7354260089686099,
+      "grad_norm": 1.0743650197982788,
+      "learning_rate": 2.9249272179359072e-05,
+      "loss": 0.7469,
+      "step": 205
+    },
+    {
+      "epoch": 0.7533632286995515,
+      "grad_norm": 0.8186368942260742,
+      "learning_rate": 2.9192719782581473e-05,
+      "loss": 0.7351,
+      "step": 210
+    },
+    {
+      "epoch": 0.7713004484304933,
+      "grad_norm": 0.8115820288658142,
+      "learning_rate": 2.9134172727362055e-05,
+      "loss": 0.7114,
+      "step": 215
+    },
+    {
+      "epoch": 0.7892376681614349,
+      "grad_norm": 0.8681687712669373,
+      "learning_rate": 2.9073639241960066e-05,
+      "loss": 0.7404,
+      "step": 220
+    },
+    {
+      "epoch": 0.8071748878923767,
+      "grad_norm": 1.064688801765442,
+      "learning_rate": 2.9011127833809565e-05,
+      "loss": 0.7015,
+      "step": 225
+    },
+    {
+      "epoch": 0.8251121076233184,
+      "grad_norm": 0.9685300588607788,
+      "learning_rate": 2.894664728832377e-05,
+      "loss": 0.7099,
+      "step": 230
+    },
+    {
+      "epoch": 0.8430493273542601,
+      "grad_norm": 0.890418291091919,
+      "learning_rate": 2.8880206667660356e-05,
+      "loss": 0.7268,
+      "step": 235
+    },
+    {
+      "epoch": 0.8609865470852018,
+      "grad_norm": 0.9415674805641174,
+      "learning_rate": 2.8811815309447822e-05,
+      "loss": 0.6661,
+      "step": 240
+    },
+    {
+      "epoch": 0.8789237668161435,
+      "grad_norm": 0.9636686444282532,
+      "learning_rate": 2.874148282547321e-05,
+      "loss": 0.7095,
+      "step": 245
+    },
+    {
+      "epoch": 0.8968609865470852,
+      "grad_norm": 0.8985582590103149,
+      "learning_rate": 2.866921910033123e-05,
+      "loss": 0.6363,
+      "step": 250
+    },
+    {
+      "epoch": 0.9147982062780269,
+      "grad_norm": 1.078536868095398,
+      "learning_rate": 2.8595034290035082e-05,
+      "loss": 0.6477,
+      "step": 255
+    },
+    {
+      "epoch": 0.9327354260089686,
+      "grad_norm": 0.9233158230781555,
+      "learning_rate": 2.85189388205891e-05,
+      "loss": 0.6918,
+      "step": 260
+    },
+    {
+      "epoch": 0.9506726457399103,
+      "grad_norm": 0.9907985329627991,
+      "learning_rate": 2.8440943386523507e-05,
+      "loss": 0.6517,
+      "step": 265
+    },
+    {
+      "epoch": 0.968609865470852,
+      "grad_norm": 0.9845608472824097,
+      "learning_rate": 2.8361058949391357e-05,
+      "loss": 0.6266,
+      "step": 270
+    },
+    {
+      "epoch": 0.9865470852017937,
+      "grad_norm": 1.2447435855865479,
+      "learning_rate": 2.827929673622801e-05,
+      "loss": 0.6589,
+      "step": 275
+    },
+    {
+      "epoch": 1.0035874439461883,
+      "grad_norm": 1.0809400081634521,
+      "learning_rate": 2.8195668237973287e-05,
+      "loss": 0.636,
+      "step": 280
+    },
+    {
+      "epoch": 1.02152466367713,
+      "grad_norm": 1.276408314704895,
+      "learning_rate": 2.811018520785648e-05,
+      "loss": 0.5796,
+      "step": 285
+    },
+    {
+      "epoch": 1.0394618834080718,
+      "grad_norm": 0.9826470613479614,
+      "learning_rate": 2.8022859659744574e-05,
+      "loss": 0.5407,
+      "step": 290
+    },
+    {
+      "epoch": 1.0573991031390135,
+      "grad_norm": 1.0762982368469238,
+      "learning_rate": 2.793370386645381e-05,
+      "loss": 0.5557,
+      "step": 295
+    },
+    {
+      "epoch": 1.075336322869955,
+      "grad_norm": 1.1283237934112549,
+      "learning_rate": 2.7842730358024824e-05,
+      "loss": 0.5067,
+      "step": 300
+    },
+    {
+      "epoch": 1.0932735426008968,
+      "grad_norm": 0.9394952654838562,
+      "learning_rate": 2.7749951919961687e-05,
+      "loss": 0.5487,
+      "step": 305
+    },
+    {
+      "epoch": 1.1112107623318386,
+      "grad_norm": 0.9326608777046204,
+      "learning_rate": 2.7655381591435017e-05,
+      "loss": 0.5323,
+      "step": 310
+    },
+    {
+      "epoch": 1.1291479820627803,
+      "grad_norm": 1.0368235111236572,
+      "learning_rate": 2.7559032663449428e-05,
+      "loss": 0.5469,
+      "step": 315
+    },
+    {
+      "epoch": 1.1470852017937219,
+      "grad_norm": 1.029537320137024,
+      "learning_rate": 2.7460918676975612e-05,
+      "loss": 0.5319,
+      "step": 320
+    },
+    {
+      "epoch": 1.1650224215246636,
+      "grad_norm": 0.9895524978637695,
+      "learning_rate": 2.7361053421047264e-05,
+      "loss": 0.491,
+      "step": 325
+    },
+    {
+      "epoch": 1.1829596412556054,
+      "grad_norm": 1.0762293338775635,
+      "learning_rate": 2.725945093082317e-05,
+      "loss": 0.5276,
+      "step": 330
+    },
+    {
+      "epoch": 1.2008968609865471,
+      "grad_norm": 0.9648789167404175,
+      "learning_rate": 2.71561254856147e-05,
+      "loss": 0.5381,
+      "step": 335
+    },
+    {
+      "epoch": 1.2188340807174889,
+      "grad_norm": 1.0773042440414429,
+      "learning_rate": 2.705109160687895e-05,
+      "loss": 0.5558,
+      "step": 340
+    },
+    {
+      "epoch": 1.2367713004484304,
+      "grad_norm": 1.087357997894287,
+      "learning_rate": 2.6944364056177923e-05,
+      "loss": 0.467,
+      "step": 345
+    },
+    {
+      "epoch": 1.2547085201793722,
+      "grad_norm": 0.9490891695022583,
+      "learning_rate": 2.6835957833103878e-05,
+      "loss": 0.4802,
+      "step": 350
+    },
+    {
+      "epoch": 1.272645739910314,
+      "grad_norm": 1.0852769613265991,
+      "learning_rate": 2.6725888173171324e-05,
+      "loss": 0.5189,
+      "step": 355
+    },
+    {
+      "epoch": 1.2905829596412555,
+      "grad_norm": 1.1488758325576782,
+      "learning_rate": 2.6614170545675763e-05,
+      "loss": 0.4788,
+      "step": 360
+    },
+    {
+      "epoch": 1.3085201793721972,
+      "grad_norm": 1.3093827962875366,
+      "learning_rate": 2.6500820651519644e-05,
+      "loss": 0.4793,
+      "step": 365
+    },
+    {
+      "epoch": 1.326457399103139,
+      "grad_norm": 1.05547297000885,
+      "learning_rate": 2.6385854421005738e-05,
+      "loss": 0.4134,
+      "step": 370
+    },
+    {
+      "epoch": 1.3443946188340807,
+      "grad_norm": 1.1201064586639404,
+      "learning_rate": 2.6269288011598273e-05,
+      "loss": 0.4493,
+      "step": 375
+    },
+    {
+      "epoch": 1.3623318385650225,
+      "grad_norm": 1.058740496635437,
+      "learning_rate": 2.6151137805652154e-05,
+      "loss": 0.4951,
+      "step": 380
+    },
+    {
+      "epoch": 1.3802690582959642,
+      "grad_norm": 0.9838550090789795,
+      "learning_rate": 2.6031420408110575e-05,
+      "loss": 0.4211,
+      "step": 385
+    },
+    {
+      "epoch": 1.3982062780269058,
+      "grad_norm": 1.001670241355896,
+      "learning_rate": 2.5910152644171335e-05,
+      "loss": 0.4272,
+      "step": 390
+    },
+    {
+      "epoch": 1.4161434977578475,
+      "grad_norm": 1.1107715368270874,
+      "learning_rate": 2.5787351556922207e-05,
+      "loss": 0.4145,
+      "step": 395
+    },
+    {
+      "epoch": 1.4340807174887893,
+      "grad_norm": 1.260655403137207,
+      "learning_rate": 2.5663034404945698e-05,
+      "loss": 0.4641,
+      "step": 400
+    },
+    {
+      "epoch": 1.4520179372197308,
+      "grad_norm": 1.086647868156433,
+      "learning_rate": 2.5537218659893505e-05,
+      "loss": 0.4421,
+      "step": 405
+    },
+    {
+      "epoch": 1.4699551569506726,
+      "grad_norm": 1.1397958993911743,
+      "learning_rate": 2.540992200403103e-05,
+      "loss": 0.4478,
+      "step": 410
+    },
+    {
+      "epoch": 1.4878923766816143,
+      "grad_norm": 1.0688517093658447,
+      "learning_rate": 2.52811623277523e-05,
+      "loss": 0.4596,
+      "step": 415
+    },
+    {
+      "epoch": 1.505829596412556,
+      "grad_norm": 1.0559828281402588,
+      "learning_rate": 2.5150957727065634e-05,
+      "loss": 0.4188,
+      "step": 420
+    },
+    {
+      "epoch": 1.5237668161434978,
+      "grad_norm": 1.0967782735824585,
+      "learning_rate": 2.5019326501050414e-05,
+      "loss": 0.4454,
+      "step": 425
+    },
+    {
+      "epoch": 1.5417040358744396,
+      "grad_norm": 1.293083906173706,
+      "learning_rate": 2.4886287149285315e-05,
+      "loss": 0.4408,
+      "step": 430
+    },
+    {
+      "epoch": 1.5596412556053811,
+      "grad_norm": 1.038537859916687,
+      "learning_rate": 2.4751858369248338e-05,
+      "loss": 0.3902,
+      "step": 435
+    },
+    {
+      "epoch": 1.577578475336323,
+      "grad_norm": 1.0545300245285034,
+      "learning_rate": 2.4616059053689093e-05,
+      "loss": 0.4084,
+      "step": 440
+    },
+    {
+      "epoch": 1.5955156950672644,
+      "grad_norm": 1.1569932699203491,
+      "learning_rate": 2.4478908287973542e-05,
+      "loss": 0.4227,
+      "step": 445
+    },
+    {
+      "epoch": 1.6134529147982062,
+      "grad_norm": 1.146126389503479,
+      "learning_rate": 2.4340425347401758e-05,
+      "loss": 0.3688,
+      "step": 450
+    },
+    {
+      "epoch": 1.631390134529148,
+      "grad_norm": 1.0478739738464355,
+      "learning_rate": 2.4200629694498952e-05,
+      "loss": 0.4408,
+      "step": 455
+    },
+    {
+      "epoch": 1.6493273542600897,
+      "grad_norm": 1.0405068397521973,
+      "learning_rate": 2.4059540976280155e-05,
+      "loss": 0.4131,
+      "step": 460
+    },
+    {
+      "epoch": 1.6672645739910315,
+      "grad_norm": 1.174987554550171,
+      "learning_rate": 2.3917179021489078e-05,
+      "loss": 0.3926,
+      "step": 465
+    },
+    {
+      "epoch": 1.6852017937219732,
+      "grad_norm": 1.1956828832626343,
+      "learning_rate": 2.3773563837811294e-05,
+      "loss": 0.3803,
+      "step": 470
+    },
+    {
+      "epoch": 1.703139013452915,
+      "grad_norm": 1.2239809036254883,
+      "learning_rate": 2.362871560906237e-05,
+      "loss": 0.3666,
+      "step": 475
+    },
+    {
+      "epoch": 1.7210762331838565,
+      "grad_norm": 1.0986831188201904,
+      "learning_rate": 2.34826546923512e-05,
+      "loss": 0.4092,
+      "step": 480
+    },
+    {
+      "epoch": 1.7390134529147983,
+      "grad_norm": 1.0886595249176025,
+      "learning_rate": 2.3335401615219037e-05,
+      "loss": 0.4012,
+      "step": 485
+    },
+    {
+      "epoch": 1.7569506726457398,
+      "grad_norm": 1.2974740266799927,
+      "learning_rate": 2.3186977072754462e-05,
+      "loss": 0.3791,
+      "step": 490
+    },
+    {
+      "epoch": 1.7748878923766815,
+      "grad_norm": 1.117116093635559,
+      "learning_rate": 2.303740192468495e-05,
+      "loss": 0.3965,
+      "step": 495
+    },
+    {
+      "epoch": 1.7928251121076233,
+      "grad_norm": 1.0683351755142212,
+      "learning_rate": 2.2886697192445175e-05,
+      "loss": 0.3661,
+      "step": 500
+    },
+    {
+      "epoch": 1.810762331838565,
+      "grad_norm": 1.212296485900879,
+      "learning_rate": 2.2734884056222643e-05,
+      "loss": 0.394,
+      "step": 505
+    },
+    {
+      "epoch": 1.8286995515695068,
+      "grad_norm": 1.1680309772491455,
+      "learning_rate": 2.2581983851981044e-05,
+      "loss": 0.3218,
+      "step": 510
+    },
+    {
+      "epoch": 1.8466367713004486,
+      "grad_norm": 1.145910620689392,
+      "learning_rate": 2.2428018068461633e-05,
+      "loss": 0.3619,
+      "step": 515
+    },
+    {
+      "epoch": 1.86457399103139,
+      "grad_norm": 1.188434362411499,
+      "learning_rate": 2.2273008344163195e-05,
+      "loss": 0.3369,
+      "step": 520
+    },
+    {
+      "epoch": 1.8825112107623319,
+      "grad_norm": 1.100471019744873,
+      "learning_rate": 2.2116976464300964e-05,
+      "loss": 0.3335,
+      "step": 525
+    },
+    {
+      "epoch": 1.9004484304932734,
+      "grad_norm": 1.056118369102478,
+      "learning_rate": 2.1959944357744887e-05,
+      "loss": 0.3826,
+      "step": 530
+    },
+    {
+      "epoch": 1.9183856502242151,
+      "grad_norm": 1.237924575805664,
+      "learning_rate": 2.180193409393773e-05,
+      "loss": 0.338,
+      "step": 535
+    },
+    {
+      "epoch": 1.936322869955157,
+      "grad_norm": 1.1316227912902832,
+      "learning_rate": 2.164296787979343e-05,
+      "loss": 0.3511,
+      "step": 540
+    },
+    {
+      "epoch": 1.9542600896860987,
+      "grad_norm": 1.0867352485656738,
+      "learning_rate": 2.1483068056576082e-05,
+      "loss": 0.299,
+      "step": 545
+    },
+    {
+      "epoch": 1.9721973094170404,
+      "grad_norm": 0.9542579650878906,
+      "learning_rate": 2.1322257096760104e-05,
+      "loss": 0.3592,
+      "step": 550
+    },
+    {
+      "epoch": 1.9901345291479822,
+      "grad_norm": 1.3199540376663208,
+      "learning_rate": 2.116055760087193e-05,
+      "loss": 0.3621,
+      "step": 555
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1395,
+  "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": 8.451719003751055e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

8/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74b47b640d1564458a44dfc258606f561c4b84de1e6d631e97464f62c56b91fe
+size 7736

8/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)