Upload folder using huggingface_hub

7/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

7/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": [
+    "k_proj",
+    "down_proj",
+    "v_proj",
+    "o_proj",
+    "gate_proj",
+    "up_proj",
+    "q_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

7/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75997d3fd098a71a5465ab0ce8ecbbeb2b6a59cb7ca0b6d8bf28f155bce416ee
+size 791751704

7/latest

@@ -0,0 +1 @@
+global_step560

7/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b24a66dc748a74acf6aaca6673e77c617c67f25d6982eda281d3a56f8efdf4
+size 14960

7/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:defbe252df7c94f62350c2f71b437d05eb60becf7eb2523d3ed39853092b32c6
+size 14960

7/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cc25ae9ce512a29297b21cba1e92bd5a8168316430619812a9d3941b8fe57a5
+size 14960

7/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2886b0fc1c909d968f1510fc24e3a8f43c25d56dda6e3ee06203b6a990cf775
+size 14960

7/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1f342196805255b489339f9b50342ae2ab2a6b2d3f159be762af00327d7cef3
+size 1064

7/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
+  }
+}

7/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
+}

7/trainer_state.json

@@ -0,0 +1,818 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 560,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.017889087656529516,
+      "grad_norm": 1.506656527519226,
+      "learning_rate": 1.7142857142857143e-06,
+      "loss": 1.2584,
+      "step": 5
+    },
+    {
+      "epoch": 0.03577817531305903,
+      "grad_norm": 0.9500389695167542,
+      "learning_rate": 3.857142857142857e-06,
+      "loss": 1.2642,
+      "step": 10
+    },
+    {
+      "epoch": 0.05366726296958855,
+      "grad_norm": 0.7999431490898132,
+      "learning_rate": 6e-06,
+      "loss": 1.3278,
+      "step": 15
+    },
+    {
+      "epoch": 0.07155635062611806,
+      "grad_norm": 0.679018497467041,
+      "learning_rate": 8.142857142857142e-06,
+      "loss": 1.2183,
+      "step": 20
+    },
+    {
+      "epoch": 0.08944543828264759,
+      "grad_norm": 0.5348811745643616,
+      "learning_rate": 1.0285714285714286e-05,
+      "loss": 1.2252,
+      "step": 25
+    },
+    {
+      "epoch": 0.1073345259391771,
+      "grad_norm": 0.6638433337211609,
+      "learning_rate": 1.242857142857143e-05,
+      "loss": 1.2385,
+      "step": 30
+    },
+    {
+      "epoch": 0.1252236135957066,
+      "grad_norm": 0.7031165361404419,
+      "learning_rate": 1.4571428571428571e-05,
+      "loss": 1.2,
+      "step": 35
+    },
+    {
+      "epoch": 0.14311270125223613,
+      "grad_norm": 0.39686235785484314,
+      "learning_rate": 1.6714285714285716e-05,
+      "loss": 1.161,
+      "step": 40
+    },
+    {
+      "epoch": 0.16100178890876565,
+      "grad_norm": 0.5343768000602722,
+      "learning_rate": 1.8857142857142856e-05,
+      "loss": 1.2075,
+      "step": 45
+    },
+    {
+      "epoch": 0.17889087656529518,
+      "grad_norm": 0.47290146350860596,
+      "learning_rate": 2.1e-05,
+      "loss": 1.186,
+      "step": 50
+    },
+    {
+      "epoch": 0.1967799642218247,
+      "grad_norm": 0.5593580007553101,
+      "learning_rate": 2.3142857142857145e-05,
+      "loss": 1.1161,
+      "step": 55
+    },
+    {
+      "epoch": 0.2146690518783542,
+      "grad_norm": 0.4979853332042694,
+      "learning_rate": 2.5285714285714285e-05,
+      "loss": 1.176,
+      "step": 60
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "grad_norm": 0.5528405904769897,
+      "learning_rate": 2.7428571428571428e-05,
+      "loss": 1.1656,
+      "step": 65
+    },
+    {
+      "epoch": 0.2504472271914132,
+      "grad_norm": 0.46704745292663574,
+      "learning_rate": 2.9571428571428575e-05,
+      "loss": 1.1207,
+      "step": 70
+    },
+    {
+      "epoch": 0.26833631484794274,
+      "grad_norm": 0.6753780245780945,
+      "learning_rate": 2.99993304631594e-05,
+      "loss": 1.0555,
+      "step": 75
+    },
+    {
+      "epoch": 0.28622540250447226,
+      "grad_norm": 0.4959678649902344,
+      "learning_rate": 2.999661057218302e-05,
+      "loss": 1.0949,
+      "step": 80
+    },
+    {
+      "epoch": 0.3041144901610018,
+      "grad_norm": 0.5355138778686523,
+      "learning_rate": 2.999179886011389e-05,
+      "loss": 1.0936,
+      "step": 85
+    },
+    {
+      "epoch": 0.3220035778175313,
+      "grad_norm": 0.5441014766693115,
+      "learning_rate": 2.9984895998119723e-05,
+      "loss": 1.0513,
+      "step": 90
+    },
+    {
+      "epoch": 0.33989266547406083,
+      "grad_norm": 0.5444539785385132,
+      "learning_rate": 2.99759029490549e-05,
+      "loss": 1.0729,
+      "step": 95
+    },
+    {
+      "epoch": 0.35778175313059035,
+      "grad_norm": 0.5886111855506897,
+      "learning_rate": 2.996482096732619e-05,
+      "loss": 1.0365,
+      "step": 100
+    },
+    {
+      "epoch": 0.3756708407871199,
+      "grad_norm": 0.6149312853813171,
+      "learning_rate": 2.9951651598717757e-05,
+      "loss": 1.0499,
+      "step": 105
+    },
+    {
+      "epoch": 0.3935599284436494,
+      "grad_norm": 0.5841829180717468,
+      "learning_rate": 2.9936396680175547e-05,
+      "loss": 0.9816,
+      "step": 110
+    },
+    {
+      "epoch": 0.41144901610017887,
+      "grad_norm": 0.7359195947647095,
+      "learning_rate": 2.9919058339551068e-05,
+      "loss": 0.9981,
+      "step": 115
+    },
+    {
+      "epoch": 0.4293381037567084,
+      "grad_norm": 0.6694313883781433,
+      "learning_rate": 2.9899638995304575e-05,
+      "loss": 0.955,
+      "step": 120
+    },
+    {
+      "epoch": 0.4472271914132379,
+      "grad_norm": 0.7220502495765686,
+      "learning_rate": 2.9878141356167725e-05,
+      "loss": 0.8753,
+      "step": 125
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "grad_norm": 0.6825446486473083,
+      "learning_rate": 2.9854568420765768e-05,
+      "loss": 0.961,
+      "step": 130
+    },
+    {
+      "epoch": 0.48300536672629696,
+      "grad_norm": 0.7498815655708313,
+      "learning_rate": 2.982892347719925e-05,
+      "loss": 0.9166,
+      "step": 135
+    },
+    {
+      "epoch": 0.5008944543828264,
+      "grad_norm": 0.7284743189811707,
+      "learning_rate": 2.9801210102585393e-05,
+      "loss": 0.9351,
+      "step": 140
+    },
+    {
+      "epoch": 0.518783542039356,
+      "grad_norm": 0.8820511102676392,
+      "learning_rate": 2.9771432162559113e-05,
+      "loss": 0.9124,
+      "step": 145
+    },
+    {
+      "epoch": 0.5366726296958855,
+      "grad_norm": 0.7539326548576355,
+      "learning_rate": 2.973959381073384e-05,
+      "loss": 0.846,
+      "step": 150
+    },
+    {
+      "epoch": 0.554561717352415,
+      "grad_norm": 0.7277606129646301,
+      "learning_rate": 2.970569948812214e-05,
+      "loss": 0.9432,
+      "step": 155
+    },
+    {
+      "epoch": 0.5724508050089445,
+      "grad_norm": 0.7533125281333923,
+      "learning_rate": 2.966975392251624e-05,
+      "loss": 0.8557,
+      "step": 160
+    },
+    {
+      "epoch": 0.590339892665474,
+      "grad_norm": 0.7943282723426819,
+      "learning_rate": 2.9631762127828584e-05,
+      "loss": 0.8635,
+      "step": 165
+    },
+    {
+      "epoch": 0.6082289803220036,
+      "grad_norm": 0.8900126814842224,
+      "learning_rate": 2.9591729403392447e-05,
+      "loss": 0.856,
+      "step": 170
+    },
+    {
+      "epoch": 0.6261180679785331,
+      "grad_norm": 0.8534295558929443,
+      "learning_rate": 2.9549661333222764e-05,
+      "loss": 0.8737,
+      "step": 175
+    },
+    {
+      "epoch": 0.6440071556350626,
+      "grad_norm": 0.7521919012069702,
+      "learning_rate": 2.950556378523723e-05,
+      "loss": 0.8129,
+      "step": 180
+    },
+    {
+      "epoch": 0.6618962432915921,
+      "grad_norm": 0.8599235415458679,
+      "learning_rate": 2.9459442910437798e-05,
+      "loss": 0.7975,
+      "step": 185
+    },
+    {
+      "epoch": 0.6797853309481217,
+      "grad_norm": 0.9698217511177063,
+      "learning_rate": 2.9411305142052725e-05,
+      "loss": 0.8128,
+      "step": 190
+    },
+    {
+      "epoch": 0.6976744186046512,
+      "grad_norm": 0.9345831871032715,
+      "learning_rate": 2.9361157194639184e-05,
+      "loss": 0.7551,
+      "step": 195
+    },
+    {
+      "epoch": 0.7155635062611807,
+      "grad_norm": 0.7919726967811584,
+      "learning_rate": 2.9309006063146716e-05,
+      "loss": 0.7606,
+      "step": 200
+    },
+    {
+      "epoch": 0.7334525939177102,
+      "grad_norm": 0.8289109468460083,
+      "learning_rate": 2.925485902194151e-05,
+      "loss": 0.7793,
+      "step": 205
+    },
+    {
+      "epoch": 0.7513416815742398,
+      "grad_norm": 0.9304677248001099,
+      "learning_rate": 2.9198723623791724e-05,
+      "loss": 0.711,
+      "step": 210
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.0459802150726318,
+      "learning_rate": 2.9140607698814e-05,
+      "loss": 0.737,
+      "step": 215
+    },
+    {
+      "epoch": 0.7871198568872988,
+      "grad_norm": 0.9233027696609497,
+      "learning_rate": 2.9080519353381243e-05,
+      "loss": 0.7234,
+      "step": 220
+    },
+    {
+      "epoch": 0.8050089445438283,
+      "grad_norm": 0.969067394733429,
+      "learning_rate": 2.9018466968991913e-05,
+      "loss": 0.712,
+      "step": 225
+    },
+    {
+      "epoch": 0.8228980322003577,
+      "grad_norm": 0.8197952508926392,
+      "learning_rate": 2.8954459201100916e-05,
+      "loss": 0.7021,
+      "step": 230
+    },
+    {
+      "epoch": 0.8407871198568873,
+      "grad_norm": 0.9914915561676025,
+      "learning_rate": 2.8888504977912284e-05,
+      "loss": 0.7463,
+      "step": 235
+    },
+    {
+      "epoch": 0.8586762075134168,
+      "grad_norm": 1.0100288391113281,
+      "learning_rate": 2.8820613499133814e-05,
+      "loss": 0.7186,
+      "step": 240
+    },
+    {
+      "epoch": 0.8765652951699463,
+      "grad_norm": 0.9397743344306946,
+      "learning_rate": 2.875079423469384e-05,
+      "loss": 0.704,
+      "step": 245
+    },
+    {
+      "epoch": 0.8944543828264758,
+      "grad_norm": 1.116220235824585,
+      "learning_rate": 2.8679056923420294e-05,
+      "loss": 0.705,
+      "step": 250
+    },
+    {
+      "epoch": 0.9123434704830053,
+      "grad_norm": 0.9566684365272522,
+      "learning_rate": 2.8605411571682295e-05,
+      "loss": 0.6645,
+      "step": 255
+    },
+    {
+      "epoch": 0.9302325581395349,
+      "grad_norm": 0.884859561920166,
+      "learning_rate": 2.8529868451994387e-05,
+      "loss": 0.6604,
+      "step": 260
+    },
+    {
+      "epoch": 0.9481216457960644,
+      "grad_norm": 1.0066081285476685,
+      "learning_rate": 2.8452438101583648e-05,
+      "loss": 0.6933,
+      "step": 265
+    },
+    {
+      "epoch": 0.9660107334525939,
+      "grad_norm": 0.992435097694397,
+      "learning_rate": 2.8373131320919936e-05,
+      "loss": 0.6576,
+      "step": 270
+    },
+    {
+      "epoch": 0.9838998211091234,
+      "grad_norm": 1.1710649728775024,
+      "learning_rate": 2.8291959172209314e-05,
+      "loss": 0.6487,
+      "step": 275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4341217279434204,
+      "learning_rate": 2.820893297785107e-05,
+      "loss": 0.6293,
+      "step": 280
+    },
+    {
+      "epoch": 1.0178890876565294,
+      "grad_norm": 1.1913830041885376,
+      "learning_rate": 2.812406431885838e-05,
+      "loss": 0.5798,
+      "step": 285
+    },
+    {
+      "epoch": 1.035778175313059,
+      "grad_norm": 0.9532171487808228,
+      "learning_rate": 2.8037365033242917e-05,
+      "loss": 0.5752,
+      "step": 290
+    },
+    {
+      "epoch": 1.0536672629695885,
+      "grad_norm": 1.3926500082015991,
+      "learning_rate": 2.794884721436361e-05,
+      "loss": 0.6038,
+      "step": 295
+    },
+    {
+      "epoch": 1.071556350626118,
+      "grad_norm": 1.0542120933532715,
+      "learning_rate": 2.7858523209239785e-05,
+      "loss": 0.5491,
+      "step": 300
+    },
+    {
+      "epoch": 1.0894454382826475,
+      "grad_norm": 1.1710113286972046,
+      "learning_rate": 2.7766405616828938e-05,
+      "loss": 0.5183,
+      "step": 305
+    },
+    {
+      "epoch": 1.1073345259391771,
+      "grad_norm": 0.9291325211524963,
+      "learning_rate": 2.7672507286269332e-05,
+      "loss": 0.5193,
+      "step": 310
+    },
+    {
+      "epoch": 1.1252236135957066,
+      "grad_norm": 0.9673988223075867,
+      "learning_rate": 2.7576841315087744e-05,
+      "loss": 0.4956,
+      "step": 315
+    },
+    {
+      "epoch": 1.1431127012522362,
+      "grad_norm": 1.3483326435089111,
+      "learning_rate": 2.747942104737252e-05,
+      "loss": 0.5407,
+      "step": 320
+    },
+    {
+      "epoch": 1.1610017889087656,
+      "grad_norm": 0.9651196002960205,
+      "learning_rate": 2.738026007191226e-05,
+      "loss": 0.5261,
+      "step": 325
+    },
+    {
+      "epoch": 1.1788908765652952,
+      "grad_norm": 1.0171095132827759,
+      "learning_rate": 2.727937222030039e-05,
+      "loss": 0.4733,
+      "step": 330
+    },
+    {
+      "epoch": 1.1967799642218246,
+      "grad_norm": 1.0522226095199585,
+      "learning_rate": 2.7176771565005804e-05,
+      "loss": 0.496,
+      "step": 335
+    },
+    {
+      "epoch": 1.2146690518783543,
+      "grad_norm": 1.230334758758545,
+      "learning_rate": 2.7072472417410002e-05,
+      "loss": 0.5143,
+      "step": 340
+    },
+    {
+      "epoch": 1.2325581395348837,
+      "grad_norm": 1.0268306732177734,
+      "learning_rate": 2.6966489325810793e-05,
+      "loss": 0.5221,
+      "step": 345
+    },
+    {
+      "epoch": 1.250447227191413,
+      "grad_norm": 0.9560613632202148,
+      "learning_rate": 2.685883707339305e-05,
+      "loss": 0.5079,
+      "step": 350
+    },
+    {
+      "epoch": 1.2683363148479427,
+      "grad_norm": 1.0826609134674072,
+      "learning_rate": 2.6749530676166633e-05,
+      "loss": 0.4486,
+      "step": 355
+    },
+    {
+      "epoch": 1.2862254025044724,
+      "grad_norm": 1.1852221488952637,
+      "learning_rate": 2.663858538087188e-05,
+      "loss": 0.5147,
+      "step": 360
+    },
+    {
+      "epoch": 1.3041144901610018,
+      "grad_norm": 1.1893274784088135,
+      "learning_rate": 2.6526016662852887e-05,
+      "loss": 0.4878,
+      "step": 365
+    },
+    {
+      "epoch": 1.3220035778175312,
+      "grad_norm": 1.0597422122955322,
+      "learning_rate": 2.6411840223898902e-05,
+      "loss": 0.4802,
+      "step": 370
+    },
+    {
+      "epoch": 1.3398926654740608,
+      "grad_norm": 1.0407767295837402,
+      "learning_rate": 2.6296071990054167e-05,
+      "loss": 0.4368,
+      "step": 375
+    },
+    {
+      "epoch": 1.3577817531305905,
+      "grad_norm": 1.2436342239379883,
+      "learning_rate": 2.6178728109396413e-05,
+      "loss": 0.4877,
+      "step": 380
+    },
+    {
+      "epoch": 1.3756708407871199,
+      "grad_norm": 1.279242753982544,
+      "learning_rate": 2.6059824949784474e-05,
+      "loss": 0.4394,
+      "step": 385
+    },
+    {
+      "epoch": 1.3935599284436493,
+      "grad_norm": 1.060095191001892,
+      "learning_rate": 2.5939379096575156e-05,
+      "loss": 0.4371,
+      "step": 390
+    },
+    {
+      "epoch": 1.411449016100179,
+      "grad_norm": 1.1264623403549194,
+      "learning_rate": 2.5817407350309825e-05,
+      "loss": 0.4361,
+      "step": 395
+    },
+    {
+      "epoch": 1.4293381037567083,
+      "grad_norm": 1.0541681051254272,
+      "learning_rate": 2.5693926724370958e-05,
+      "loss": 0.4431,
+      "step": 400
+    },
+    {
+      "epoch": 1.447227191413238,
+      "grad_norm": 1.0380645990371704,
+      "learning_rate": 2.5568954442609016e-05,
+      "loss": 0.4025,
+      "step": 405
+    },
+    {
+      "epoch": 1.4651162790697674,
+      "grad_norm": 1.122382640838623,
+      "learning_rate": 2.544250793693995e-05,
+      "loss": 0.4379,
+      "step": 410
+    },
+    {
+      "epoch": 1.483005366726297,
+      "grad_norm": 1.0751020908355713,
+      "learning_rate": 2.531460484491368e-05,
+      "loss": 0.4328,
+      "step": 415
+    },
+    {
+      "epoch": 1.5008944543828264,
+      "grad_norm": 1.234891653060913,
+      "learning_rate": 2.5185263007253912e-05,
+      "loss": 0.4375,
+      "step": 420
+    },
+    {
+      "epoch": 1.518783542039356,
+      "grad_norm": 1.4189293384552002,
+      "learning_rate": 2.5054500465369597e-05,
+      "loss": 0.4321,
+      "step": 425
+    },
+    {
+      "epoch": 1.5366726296958855,
+      "grad_norm": 1.1898143291473389,
+      "learning_rate": 2.4922335458838397e-05,
+      "loss": 0.4428,
+      "step": 430
+    },
+    {
+      "epoch": 1.5545617173524149,
+      "grad_norm": 0.9852802753448486,
+      "learning_rate": 2.478878642286253e-05,
+      "loss": 0.4321,
+      "step": 435
+    },
+    {
+      "epoch": 1.5724508050089445,
+      "grad_norm": 1.0897939205169678,
+      "learning_rate": 2.465387198569729e-05,
+      "loss": 0.4173,
+      "step": 440
+    },
+    {
+      "epoch": 1.5903398926654742,
+      "grad_norm": 1.1859195232391357,
+      "learning_rate": 2.4517610966052682e-05,
+      "loss": 0.3929,
+      "step": 445
+    },
+    {
+      "epoch": 1.6082289803220036,
+      "grad_norm": 1.2279844284057617,
+      "learning_rate": 2.4380022370468464e-05,
+      "loss": 0.4279,
+      "step": 450
+    },
+    {
+      "epoch": 1.626118067978533,
+      "grad_norm": 1.1750357151031494,
+      "learning_rate": 2.4241125390662982e-05,
+      "loss": 0.3863,
+      "step": 455
+    },
+    {
+      "epoch": 1.6440071556350626,
+      "grad_norm": 1.1458582878112793,
+      "learning_rate": 2.4100939400856216e-05,
+      "loss": 0.4131,
+      "step": 460
+    },
+    {
+      "epoch": 1.6618962432915922,
+      "grad_norm": 1.0942723751068115,
+      "learning_rate": 2.395948395506731e-05,
+      "loss": 0.3943,
+      "step": 465
+    },
+    {
+      "epoch": 1.6797853309481217,
+      "grad_norm": 0.9955680966377258,
+      "learning_rate": 2.3816778784387097e-05,
+      "loss": 0.3774,
+      "step": 470
+    },
+    {
+      "epoch": 1.697674418604651,
+      "grad_norm": 1.2159160375595093,
+      "learning_rate": 2.367284379422584e-05,
+      "loss": 0.3674,
+      "step": 475
+    },
+    {
+      "epoch": 1.7155635062611807,
+      "grad_norm": 1.143467664718628,
+      "learning_rate": 2.3527699061536726e-05,
+      "loss": 0.3777,
+      "step": 480
+    },
+    {
+      "epoch": 1.7334525939177103,
+      "grad_norm": 1.0947456359863281,
+      "learning_rate": 2.338136483201539e-05,
+      "loss": 0.3947,
+      "step": 485
+    },
+    {
+      "epoch": 1.7513416815742398,
+      "grad_norm": 1.0965192317962646,
+      "learning_rate": 2.323386151727595e-05,
+      "loss": 0.3732,
+      "step": 490
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 1.1917636394500732,
+      "learning_rate": 2.3085209692003836e-05,
+      "loss": 0.369,
+      "step": 495
+    },
+    {
+      "epoch": 1.7871198568872988,
+      "grad_norm": 1.117609977722168,
+      "learning_rate": 2.2935430091085904e-05,
+      "loss": 0.3933,
+      "step": 500
+    },
+    {
+      "epoch": 1.8050089445438284,
+      "grad_norm": 1.2263697385787964,
+      "learning_rate": 2.2784543606718227e-05,
+      "loss": 0.4066,
+      "step": 505
+    },
+    {
+      "epoch": 1.8228980322003578,
+      "grad_norm": 1.1519149541854858,
+      "learning_rate": 2.263257128549191e-05,
+      "loss": 0.3776,
+      "step": 510
+    },
+    {
+      "epoch": 1.8407871198568873,
+      "grad_norm": 1.001869797706604,
+      "learning_rate": 2.2479534325457374e-05,
+      "loss": 0.3576,
+      "step": 515
+    },
+    {
+      "epoch": 1.8586762075134167,
+      "grad_norm": 1.0438801050186157,
+      "learning_rate": 2.2325454073167518e-05,
+      "loss": 0.3951,
+      "step": 520
+    },
+    {
+      "epoch": 1.8765652951699463,
+      "grad_norm": 1.1146963834762573,
+      "learning_rate": 2.2170352020700187e-05,
+      "loss": 0.3637,
+      "step": 525
+    },
+    {
+      "epoch": 1.894454382826476,
+      "grad_norm": 1.0773694515228271,
+      "learning_rate": 2.2014249802660297e-05,
+      "loss": 0.3596,
+      "step": 530
+    },
+    {
+      "epoch": 1.9123434704830053,
+      "grad_norm": 1.0742276906967163,
+      "learning_rate": 2.185716919316212e-05,
+      "loss": 0.3641,
+      "step": 535
+    },
+    {
+      "epoch": 1.9302325581395348,
+      "grad_norm": 1.118898630142212,
+      "learning_rate": 2.16991321027921e-05,
+      "loss": 0.3598,
+      "step": 540
+    },
+    {
+      "epoch": 1.9481216457960644,
+      "grad_norm": 1.1367541551589966,
+      "learning_rate": 2.1540160575552604e-05,
+      "loss": 0.3532,
+      "step": 545
+    },
+    {
+      "epoch": 1.966010733452594,
+      "grad_norm": 1.0174490213394165,
+      "learning_rate": 2.138027678578712e-05,
+      "loss": 0.3277,
+      "step": 550
+    },
+    {
+      "epoch": 1.9838998211091234,
+      "grad_norm": 1.1238727569580078,
+      "learning_rate": 2.1219503035087202e-05,
+      "loss": 0.302,
+      "step": 555
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.766844391822815,
+      "learning_rate": 2.1057861749181743e-05,
+      "loss": 0.3191,
+      "step": 560
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1400,
+  "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": 8.135566147023012e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

7/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e286434b66ce83f6a7b0cd5d442242ac85f0a68c6b34379235c2b0ade0da66d9
+size 7736

7/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)