Upload folder using huggingface_hub

base/29_128_e3_3e-5/README.md

@@ -0,0 +1,202 @@
+---
+base_model: ibm-granite/granite-3.3-8b-base
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.15.2

base/29_128_e3_3e-5/adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 256,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "k_proj",
+    "o_proj",
+    "q_proj",
+    "gate_proj",
+    "up_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

base/29_128_e3_3e-5/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c750d08898daa3b7dd2e6e5e345e4c2a84e3780357afff6d0a37fda3b8ef97e9
+size 791751704

base/29_128_e3_3e-5/latest

@@ -0,0 +1 @@
+global_step366

base/29_128_e3_3e-5/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81c78c2df4bd397fb446a3e4c7e99ed72d8a19c1a29d2b017188b66a5652bcc0
+size 16389

base/29_128_e3_3e-5/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a44c5816fca30392989c85520c3c1a5d22ee1a1c3e4d9899c8c9bbd96888d2d2
+size 16389

base/29_128_e3_3e-5/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:def9ed3d056d3973bacf563c8b97348904919526a47a2de57e7c647967e9554f
+size 16389

base/29_128_e3_3e-5/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b64175f691ac579351fed4cb41fd99fd6d3675710d56c07029d671008f16507
+size 16389

base/29_128_e3_3e-5/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21d161ec81f5af4719154de96eadaf33945b41f298d8aa63ed389d4ebcc4b318
+size 16389

base/29_128_e3_3e-5/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c03b2928be235963a139c8455db2e06c305fdde0ee04808e4e89f84f8993ffb
+size 16389

base/29_128_e3_3e-5/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:483193519bc5712396db63ff065cc6f4762ac3d1a38473e2c9dde54f99b85df1
+size 16389

base/29_128_e3_3e-5/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52156af4bca643330eac9f98623860b63f670715b56049fa688ba117ead1b46e
+size 16389

base/29_128_e3_3e-5/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:135d640e86d01191013ce2787ac8565de514e0eae03ce8945cc93d23efb20ed0
+size 1401

base/29_128_e3_3e-5/special_tokens_map.json

@@ -0,0 +1,45 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<reponame>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

base/29_128_e3_3e-5/tokenizer_config.json

@@ -0,0 +1,188 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<fim_prefix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<fim_middle>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<fim_suffix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<fim_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<filename>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "<gh_stars>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "7": {
+      "content": "<issue_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "8": {
+      "content": "<issue_comment>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "9": {
+      "content": "<issue_closed>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "10": {
+      "content": "<jupyter_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "11": {
+      "content": "<jupyter_text>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "12": {
+      "content": "<jupyter_code>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "13": {
+      "content": "<jupyter_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "14": {
+      "content": "<empty_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "15": {
+      "content": "<commit_before>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "16": {
+      "content": "<commit_msg>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "17": {
+      "content": "<commit_after>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "18": {
+      "content": "<reponame>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<reponame>",
+  "padding_side": "left",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

base/29_128_e3_3e-5/trainer_state.json

@@ -0,0 +1,545 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 366,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.040983606557377046,
+      "grad_norm": 1.868965744972229,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 1.1739,
+      "step": 5
+    },
+    {
+      "epoch": 0.08196721311475409,
+      "grad_norm": 2.0415163040161133,
+      "learning_rate": 1.4210526315789473e-05,
+      "loss": 1.1738,
+      "step": 10
+    },
+    {
+      "epoch": 0.12295081967213115,
+      "grad_norm": 1.2948756217956543,
+      "learning_rate": 2.2105263157894736e-05,
+      "loss": 1.0444,
+      "step": 15
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 0.9190149307250977,
+      "learning_rate": 3e-05,
+      "loss": 0.9243,
+      "step": 20
+    },
+    {
+      "epoch": 0.20491803278688525,
+      "grad_norm": 0.8775644302368164,
+      "learning_rate": 2.9984633749661812e-05,
+      "loss": 0.8608,
+      "step": 25
+    },
+    {
+      "epoch": 0.2459016393442623,
+      "grad_norm": 0.8844294548034668,
+      "learning_rate": 2.9938566481533845e-05,
+      "loss": 0.8358,
+      "step": 30
+    },
+    {
+      "epoch": 0.28688524590163933,
+      "grad_norm": 0.8944787979125977,
+      "learning_rate": 2.9861892579772688e-05,
+      "loss": 0.7618,
+      "step": 35
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 1.0438920259475708,
+      "learning_rate": 2.975476913642753e-05,
+      "loss": 0.6601,
+      "step": 40
+    },
+    {
+      "epoch": 0.36885245901639346,
+      "grad_norm": 0.8531023263931274,
+      "learning_rate": 2.9617415629584705e-05,
+      "loss": 0.5995,
+      "step": 45
+    },
+    {
+      "epoch": 0.4098360655737705,
+      "grad_norm": 0.9162529110908508,
+      "learning_rate": 2.9450113473693675e-05,
+      "loss": 0.5708,
+      "step": 50
+    },
+    {
+      "epoch": 0.45081967213114754,
+      "grad_norm": 1.0522689819335938,
+      "learning_rate": 2.925320544299571e-05,
+      "loss": 0.6,
+      "step": 55
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 1.0933209657669067,
+      "learning_rate": 2.9027094969236592e-05,
+      "loss": 0.4517,
+      "step": 60
+    },
+    {
+      "epoch": 0.5327868852459017,
+      "grad_norm": 1.015614628791809,
+      "learning_rate": 2.8772245315102164e-05,
+      "loss": 0.4539,
+      "step": 65
+    },
+    {
+      "epoch": 0.5737704918032787,
+      "grad_norm": 1.367930293083191,
+      "learning_rate": 2.8489178625070296e-05,
+      "loss": 0.4673,
+      "step": 70
+    },
+    {
+      "epoch": 0.6147540983606558,
+      "grad_norm": 1.2692198753356934,
+      "learning_rate": 2.817847485562385e-05,
+      "loss": 0.4439,
+      "step": 75
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 1.2608386278152466,
+      "learning_rate": 2.7840770587016466e-05,
+      "loss": 0.3984,
+      "step": 80
+    },
+    {
+      "epoch": 0.6967213114754098,
+      "grad_norm": 1.3799766302108765,
+      "learning_rate": 2.7476757719025706e-05,
+      "loss": 0.3598,
+      "step": 85
+    },
+    {
+      "epoch": 0.7377049180327869,
+      "grad_norm": 1.0914649963378906,
+      "learning_rate": 2.7087182053365686e-05,
+      "loss": 0.3581,
+      "step": 90
+    },
+    {
+      "epoch": 0.7786885245901639,
+      "grad_norm": 1.2499964237213135,
+      "learning_rate": 2.6672841765663627e-05,
+      "loss": 0.3118,
+      "step": 95
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 0.9599180221557617,
+      "learning_rate": 2.6234585770131002e-05,
+      "loss": 0.32,
+      "step": 100
+    },
+    {
+      "epoch": 0.860655737704918,
+      "grad_norm": 1.4201722145080566,
+      "learning_rate": 2.5773311980279764e-05,
+      "loss": 0.2952,
+      "step": 105
+    },
+    {
+      "epoch": 0.9016393442622951,
+      "grad_norm": 1.2638850212097168,
+      "learning_rate": 2.5289965469247136e-05,
+      "loss": 0.268,
+      "step": 110
+    },
+    {
+      "epoch": 0.9426229508196722,
+      "grad_norm": 1.5408021211624146,
+      "learning_rate": 2.4785536533498274e-05,
+      "loss": 0.2299,
+      "step": 115
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 0.9972408413887024,
+      "learning_rate": 2.4261058663873784e-05,
+      "loss": 0.2262,
+      "step": 120
+    },
+    {
+      "epoch": 1.0245901639344261,
+      "grad_norm": 0.9083095192909241,
+      "learning_rate": 2.37176064281392e-05,
+      "loss": 0.19,
+      "step": 125
+    },
+    {
+      "epoch": 1.0655737704918034,
+      "grad_norm": 0.8827641010284424,
+      "learning_rate": 2.3156293269374673e-05,
+      "loss": 0.1752,
+      "step": 130
+    },
+    {
+      "epoch": 1.1065573770491803,
+      "grad_norm": 0.9249515533447266,
+      "learning_rate": 2.2578269224715627e-05,
+      "loss": 0.1619,
+      "step": 135
+    },
+    {
+      "epoch": 1.1475409836065573,
+      "grad_norm": 1.3332761526107788,
+      "learning_rate": 2.1984718569118292e-05,
+      "loss": 0.1559,
+      "step": 140
+    },
+    {
+      "epoch": 1.1885245901639343,
+      "grad_norm": 1.1348826885223389,
+      "learning_rate": 2.1376857388977642e-05,
+      "loss": 0.1666,
+      "step": 145
+    },
+    {
+      "epoch": 1.2295081967213115,
+      "grad_norm": 0.9556555151939392,
+      "learning_rate": 2.0755931090568998e-05,
+      "loss": 0.1249,
+      "step": 150
+    },
+    {
+      "epoch": 1.2704918032786885,
+      "grad_norm": 0.948033332824707,
+      "learning_rate": 2.0123211848418087e-05,
+      "loss": 0.1374,
+      "step": 155
+    },
+    {
+      "epoch": 1.3114754098360657,
+      "grad_norm": 1.4530562162399292,
+      "learning_rate": 1.9479995998827394e-05,
+      "loss": 0.1319,
+      "step": 160
+    },
+    {
+      "epoch": 1.3524590163934427,
+      "grad_norm": 1.0472363233566284,
+      "learning_rate": 1.882760138389909e-05,
+      "loss": 0.1195,
+      "step": 165
+    },
+    {
+      "epoch": 1.3934426229508197,
+      "grad_norm": 1.064152717590332,
+      "learning_rate": 1.816736465149615e-05,
+      "loss": 0.1204,
+      "step": 170
+    },
+    {
+      "epoch": 1.4344262295081966,
+      "grad_norm": 1.2208365201950073,
+      "learning_rate": 1.7500638516673578e-05,
+      "loss": 0.1296,
+      "step": 175
+    },
+    {
+      "epoch": 1.4754098360655736,
+      "grad_norm": 1.073548436164856,
+      "learning_rate": 1.6828788990190675e-05,
+      "loss": 0.105,
+      "step": 180
+    },
+    {
+      "epoch": 1.5163934426229508,
+      "grad_norm": 0.9216892719268799,
+      "learning_rate": 1.6153192579782568e-05,
+      "loss": 0.1157,
+      "step": 185
+    },
+    {
+      "epoch": 1.5573770491803278,
+      "grad_norm": 1.0411626100540161,
+      "learning_rate": 1.5475233469925254e-05,
+      "loss": 0.0947,
+      "step": 190
+    },
+    {
+      "epoch": 1.598360655737705,
+      "grad_norm": 0.9014426469802856,
+      "learning_rate": 1.4796300685872216e-05,
+      "loss": 0.0815,
+      "step": 195
+    },
+    {
+      "epoch": 1.639344262295082,
+      "grad_norm": 0.9626762866973877,
+      "learning_rate": 1.4117785247773118e-05,
+      "loss": 0.0755,
+      "step": 200
+    },
+    {
+      "epoch": 1.680327868852459,
+      "grad_norm": 1.0149961709976196,
+      "learning_rate": 1.3441077320705323e-05,
+      "loss": 0.0919,
+      "step": 205
+    },
+    {
+      "epoch": 1.721311475409836,
+      "grad_norm": 0.6428225040435791,
+      "learning_rate": 1.2767563366457258e-05,
+      "loss": 0.0714,
+      "step": 210
+    },
+    {
+      "epoch": 1.762295081967213,
+      "grad_norm": 0.91527259349823,
+      "learning_rate": 1.2098623302899206e-05,
+      "loss": 0.0636,
+      "step": 215
+    },
+    {
+      "epoch": 1.8032786885245902,
+      "grad_norm": 0.7494262456893921,
+      "learning_rate": 1.1435627676761559e-05,
+      "loss": 0.0676,
+      "step": 220
+    },
+    {
+      "epoch": 1.8442622950819674,
+      "grad_norm": 0.8445521593093872,
+      "learning_rate": 1.0779934855612898e-05,
+      "loss": 0.0667,
+      "step": 225
+    },
+    {
+      "epoch": 1.8852459016393444,
+      "grad_norm": 0.6250991225242615,
+      "learning_rate": 1.0132888244791187e-05,
+      "loss": 0.0621,
+      "step": 230
+    },
+    {
+      "epoch": 1.9262295081967213,
+      "grad_norm": 0.9637157917022705,
+      "learning_rate": 9.495813534990069e-06,
+      "loss": 0.0613,
+      "step": 235
+    },
+    {
+      "epoch": 1.9672131147540983,
+      "grad_norm": 0.8252452611923218,
+      "learning_rate": 8.870015986139541e-06,
+      "loss": 0.0551,
+      "step": 240
+    },
+    {
+      "epoch": 2.0081967213114753,
+      "grad_norm": 0.5627564787864685,
+      "learning_rate": 8.256777753145819e-06,
+      "loss": 0.0532,
+      "step": 245
+    },
+    {
+      "epoch": 2.0491803278688523,
+      "grad_norm": 0.711893618106842,
+      "learning_rate": 7.657355258969594e-06,
+      "loss": 0.0511,
+      "step": 250
+    },
+    {
+      "epoch": 2.0901639344262297,
+      "grad_norm": 0.6451098322868347,
+      "learning_rate": 7.072976620424707e-06,
+      "loss": 0.0411,
+      "step": 255
+    },
+    {
+      "epoch": 2.1311475409836067,
+      "grad_norm": 0.5286330580711365,
+      "learning_rate": 6.5048391319714466e-06,
+      "loss": 0.044,
+      "step": 260
+    },
+    {
+      "epoch": 2.1721311475409837,
+      "grad_norm": 0.5646039247512817,
+      "learning_rate": 5.954106812659697e-06,
+      "loss": 0.0398,
+      "step": 265
+    },
+    {
+      "epoch": 2.2131147540983607,
+      "grad_norm": 0.6554955840110779,
+      "learning_rate": 5.421908021247834e-06,
+      "loss": 0.0396,
+      "step": 270
+    },
+    {
+      "epoch": 2.2540983606557377,
+      "grad_norm": 0.5206019282341003,
+      "learning_rate": 4.9093331443836616e-06,
+      "loss": 0.0436,
+      "step": 275
+    },
+    {
+      "epoch": 2.2950819672131146,
+      "grad_norm": 0.39405640959739685,
+      "learning_rate": 4.417432362583843e-06,
+      "loss": 0.0364,
+      "step": 280
+    },
+    {
+      "epoch": 2.3360655737704916,
+      "grad_norm": 0.6238886713981628,
+      "learning_rate": 3.947213498589008e-06,
+      "loss": 0.0409,
+      "step": 285
+    },
+    {
+      "epoch": 2.3770491803278686,
+      "grad_norm": 0.47585681080818176,
+      "learning_rate": 3.4996399525028627e-06,
+      "loss": 0.0423,
+      "step": 290
+    },
+    {
+      "epoch": 2.418032786885246,
+      "grad_norm": 0.6819049715995789,
+      "learning_rate": 3.0756287279459354e-06,
+      "loss": 0.0345,
+      "step": 295
+    },
+    {
+      "epoch": 2.459016393442623,
+      "grad_norm": 0.4804767072200775,
+      "learning_rate": 2.6760485532679247e-06,
+      "loss": 0.0343,
+      "step": 300
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6551392078399658,
+      "learning_rate": 2.3017181016680704e-06,
+      "loss": 0.0382,
+      "step": 305
+    },
+    {
+      "epoch": 2.540983606557377,
+      "grad_norm": 0.5764215588569641,
+      "learning_rate": 1.9534043138701634e-06,
+      "loss": 0.0288,
+      "step": 310
+    },
+    {
+      "epoch": 2.581967213114754,
+      "grad_norm": 0.560716450214386,
+      "learning_rate": 1.6318208267888157e-06,
+      "loss": 0.0412,
+      "step": 315
+    },
+    {
+      "epoch": 2.6229508196721314,
+      "grad_norm": 0.446825236082077,
+      "learning_rate": 1.3376265114063086e-06,
+      "loss": 0.0404,
+      "step": 320
+    },
+    {
+      "epoch": 2.663934426229508,
+      "grad_norm": 0.44851160049438477,
+      "learning_rate": 1.0714241228557426e-06,
+      "loss": 0.0289,
+      "step": 325
+    },
+    {
+      "epoch": 2.7049180327868854,
+      "grad_norm": 0.37918326258659363,
+      "learning_rate": 8.337590654761928e-07,
+      "loss": 0.0337,
+      "step": 330
+    },
+    {
+      "epoch": 2.7459016393442623,
+      "grad_norm": 0.42671921849250793,
+      "learning_rate": 6.25118275370104e-07,
+      "loss": 0.039,
+      "step": 335
+    },
+    {
+      "epoch": 2.7868852459016393,
+      "grad_norm": 0.2966732978820801,
+      "learning_rate": 4.459292227523498e-07,
+      "loss": 0.0258,
+      "step": 340
+    },
+    {
+      "epoch": 2.8278688524590163,
+      "grad_norm": 0.43605688214302063,
+      "learning_rate": 2.9655903613497957e-07,
+      "loss": 0.0242,
+      "step": 345
+    },
+    {
+      "epoch": 2.8688524590163933,
+      "grad_norm": 0.42670899629592896,
+      "learning_rate": 1.7731375014207574e-07,
+      "loss": 0.0257,
+      "step": 350
+    },
+    {
+      "epoch": 2.9098360655737707,
+      "grad_norm": 0.3986416757106781,
+      "learning_rate": 8.843767849580465e-08,
+      "loss": 0.0293,
+      "step": 355
+    },
+    {
+      "epoch": 2.9508196721311473,
+      "grad_norm": 0.576945960521698,
+      "learning_rate": 3.0112913458293365e-08,
+      "loss": 0.0441,
+      "step": 360
+    },
+    {
+      "epoch": 2.9918032786885247,
+      "grad_norm": 0.4272902309894562,
+      "learning_rate": 2.45895275494179e-09,
+      "loss": 0.0308,
+      "step": 365
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 366,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.483131539277742e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

base/29_128_e3_3e-5/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:285e2f89584936a60857b93d191a0aa6c544ef81a8dc834f12fb2f8cb10f9ac5
+size 8273

base/29_128_e3_3e-5/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)