Upload folder using huggingface_hub

README.md

@@ -0,0 +1,202 @@
+---
+base_model: codellama/CodeLlama-7b-hf
+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 0.14.0

adapter_config.json

@@ -0,0 +1,15 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "codellama/CodeLlama-7b-hf",
+  "encoder_hidden_size": 4096,
+  "inference_mode": true,
+  "num_attention_heads": 32,
+  "num_layers": 32,
+  "num_transformer_submodules": 1,
+  "num_virtual_tokens": 20,
+  "peft_type": "PREFIX_TUNING",
+  "prefix_projection": true,
+  "revision": null,
+  "task_type": "CAUSAL_LM",
+  "token_dim": 4096
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16bc5a7b19cfae6ebe631d3ec5a64df1f4092a9bd27334ebe7c1c7736249a098
+size 10485888

checkpoint-1484/README.md

@@ -0,0 +1,202 @@
+---
+base_model: codellama/CodeLlama-7b-hf
+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 0.14.0

checkpoint-1484/adapter_config.json

@@ -0,0 +1,15 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "codellama/CodeLlama-7b-hf",
+  "encoder_hidden_size": 4096,
+  "inference_mode": true,
+  "num_attention_heads": 32,
+  "num_layers": 32,
+  "num_transformer_submodules": 1,
+  "num_virtual_tokens": 20,
+  "peft_type": "PREFIX_TUNING",
+  "prefix_projection": true,
+  "revision": null,
+  "task_type": "CAUSAL_LM",
+  "token_dim": 4096
+}

checkpoint-1484/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16bc5a7b19cfae6ebe631d3ec5a64df1f4092a9bd27334ebe7c1c7736249a098
+size 10485888

checkpoint-1484/global_step1484/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7220a5d92557d0df3d04687fe04e717890d840a9d8c39b734ca57afd81d3b8c
+size 3272607024

checkpoint-1484/global_step1484/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:210a9e497b1bb3a1178457df8659ae0332ab337379a320f7f2042047b6b31883
+size 3272606832

checkpoint-1484/global_step1484/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1aaf307245004251c97cdef5a1f4f6dd57854027f3e3be294611716af5484149
+size 3272606832

checkpoint-1484/global_step1484/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:795d96a805714a3f7d30002b8a122d57a6112ba45b90e081b3897f91ea798751
+size 3272606832

checkpoint-1484/global_step1484/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01291585acfd90e533c0c86cbb3e4e2e2cd6cfc2ecad8fa646d5f82bb4e8f485
+size 2444036868

checkpoint-1484/latest

@@ -0,0 +1 @@
+global_step1484

checkpoint-1484/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0463657eb196db696840f1d09b3adadac520fe49264add5251aeb9addf055fc
+size 15024

checkpoint-1484/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ffe320ec96f869bb8919b462c34efb3292a69842530bead39d0d3b341234bd7
+size 15024

checkpoint-1484/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6596162c350ad095f608ad419bda73c8a08939691194e48e6a039213c5867009
+size 15024

checkpoint-1484/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:612d8065c6426c751f372c28c3f621bde42e79eaeade12ce126c4b79412e74e9
+size 15024

checkpoint-1484/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48a72f1f07c0dc1b2913fffde3ec12d88bfb3e89442dc52ba087138ee562d340
+size 1064

checkpoint-1484/trainer_state.json

@@ -0,0 +1,1076 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997473684210526,
+  "eval_steps": 500,
+  "global_step": 1484,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006736842105263158,
+      "grad_norm": 3.6967623233795166,
+      "learning_rate": 1.3422818791946309e-06,
+      "loss": 2.4093,
+      "step": 1
+    },
+    {
+      "epoch": 0.006736842105263158,
+      "grad_norm": 2.5490193367004395,
+      "learning_rate": 1.3422818791946309e-05,
+      "loss": 2.4939,
+      "step": 10
+    },
+    {
+      "epoch": 0.013473684210526317,
+      "grad_norm": 0.18483224511146545,
+      "learning_rate": 2.6845637583892618e-05,
+      "loss": 1.1877,
+      "step": 20
+    },
+    {
+      "epoch": 0.020210526315789474,
+      "grad_norm": 0.2031693309545517,
+      "learning_rate": 4.026845637583892e-05,
+      "loss": 0.8909,
+      "step": 30
+    },
+    {
+      "epoch": 0.026947368421052633,
+      "grad_norm": 0.6876732707023621,
+      "learning_rate": 5.3691275167785237e-05,
+      "loss": 0.7581,
+      "step": 40
+    },
+    {
+      "epoch": 0.03368421052631579,
+      "grad_norm": 0.09247241914272308,
+      "learning_rate": 6.711409395973155e-05,
+      "loss": 0.7594,
+      "step": 50
+    },
+    {
+      "epoch": 0.04042105263157895,
+      "grad_norm": 0.1324968934059143,
+      "learning_rate": 8.053691275167784e-05,
+      "loss": 0.7405,
+      "step": 60
+    },
+    {
+      "epoch": 0.04715789473684211,
+      "grad_norm": 0.05673883110284805,
+      "learning_rate": 9.395973154362417e-05,
+      "loss": 0.7065,
+      "step": 70
+    },
+    {
+      "epoch": 0.053894736842105266,
+      "grad_norm": 0.04617280140519142,
+      "learning_rate": 0.00010738255033557047,
+      "loss": 0.6817,
+      "step": 80
+    },
+    {
+      "epoch": 0.06063157894736842,
+      "grad_norm": 0.04381496459245682,
+      "learning_rate": 0.0001208053691275168,
+      "loss": 0.6789,
+      "step": 90
+    },
+    {
+      "epoch": 0.06736842105263158,
+      "grad_norm": 0.07428538799285889,
+      "learning_rate": 0.0001342281879194631,
+      "loss": 0.6816,
+      "step": 100
+    },
+    {
+      "epoch": 0.07410526315789474,
+      "grad_norm": 0.04249708354473114,
+      "learning_rate": 0.00014765100671140942,
+      "loss": 0.6997,
+      "step": 110
+    },
+    {
+      "epoch": 0.0808421052631579,
+      "grad_norm": 0.05957937240600586,
+      "learning_rate": 0.0001610738255033557,
+      "loss": 0.6807,
+      "step": 120
+    },
+    {
+      "epoch": 0.08757894736842105,
+      "grad_norm": 0.03975442424416542,
+      "learning_rate": 0.000174496644295302,
+      "loss": 0.6733,
+      "step": 130
+    },
+    {
+      "epoch": 0.09431578947368421,
+      "grad_norm": 0.04079463332891464,
+      "learning_rate": 0.00018791946308724833,
+      "loss": 0.6556,
+      "step": 140
+    },
+    {
+      "epoch": 0.10105263157894737,
+      "grad_norm": 0.04245497286319733,
+      "learning_rate": 0.00019985018726591762,
+      "loss": 0.6575,
+      "step": 150
+    },
+    {
+      "epoch": 0.10778947368421053,
+      "grad_norm": 0.09695123136043549,
+      "learning_rate": 0.00019835205992509364,
+      "loss": 0.6916,
+      "step": 160
+    },
+    {
+      "epoch": 0.11452631578947368,
+      "grad_norm": 0.03505201265215874,
+      "learning_rate": 0.00019685393258426966,
+      "loss": 0.6622,
+      "step": 170
+    },
+    {
+      "epoch": 0.12126315789473684,
+      "grad_norm": 0.02820334956049919,
+      "learning_rate": 0.0001953558052434457,
+      "loss": 0.6497,
+      "step": 180
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.04135354235768318,
+      "learning_rate": 0.00019385767790262173,
+      "loss": 0.6671,
+      "step": 190
+    },
+    {
+      "epoch": 0.13473684210526315,
+      "grad_norm": 0.031461067497730255,
+      "learning_rate": 0.00019235955056179775,
+      "loss": 0.657,
+      "step": 200
+    },
+    {
+      "epoch": 0.1414736842105263,
+      "grad_norm": 0.04208710789680481,
+      "learning_rate": 0.0001908614232209738,
+      "loss": 0.6766,
+      "step": 210
+    },
+    {
+      "epoch": 0.1482105263157895,
+      "grad_norm": 3.495147705078125,
+      "learning_rate": 0.00018936329588014982,
+      "loss": 3.9378,
+      "step": 220
+    },
+    {
+      "epoch": 0.15494736842105264,
+      "grad_norm": 0.18893112242221832,
+      "learning_rate": 0.00018786516853932586,
+      "loss": 7.1374,
+      "step": 230
+    },
+    {
+      "epoch": 0.1616842105263158,
+      "grad_norm": 0.0959916040301323,
+      "learning_rate": 0.00018636704119850189,
+      "loss": 5.8104,
+      "step": 240
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 0.08286964148283005,
+      "learning_rate": 0.0001848689138576779,
+      "loss": 4.7292,
+      "step": 250
+    },
+    {
+      "epoch": 0.1751578947368421,
+      "grad_norm": 0.04510454833507538,
+      "learning_rate": 0.00018337078651685393,
+      "loss": 4.9858,
+      "step": 260
+    },
+    {
+      "epoch": 0.18189473684210528,
+      "grad_norm": 0.2256896197795868,
+      "learning_rate": 0.00018187265917602997,
+      "loss": 4.7463,
+      "step": 270
+    },
+    {
+      "epoch": 0.18863157894736843,
+      "grad_norm": 0.06342379748821259,
+      "learning_rate": 0.00018037453183520602,
+      "loss": 4.517,
+      "step": 280
+    },
+    {
+      "epoch": 0.19536842105263158,
+      "grad_norm": 0.07497289776802063,
+      "learning_rate": 0.00017887640449438204,
+      "loss": 4.4052,
+      "step": 290
+    },
+    {
+      "epoch": 0.20210526315789473,
+      "grad_norm": 0.08952877670526505,
+      "learning_rate": 0.00017737827715355806,
+      "loss": 3.9614,
+      "step": 300
+    },
+    {
+      "epoch": 0.20884210526315788,
+      "grad_norm": 0.044066932052373886,
+      "learning_rate": 0.00017588014981273408,
+      "loss": 4.5861,
+      "step": 310
+    },
+    {
+      "epoch": 0.21557894736842106,
+      "grad_norm": 0.08251778781414032,
+      "learning_rate": 0.0001743820224719101,
+      "loss": 4.5163,
+      "step": 320
+    },
+    {
+      "epoch": 0.22231578947368422,
+      "grad_norm": 0.04723803699016571,
+      "learning_rate": 0.00017288389513108615,
+      "loss": 4.1904,
+      "step": 330
+    },
+    {
+      "epoch": 0.22905263157894737,
+      "grad_norm": 0.09082615375518799,
+      "learning_rate": 0.0001713857677902622,
+      "loss": 4.1982,
+      "step": 340
+    },
+    {
+      "epoch": 0.23578947368421052,
+      "grad_norm": 0.04866361245512962,
+      "learning_rate": 0.00016988764044943822,
+      "loss": 3.8506,
+      "step": 350
+    },
+    {
+      "epoch": 0.24252631578947367,
+      "grad_norm": 0.04515402019023895,
+      "learning_rate": 0.00016838951310861424,
+      "loss": 4.4254,
+      "step": 360
+    },
+    {
+      "epoch": 0.24926315789473685,
+      "grad_norm": 0.14205284416675568,
+      "learning_rate": 0.00016689138576779026,
+      "loss": 4.4111,
+      "step": 370
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.16082021594047546,
+      "learning_rate": 0.0001653932584269663,
+      "loss": 4.1119,
+      "step": 380
+    },
+    {
+      "epoch": 0.26273684210526316,
+      "grad_norm": 0.061411116272211075,
+      "learning_rate": 0.00016389513108614235,
+      "loss": 4.059,
+      "step": 390
+    },
+    {
+      "epoch": 0.2694736842105263,
+      "grad_norm": 0.058379318565130234,
+      "learning_rate": 0.00016239700374531837,
+      "loss": 3.7307,
+      "step": 400
+    },
+    {
+      "epoch": 0.27621052631578946,
+      "grad_norm": 0.048859789967536926,
+      "learning_rate": 0.0001608988764044944,
+      "loss": 4.3039,
+      "step": 410
+    },
+    {
+      "epoch": 0.2829473684210526,
+      "grad_norm": 0.06003361940383911,
+      "learning_rate": 0.0001594007490636704,
+      "loss": 4.2032,
+      "step": 420
+    },
+    {
+      "epoch": 0.28968421052631577,
+      "grad_norm": 0.10120591521263123,
+      "learning_rate": 0.00015790262172284646,
+      "loss": 3.9567,
+      "step": 430
+    },
+    {
+      "epoch": 0.296421052631579,
+      "grad_norm": 0.21033401787281036,
+      "learning_rate": 0.00015640449438202248,
+      "loss": 3.9369,
+      "step": 440
+    },
+    {
+      "epoch": 0.3031578947368421,
+      "grad_norm": 0.06378967314958572,
+      "learning_rate": 0.00015490636704119852,
+      "loss": 3.6318,
+      "step": 450
+    },
+    {
+      "epoch": 0.3098947368421053,
+      "grad_norm": 0.042198359966278076,
+      "learning_rate": 0.00015340823970037455,
+      "loss": 4.1789,
+      "step": 460
+    },
+    {
+      "epoch": 0.31663157894736843,
+      "grad_norm": 0.053648848086595535,
+      "learning_rate": 0.00015191011235955057,
+      "loss": 4.1562,
+      "step": 470
+    },
+    {
+      "epoch": 0.3233684210526316,
+      "grad_norm": 0.0808805301785469,
+      "learning_rate": 0.00015041198501872659,
+      "loss": 3.8883,
+      "step": 480
+    },
+    {
+      "epoch": 0.33010526315789473,
+      "grad_norm": 0.13895294070243835,
+      "learning_rate": 0.00014891385767790263,
+      "loss": 3.9055,
+      "step": 490
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 0.11999215185642242,
+      "learning_rate": 0.00014741573033707865,
+      "loss": 3.6025,
+      "step": 500
+    },
+    {
+      "epoch": 0.34357894736842104,
+      "grad_norm": 0.0969998687505722,
+      "learning_rate": 0.0001459176029962547,
+      "loss": 4.2401,
+      "step": 510
+    },
+    {
+      "epoch": 0.3503157894736842,
+      "grad_norm": 0.2578948438167572,
+      "learning_rate": 0.00014441947565543072,
+      "loss": 4.1355,
+      "step": 520
+    },
+    {
+      "epoch": 0.35705263157894734,
+      "grad_norm": 0.067634217441082,
+      "learning_rate": 0.00014292134831460674,
+      "loss": 3.8735,
+      "step": 530
+    },
+    {
+      "epoch": 0.36378947368421055,
+      "grad_norm": 0.1961352676153183,
+      "learning_rate": 0.0001414232209737828,
+      "loss": 3.7641,
+      "step": 540
+    },
+    {
+      "epoch": 0.3705263157894737,
+      "grad_norm": 0.07940343767404556,
+      "learning_rate": 0.0001399250936329588,
+      "loss": 3.5177,
+      "step": 550
+    },
+    {
+      "epoch": 0.37726315789473686,
+      "grad_norm": 1.3029491901397705,
+      "learning_rate": 0.00013842696629213483,
+      "loss": 4.1854,
+      "step": 560
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.10544762760400772,
+      "learning_rate": 0.00013692883895131088,
+      "loss": 4.3064,
+      "step": 570
+    },
+    {
+      "epoch": 0.39073684210526316,
+      "grad_norm": 0.150394469499588,
+      "learning_rate": 0.0001354307116104869,
+      "loss": 3.9517,
+      "step": 580
+    },
+    {
+      "epoch": 0.3974736842105263,
+      "grad_norm": 0.06921563297510147,
+      "learning_rate": 0.00013393258426966294,
+      "loss": 3.8917,
+      "step": 590
+    },
+    {
+      "epoch": 0.40421052631578946,
+      "grad_norm": 0.06402010470628738,
+      "learning_rate": 0.00013243445692883896,
+      "loss": 3.5635,
+      "step": 600
+    },
+    {
+      "epoch": 0.4109473684210526,
+      "grad_norm": 0.08918313682079315,
+      "learning_rate": 0.00013093632958801498,
+      "loss": 4.1197,
+      "step": 610
+    },
+    {
+      "epoch": 0.41768421052631577,
+      "grad_norm": 0.054397523403167725,
+      "learning_rate": 0.000129438202247191,
+      "loss": 4.0442,
+      "step": 620
+    },
+    {
+      "epoch": 0.4244210526315789,
+      "grad_norm": 0.068702831864357,
+      "learning_rate": 0.00012794007490636705,
+      "loss": 3.7506,
+      "step": 630
+    },
+    {
+      "epoch": 0.43115789473684213,
+      "grad_norm": 0.14575353264808655,
+      "learning_rate": 0.0001264419475655431,
+      "loss": 3.7359,
+      "step": 640
+    },
+    {
+      "epoch": 0.4378947368421053,
+      "grad_norm": 0.1481335461139679,
+      "learning_rate": 0.00012494382022471912,
+      "loss": 3.3705,
+      "step": 650
+    },
+    {
+      "epoch": 0.44463157894736843,
+      "grad_norm": 0.06438197940587997,
+      "learning_rate": 0.00012344569288389514,
+      "loss": 4.0248,
+      "step": 660
+    },
+    {
+      "epoch": 0.4513684210526316,
+      "grad_norm": 0.38855019211769104,
+      "learning_rate": 0.00012194756554307116,
+      "loss": 4.0265,
+      "step": 670
+    },
+    {
+      "epoch": 0.45810526315789474,
+      "grad_norm": 0.20793034136295319,
+      "learning_rate": 0.00012044943820224719,
+      "loss": 3.7305,
+      "step": 680
+    },
+    {
+      "epoch": 0.4648421052631579,
+      "grad_norm": 0.11011853814125061,
+      "learning_rate": 0.00011895131086142324,
+      "loss": 3.6933,
+      "step": 690
+    },
+    {
+      "epoch": 0.47157894736842104,
+      "grad_norm": 0.06795340031385422,
+      "learning_rate": 0.00011745318352059926,
+      "loss": 3.3734,
+      "step": 700
+    },
+    {
+      "epoch": 0.4783157894736842,
+      "grad_norm": 0.07788679003715515,
+      "learning_rate": 0.00011595505617977529,
+      "loss": 3.9053,
+      "step": 710
+    },
+    {
+      "epoch": 0.48505263157894735,
+      "grad_norm": 0.07339611649513245,
+      "learning_rate": 0.00011445692883895131,
+      "loss": 3.8685,
+      "step": 720
+    },
+    {
+      "epoch": 0.4917894736842105,
+      "grad_norm": 0.16048288345336914,
+      "learning_rate": 0.00011295880149812735,
+      "loss": 3.5673,
+      "step": 730
+    },
+    {
+      "epoch": 0.4985263157894737,
+      "grad_norm": 0.2596355974674225,
+      "learning_rate": 0.00011146067415730337,
+      "loss": 3.5684,
+      "step": 740
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 0.10115884989500046,
+      "learning_rate": 0.00010996254681647941,
+      "loss": 3.2226,
+      "step": 750
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.13997367024421692,
+      "learning_rate": 0.00010846441947565545,
+      "loss": 3.8579,
+      "step": 760
+    },
+    {
+      "epoch": 0.5187368421052632,
+      "grad_norm": 0.08359155058860779,
+      "learning_rate": 0.00010696629213483147,
+      "loss": 3.8313,
+      "step": 770
+    },
+    {
+      "epoch": 0.5254736842105263,
+      "grad_norm": 0.2407791018486023,
+      "learning_rate": 0.0001054681647940075,
+      "loss": 3.5257,
+      "step": 780
+    },
+    {
+      "epoch": 0.5322105263157895,
+      "grad_norm": 0.34615418314933777,
+      "learning_rate": 0.00010397003745318352,
+      "loss": 3.5113,
+      "step": 790
+    },
+    {
+      "epoch": 0.5389473684210526,
+      "grad_norm": 0.06987264007329941,
+      "learning_rate": 0.00010247191011235954,
+      "loss": 3.1525,
+      "step": 800
+    },
+    {
+      "epoch": 0.5456842105263158,
+      "grad_norm": 0.07933894544839859,
+      "learning_rate": 0.00010097378277153558,
+      "loss": 3.718,
+      "step": 810
+    },
+    {
+      "epoch": 0.5524210526315789,
+      "grad_norm": 0.12424171715974808,
+      "learning_rate": 9.947565543071161e-05,
+      "loss": 3.6641,
+      "step": 820
+    },
+    {
+      "epoch": 0.5591578947368421,
+      "grad_norm": 0.2515564262866974,
+      "learning_rate": 9.797752808988764e-05,
+      "loss": 3.4268,
+      "step": 830
+    },
+    {
+      "epoch": 0.5658947368421052,
+      "grad_norm": 0.30851560831069946,
+      "learning_rate": 9.647940074906368e-05,
+      "loss": 3.3856,
+      "step": 840
+    },
+    {
+      "epoch": 0.5726315789473684,
+      "grad_norm": 0.05149822682142258,
+      "learning_rate": 9.49812734082397e-05,
+      "loss": 3.1259,
+      "step": 850
+    },
+    {
+      "epoch": 0.5793684210526315,
+      "grad_norm": 0.17960771918296814,
+      "learning_rate": 9.348314606741574e-05,
+      "loss": 3.6767,
+      "step": 860
+    },
+    {
+      "epoch": 0.5861052631578947,
+      "grad_norm": 0.17523854970932007,
+      "learning_rate": 9.198501872659176e-05,
+      "loss": 3.5995,
+      "step": 870
+    },
+    {
+      "epoch": 0.592842105263158,
+      "grad_norm": 0.3186163008213043,
+      "learning_rate": 9.04868913857678e-05,
+      "loss": 3.3966,
+      "step": 880
+    },
+    {
+      "epoch": 0.5995789473684211,
+      "grad_norm": 0.21263690292835236,
+      "learning_rate": 8.898876404494383e-05,
+      "loss": 3.3526,
+      "step": 890
+    },
+    {
+      "epoch": 0.6063157894736843,
+      "grad_norm": 0.10399254411458969,
+      "learning_rate": 8.749063670411985e-05,
+      "loss": 3.0519,
+      "step": 900
+    },
+    {
+      "epoch": 0.6130526315789474,
+      "grad_norm": 0.13143524527549744,
+      "learning_rate": 8.599250936329589e-05,
+      "loss": 3.629,
+      "step": 910
+    },
+    {
+      "epoch": 0.6197894736842106,
+      "grad_norm": 0.15374666452407837,
+      "learning_rate": 8.449438202247192e-05,
+      "loss": 3.6895,
+      "step": 920
+    },
+    {
+      "epoch": 0.6265263157894737,
+      "grad_norm": 0.23757484555244446,
+      "learning_rate": 8.299625468164794e-05,
+      "loss": 3.3622,
+      "step": 930
+    },
+    {
+      "epoch": 0.6332631578947369,
+      "grad_norm": 0.1661984622478485,
+      "learning_rate": 8.149812734082397e-05,
+      "loss": 3.3248,
+      "step": 940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08603614568710327,
+      "learning_rate": 8e-05,
+      "loss": 3.0086,
+      "step": 950
+    },
+    {
+      "epoch": 0.6467368421052632,
+      "grad_norm": 0.07694745808839798,
+      "learning_rate": 7.850187265917604e-05,
+      "loss": 3.5162,
+      "step": 960
+    },
+    {
+      "epoch": 0.6534736842105263,
+      "grad_norm": 0.16395558416843414,
+      "learning_rate": 7.700374531835206e-05,
+      "loss": 3.4812,
+      "step": 970
+    },
+    {
+      "epoch": 0.6602105263157895,
+      "grad_norm": 0.13817398250102997,
+      "learning_rate": 7.55056179775281e-05,
+      "loss": 3.2516,
+      "step": 980
+    },
+    {
+      "epoch": 0.6669473684210526,
+      "grad_norm": 0.25807198882102966,
+      "learning_rate": 7.400749063670413e-05,
+      "loss": 3.2101,
+      "step": 990
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 0.06848172843456268,
+      "learning_rate": 7.250936329588015e-05,
+      "loss": 2.93,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6804210526315789,
+      "grad_norm": 1.089575171470642,
+      "learning_rate": 7.101123595505618e-05,
+      "loss": 3.4925,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6871578947368421,
+      "grad_norm": 0.20126965641975403,
+      "learning_rate": 6.951310861423222e-05,
+      "loss": 3.4603,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6938947368421052,
+      "grad_norm": 0.21779027581214905,
+      "learning_rate": 6.801498127340824e-05,
+      "loss": 3.1723,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7006315789473684,
+      "grad_norm": 0.18239159882068634,
+      "learning_rate": 6.651685393258428e-05,
+      "loss": 3.1903,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7073684210526315,
+      "grad_norm": 0.06677573919296265,
+      "learning_rate": 6.50187265917603e-05,
+      "loss": 2.8445,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7141052631578947,
+      "grad_norm": 0.42619746923446655,
+      "learning_rate": 6.352059925093634e-05,
+      "loss": 3.4319,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7208421052631578,
+      "grad_norm": 0.12023507058620453,
+      "learning_rate": 6.202247191011237e-05,
+      "loss": 3.3826,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7275789473684211,
+      "grad_norm": 0.15099403262138367,
+      "learning_rate": 6.052434456928839e-05,
+      "loss": 3.1425,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7343157894736843,
+      "grad_norm": 0.3474717438220978,
+      "learning_rate": 5.902621722846442e-05,
+      "loss": 3.1279,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7410526315789474,
+      "grad_norm": 0.12225649505853653,
+      "learning_rate": 5.752808988764046e-05,
+      "loss": 2.9033,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7477894736842106,
+      "grad_norm": 0.19639068841934204,
+      "learning_rate": 5.6029962546816485e-05,
+      "loss": 3.3681,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7545263157894737,
+      "grad_norm": 0.10571427643299103,
+      "learning_rate": 5.453183520599251e-05,
+      "loss": 3.335,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7612631578947369,
+      "grad_norm": 0.5154901146888733,
+      "learning_rate": 5.3033707865168545e-05,
+      "loss": 3.0952,
+      "step": 1130
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.6122628450393677,
+      "learning_rate": 5.153558052434457e-05,
+      "loss": 3.1269,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7747368421052632,
+      "grad_norm": 0.19698569178581238,
+      "learning_rate": 5.00374531835206e-05,
+      "loss": 2.8233,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7814736842105263,
+      "grad_norm": 0.13018374145030975,
+      "learning_rate": 4.853932584269663e-05,
+      "loss": 3.3094,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7882105263157895,
+      "grad_norm": 0.09522128850221634,
+      "learning_rate": 4.704119850187266e-05,
+      "loss": 3.2765,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7949473684210526,
+      "grad_norm": 0.10098107159137726,
+      "learning_rate": 4.554307116104869e-05,
+      "loss": 3.0807,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8016842105263158,
+      "grad_norm": 0.18019132316112518,
+      "learning_rate": 4.404494382022472e-05,
+      "loss": 3.0332,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8084210526315789,
+      "grad_norm": 0.16289708018302917,
+      "learning_rate": 4.2546816479400754e-05,
+      "loss": 2.7374,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8151578947368421,
+      "grad_norm": 0.12666673958301544,
+      "learning_rate": 4.104868913857678e-05,
+      "loss": 3.2118,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8218947368421052,
+      "grad_norm": 0.16891352832317352,
+      "learning_rate": 3.955056179775281e-05,
+      "loss": 3.1902,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8286315789473684,
+      "grad_norm": 0.10958009213209152,
+      "learning_rate": 3.805243445692884e-05,
+      "loss": 2.9862,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8353684210526315,
+      "grad_norm": 0.10642745345830917,
+      "learning_rate": 3.655430711610487e-05,
+      "loss": 3.0052,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.05656813085079193,
+      "learning_rate": 3.50561797752809e-05,
+      "loss": 2.723,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8488421052631578,
+      "grad_norm": 0.08322717994451523,
+      "learning_rate": 3.355805243445693e-05,
+      "loss": 3.234,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8555789473684211,
+      "grad_norm": 0.13246551156044006,
+      "learning_rate": 3.2059925093632956e-05,
+      "loss": 3.212,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8623157894736843,
+      "grad_norm": 0.10225304961204529,
+      "learning_rate": 3.056179775280899e-05,
+      "loss": 2.9484,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8690526315789474,
+      "grad_norm": 0.19440552592277527,
+      "learning_rate": 2.9063670411985024e-05,
+      "loss": 2.9266,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8757894736842106,
+      "grad_norm": 0.08913037180900574,
+      "learning_rate": 2.7565543071161047e-05,
+      "loss": 2.6801,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8825263157894737,
+      "grad_norm": 0.10815408080816269,
+      "learning_rate": 2.606741573033708e-05,
+      "loss": 3.1505,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8892631578947369,
+      "grad_norm": 0.14371147751808167,
+      "learning_rate": 2.4569288389513108e-05,
+      "loss": 3.1293,
+      "step": 1320
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.1680973470211029,
+      "learning_rate": 2.3071161048689138e-05,
+      "loss": 2.8961,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9027368421052632,
+      "grad_norm": 0.19012019038200378,
+      "learning_rate": 2.157303370786517e-05,
+      "loss": 2.9096,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9094736842105263,
+      "grad_norm": 0.060957688838243484,
+      "learning_rate": 2.00749063670412e-05,
+      "loss": 2.6879,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9162105263157895,
+      "grad_norm": 0.15055014193058014,
+      "learning_rate": 1.857677902621723e-05,
+      "loss": 3.108,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9229473684210526,
+      "grad_norm": 0.1378874033689499,
+      "learning_rate": 1.707865168539326e-05,
+      "loss": 3.0428,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9296842105263158,
+      "grad_norm": 0.14901022613048553,
+      "learning_rate": 1.558052434456929e-05,
+      "loss": 2.8589,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9364210526315789,
+      "grad_norm": 0.17515867948532104,
+      "learning_rate": 1.4082397003745318e-05,
+      "loss": 2.8563,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9431578947368421,
+      "grad_norm": 0.11909812688827515,
+      "learning_rate": 1.258426966292135e-05,
+      "loss": 2.5759,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9498947368421052,
+      "grad_norm": 0.16348549723625183,
+      "learning_rate": 1.1086142322097379e-05,
+      "loss": 3.089,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9566315789473684,
+      "grad_norm": 0.08107765763998032,
+      "learning_rate": 9.588014981273409e-06,
+      "loss": 3.0145,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9633684210526315,
+      "grad_norm": 0.13251617550849915,
+      "learning_rate": 8.089887640449438e-06,
+      "loss": 2.8256,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9701052631578947,
+      "grad_norm": 0.10319063812494278,
+      "learning_rate": 6.591760299625469e-06,
+      "loss": 2.8456,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9768421052631578,
+      "grad_norm": 0.08950542658567429,
+      "learning_rate": 5.093632958801498e-06,
+      "loss": 2.605,
+      "step": 1450
+    },
+    {
+      "epoch": 0.983578947368421,
+      "grad_norm": 0.08379487693309784,
+      "learning_rate": 3.5955056179775286e-06,
+      "loss": 3.0334,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9903157894736843,
+      "grad_norm": 0.1561821848154068,
+      "learning_rate": 2.097378277153558e-06,
+      "loss": 3.0357,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9970526315789474,
+      "grad_norm": 0.07574011385440826,
+      "learning_rate": 5.992509363295881e-07,
+      "loss": 2.7458,
+      "step": 1480
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1484,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.32780044727799e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1484/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28da48262e760c5feb1c1e1f08d029374a06fa8d076c01d714f0912a196ad9d7
+size 6904

checkpoint-1484/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/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 . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+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, weights_only=False)
+
+        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 tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # 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}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] 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")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+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 = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * 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
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        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}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        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 to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=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
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    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.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    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")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               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_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``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
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+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_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    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_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

logs/events.out.tfevents.1738806841.apolo.2066103.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd129cec2e7b2768185ecf810b2a439324462aa63fa860020502e7dd4cbc0815
+size 8328

logs/events.out.tfevents.1738813546.apolo.2108330.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74b32b9d52814ce91e481a2b4c7676a0ae9fe5f03c5507ded89a5ca3b06e6ae0
+size 37585

special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "additional_special_tokens": [
+    "▁<PRE>",
+    "▁<MID>",
+    "▁<SUF>",
+    "▁<EOT>"
+  ],
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45ccb9c8b6b561889acea59191d66986d314e7cbd6a78abc6e49b139ca91c1e6
+size 500058

tokenizer_config.json

@@ -0,0 +1,85 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32007": {
+      "content": "▁<PRE>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32008": {
+      "content": "▁<SUF>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32009": {
+      "content": "▁<MID>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32010": {
+      "content": "▁<EOT>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "▁<PRE>",
+    "▁<MID>",
+    "▁<SUF>",
+    "▁<EOT>"
+  ],
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "eot_token": "▁<EOT>",
+  "extra_special_tokens": {},
+  "fill_token": "<FILL_ME>",
+  "legacy": null,
+  "middle_token": "▁<MID>",
+  "model_max_length": 2048,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "prefix_token": "▁<PRE>",
+  "sp_model_kwargs": {},
+  "suffix_token": "▁<SUF>",
+  "tokenizer_class": "CodeLlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}