Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,202 @@
+---
+base_model: mistralai/Mistral-Nemo-Instruct-2407
+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 0.15.1

adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-Nemo-Instruct-2407",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.1,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08d19f934ca1a14a01857cfb4940605cec545447de6279dbf541e011c4e94c6e
+size 228141160

global_step599/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f38ac9c77df5534f00d084b7400bf12168c3774dcaa7d09beb48e74ee1e0d46
+size 1368456848

global_step599/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d0fccad359da002ccc73788d7110556776b2c13fdae46572b1e022df98984bc
+size 402429729

latest

@@ -0,0 +1 @@
+global_step599

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4587d199845342f90878f4dc0e7dbce541f83908f1a4d8e2e8f87afa3831c475
+size 14244

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d88891425b6da9fdf531ec7fe7e0434c0cb51007e555d2cc2480e16d7c4600c
+size 1064

special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "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": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0240ce510f08e6c2041724e9043e33be9d251d1e4a4d94eb68cd47b954b61d2
+size 17078292

trainer_state.json

@@ -0,0 +1,874 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0309544282029235,
+  "eval_steps": 500,
+  "global_step": 600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008598452278589854,
+      "grad_norm": 0.5736560821533203,
+      "learning_rate": 5.517241379310345e-08,
+      "loss": 1.4983,
+      "step": 5
+    },
+    {
+      "epoch": 0.017196904557179708,
+      "grad_norm": 0.47119423747062683,
+      "learning_rate": 1.2413793103448275e-07,
+      "loss": 1.2594,
+      "step": 10
+    },
+    {
+      "epoch": 0.025795356835769563,
+      "grad_norm": 0.6366227865219116,
+      "learning_rate": 1.9310344827586205e-07,
+      "loss": 1.5001,
+      "step": 15
+    },
+    {
+      "epoch": 0.034393809114359415,
+      "grad_norm": 0.5710499882698059,
+      "learning_rate": 2.620689655172414e-07,
+      "loss": 1.5602,
+      "step": 20
+    },
+    {
+      "epoch": 0.04299226139294927,
+      "grad_norm": 0.4520650804042816,
+      "learning_rate": 3.310344827586207e-07,
+      "loss": 1.4316,
+      "step": 25
+    },
+    {
+      "epoch": 0.051590713671539126,
+      "grad_norm": 0.5874906778335571,
+      "learning_rate": 4e-07,
+      "loss": 1.5158,
+      "step": 30
+    },
+    {
+      "epoch": 0.06018916595012898,
+      "grad_norm": 0.4618767201900482,
+      "learning_rate": 4.689655172413793e-07,
+      "loss": 1.3945,
+      "step": 35
+    },
+    {
+      "epoch": 0.06878761822871883,
+      "grad_norm": 0.47305673360824585,
+      "learning_rate": 5.379310344827586e-07,
+      "loss": 1.4154,
+      "step": 40
+    },
+    {
+      "epoch": 0.07738607050730868,
+      "grad_norm": 0.5665937066078186,
+      "learning_rate": 6.068965517241379e-07,
+      "loss": 1.3776,
+      "step": 45
+    },
+    {
+      "epoch": 0.08598452278589853,
+      "grad_norm": 0.5359976291656494,
+      "learning_rate": 6.758620689655172e-07,
+      "loss": 1.5018,
+      "step": 50
+    },
+    {
+      "epoch": 0.09458297506448839,
+      "grad_norm": 0.5621670484542847,
+      "learning_rate": 7.448275862068965e-07,
+      "loss": 1.4493,
+      "step": 55
+    },
+    {
+      "epoch": 0.10318142734307825,
+      "grad_norm": 0.5842466354370117,
+      "learning_rate": 8.137931034482758e-07,
+      "loss": 1.4241,
+      "step": 60
+    },
+    {
+      "epoch": 0.1117798796216681,
+      "grad_norm": 0.6147713661193848,
+      "learning_rate": 8.827586206896551e-07,
+      "loss": 1.4748,
+      "step": 65
+    },
+    {
+      "epoch": 0.12037833190025796,
+      "grad_norm": 0.6653420925140381,
+      "learning_rate": 9.517241379310345e-07,
+      "loss": 1.6576,
+      "step": 70
+    },
+    {
+      "epoch": 0.1289767841788478,
+      "grad_norm": 0.5247882604598999,
+      "learning_rate": 1.0206896551724139e-06,
+      "loss": 1.439,
+      "step": 75
+    },
+    {
+      "epoch": 0.13757523645743766,
+      "grad_norm": 0.6043932437896729,
+      "learning_rate": 1.089655172413793e-06,
+      "loss": 1.475,
+      "step": 80
+    },
+    {
+      "epoch": 0.1461736887360275,
+      "grad_norm": 0.614253580570221,
+      "learning_rate": 1.1586206896551724e-06,
+      "loss": 1.4928,
+      "step": 85
+    },
+    {
+      "epoch": 0.15477214101461736,
+      "grad_norm": 0.6523554921150208,
+      "learning_rate": 1.2275862068965516e-06,
+      "loss": 1.5111,
+      "step": 90
+    },
+    {
+      "epoch": 0.16337059329320722,
+      "grad_norm": 0.4596118628978729,
+      "learning_rate": 1.2965517241379309e-06,
+      "loss": 1.4672,
+      "step": 95
+    },
+    {
+      "epoch": 0.17196904557179707,
+      "grad_norm": 0.5627196431159973,
+      "learning_rate": 1.3655172413793103e-06,
+      "loss": 1.4621,
+      "step": 100
+    },
+    {
+      "epoch": 0.18056749785038692,
+      "grad_norm": 0.5366098284721375,
+      "learning_rate": 1.4344827586206896e-06,
+      "loss": 1.3875,
+      "step": 105
+    },
+    {
+      "epoch": 0.18916595012897677,
+      "grad_norm": 0.48968908190727234,
+      "learning_rate": 1.5034482758620688e-06,
+      "loss": 1.4264,
+      "step": 110
+    },
+    {
+      "epoch": 0.19776440240756663,
+      "grad_norm": 0.5644566416740417,
+      "learning_rate": 1.5724137931034483e-06,
+      "loss": 1.3822,
+      "step": 115
+    },
+    {
+      "epoch": 0.2063628546861565,
+      "grad_norm": 0.49979278445243835,
+      "learning_rate": 1.6413793103448275e-06,
+      "loss": 1.4866,
+      "step": 120
+    },
+    {
+      "epoch": 0.21496130696474636,
+      "grad_norm": 0.5352433919906616,
+      "learning_rate": 1.710344827586207e-06,
+      "loss": 1.3047,
+      "step": 125
+    },
+    {
+      "epoch": 0.2235597592433362,
+      "grad_norm": 0.5073248744010925,
+      "learning_rate": 1.7793103448275862e-06,
+      "loss": 1.4763,
+      "step": 130
+    },
+    {
+      "epoch": 0.23215821152192606,
+      "grad_norm": 0.48227864503860474,
+      "learning_rate": 1.8482758620689653e-06,
+      "loss": 1.4552,
+      "step": 135
+    },
+    {
+      "epoch": 0.2407566638005159,
+      "grad_norm": 0.4420427978038788,
+      "learning_rate": 1.9172413793103447e-06,
+      "loss": 1.3285,
+      "step": 140
+    },
+    {
+      "epoch": 0.24935511607910577,
+      "grad_norm": 0.4224660396575928,
+      "learning_rate": 1.986206896551724e-06,
+      "loss": 1.4294,
+      "step": 145
+    },
+    {
+      "epoch": 0.2579535683576956,
+      "grad_norm": 0.3935889005661011,
+      "learning_rate": 1.999989634963924e-06,
+      "loss": 1.32,
+      "step": 150
+    },
+    {
+      "epoch": 0.26655202063628547,
+      "grad_norm": 0.352851539850235,
+      "learning_rate": 1.9999475273731217e-06,
+      "loss": 1.326,
+      "step": 155
+    },
+    {
+      "epoch": 0.2751504729148753,
+      "grad_norm": 0.4401000142097473,
+      "learning_rate": 1.9998730307756826e-06,
+      "loss": 1.352,
+      "step": 160
+    },
+    {
+      "epoch": 0.2837489251934652,
+      "grad_norm": 0.3836155831813812,
+      "learning_rate": 1.9997661475846052e-06,
+      "loss": 1.3492,
+      "step": 165
+    },
+    {
+      "epoch": 0.292347377472055,
+      "grad_norm": 0.32301604747772217,
+      "learning_rate": 1.9996268812619105e-06,
+      "loss": 1.309,
+      "step": 170
+    },
+    {
+      "epoch": 0.3009458297506449,
+      "grad_norm": 0.37157368659973145,
+      "learning_rate": 1.999455236318534e-06,
+      "loss": 1.4444,
+      "step": 175
+    },
+    {
+      "epoch": 0.30954428202923473,
+      "grad_norm": 0.340994268655777,
+      "learning_rate": 1.999251218314176e-06,
+      "loss": 1.2391,
+      "step": 180
+    },
+    {
+      "epoch": 0.3181427343078246,
+      "grad_norm": 0.3156416714191437,
+      "learning_rate": 1.999014833857124e-06,
+      "loss": 1.2496,
+      "step": 185
+    },
+    {
+      "epoch": 0.32674118658641443,
+      "grad_norm": 0.35868847370147705,
+      "learning_rate": 1.998746090604037e-06,
+      "loss": 1.3442,
+      "step": 190
+    },
+    {
+      "epoch": 0.3353396388650043,
+      "grad_norm": 0.31525614857673645,
+      "learning_rate": 1.9984449972597e-06,
+      "loss": 1.3011,
+      "step": 195
+    },
+    {
+      "epoch": 0.34393809114359414,
+      "grad_norm": 0.47816550731658936,
+      "learning_rate": 1.9981115635767376e-06,
+      "loss": 1.2962,
+      "step": 200
+    },
+    {
+      "epoch": 0.352536543422184,
+      "grad_norm": 0.31586989760398865,
+      "learning_rate": 1.9977458003553037e-06,
+      "loss": 1.2353,
+      "step": 205
+    },
+    {
+      "epoch": 0.36113499570077384,
+      "grad_norm": 0.3226100206375122,
+      "learning_rate": 1.9973477194427262e-06,
+      "loss": 1.2135,
+      "step": 210
+    },
+    {
+      "epoch": 0.3697334479793637,
+      "grad_norm": 0.3332583010196686,
+      "learning_rate": 1.996917333733128e-06,
+      "loss": 1.2553,
+      "step": 215
+    },
+    {
+      "epoch": 0.37833190025795355,
+      "grad_norm": 0.33345943689346313,
+      "learning_rate": 1.9964546571670053e-06,
+      "loss": 1.2652,
+      "step": 220
+    },
+    {
+      "epoch": 0.3869303525365434,
+      "grad_norm": 0.37145930528640747,
+      "learning_rate": 1.995959704730779e-06,
+      "loss": 1.1999,
+      "step": 225
+    },
+    {
+      "epoch": 0.39552880481513325,
+      "grad_norm": 0.41526567935943604,
+      "learning_rate": 1.9954324924563086e-06,
+      "loss": 1.3373,
+      "step": 230
+    },
+    {
+      "epoch": 0.4041272570937231,
+      "grad_norm": 0.2987145781517029,
+      "learning_rate": 1.9948730374203715e-06,
+      "loss": 1.2249,
+      "step": 235
+    },
+    {
+      "epoch": 0.412725709372313,
+      "grad_norm": 0.49898993968963623,
+      "learning_rate": 1.994281357744112e-06,
+      "loss": 1.2323,
+      "step": 240
+    },
+    {
+      "epoch": 0.42132416165090286,
+      "grad_norm": 0.2862403392791748,
+      "learning_rate": 1.9936574725924525e-06,
+      "loss": 1.1908,
+      "step": 245
+    },
+    {
+      "epoch": 0.4299226139294927,
+      "grad_norm": 0.31127598881721497,
+      "learning_rate": 1.9930014021734732e-06,
+      "loss": 1.3326,
+      "step": 250
+    },
+    {
+      "epoch": 0.43852106620808257,
+      "grad_norm": 0.44075530767440796,
+      "learning_rate": 1.9923131677377585e-06,
+      "loss": 1.1721,
+      "step": 255
+    },
+    {
+      "epoch": 0.4471195184866724,
+      "grad_norm": 0.29328110814094543,
+      "learning_rate": 1.991592791577708e-06,
+      "loss": 1.2735,
+      "step": 260
+    },
+    {
+      "epoch": 0.45571797076526227,
+      "grad_norm": 0.3364109396934509,
+      "learning_rate": 1.9908402970268145e-06,
+      "loss": 1.1244,
+      "step": 265
+    },
+    {
+      "epoch": 0.4643164230438521,
+      "grad_norm": 0.3128204047679901,
+      "learning_rate": 1.9900557084589077e-06,
+      "loss": 1.2281,
+      "step": 270
+    },
+    {
+      "epoch": 0.472914875322442,
+      "grad_norm": 0.30848929286003113,
+      "learning_rate": 1.989239051287366e-06,
+      "loss": 1.2333,
+      "step": 275
+    },
+    {
+      "epoch": 0.4815133276010318,
+      "grad_norm": 0.36058536171913147,
+      "learning_rate": 1.988390351964291e-06,
+      "loss": 1.2995,
+      "step": 280
+    },
+    {
+      "epoch": 0.4901117798796217,
+      "grad_norm": 0.3388177752494812,
+      "learning_rate": 1.9875096379796535e-06,
+      "loss": 1.3073,
+      "step": 285
+    },
+    {
+      "epoch": 0.49871023215821153,
+      "grad_norm": 0.3674142062664032,
+      "learning_rate": 1.986596937860402e-06,
+      "loss": 1.2673,
+      "step": 290
+    },
+    {
+      "epoch": 0.5073086844368013,
+      "grad_norm": 0.30132225155830383,
+      "learning_rate": 1.9856522811695374e-06,
+      "loss": 1.2107,
+      "step": 295
+    },
+    {
+      "epoch": 0.5159071367153912,
+      "grad_norm": 0.3170187473297119,
+      "learning_rate": 1.9846756985051573e-06,
+      "loss": 1.1624,
+      "step": 300
+    },
+    {
+      "epoch": 0.524505588993981,
+      "grad_norm": 0.3753427565097809,
+      "learning_rate": 1.9836672214994637e-06,
+      "loss": 1.2782,
+      "step": 305
+    },
+    {
+      "epoch": 0.5331040412725709,
+      "grad_norm": 0.36705997586250305,
+      "learning_rate": 1.9826268828177393e-06,
+      "loss": 1.134,
+      "step": 310
+    },
+    {
+      "epoch": 0.5417024935511608,
+      "grad_norm": 0.3330920338630676,
+      "learning_rate": 1.9815547161572892e-06,
+      "loss": 1.1611,
+      "step": 315
+    },
+    {
+      "epoch": 0.5503009458297506,
+      "grad_norm": 0.3084378242492676,
+      "learning_rate": 1.980450756246348e-06,
+      "loss": 1.2008,
+      "step": 320
+    },
+    {
+      "epoch": 0.5588993981083406,
+      "grad_norm": 0.3422742486000061,
+      "learning_rate": 1.979315038842957e-06,
+      "loss": 1.136,
+      "step": 325
+    },
+    {
+      "epoch": 0.5674978503869303,
+      "grad_norm": 0.4076729118824005,
+      "learning_rate": 1.9781476007338054e-06,
+      "loss": 1.179,
+      "step": 330
+    },
+    {
+      "epoch": 0.5760963026655203,
+      "grad_norm": 0.38809648156166077,
+      "learning_rate": 1.976948479733038e-06,
+      "loss": 1.1762,
+      "step": 335
+    },
+    {
+      "epoch": 0.58469475494411,
+      "grad_norm": 0.38286837935447693,
+      "learning_rate": 1.9757177146810307e-06,
+      "loss": 1.0976,
+      "step": 340
+    },
+    {
+      "epoch": 0.5932932072227,
+      "grad_norm": 0.33073902130126953,
+      "learning_rate": 1.9744553454431325e-06,
+      "loss": 1.1775,
+      "step": 345
+    },
+    {
+      "epoch": 0.6018916595012898,
+      "grad_norm": 0.30669665336608887,
+      "learning_rate": 1.9731614129083753e-06,
+      "loss": 1.1242,
+      "step": 350
+    },
+    {
+      "epoch": 0.6104901117798797,
+      "grad_norm": 0.36842820048332214,
+      "learning_rate": 1.9718359589881475e-06,
+      "loss": 1.233,
+      "step": 355
+    },
+    {
+      "epoch": 0.6190885640584695,
+      "grad_norm": 0.28934866189956665,
+      "learning_rate": 1.970479026614837e-06,
+      "loss": 1.0545,
+      "step": 360
+    },
+    {
+      "epoch": 0.6276870163370594,
+      "grad_norm": 0.36205312609672546,
+      "learning_rate": 1.9690906597404428e-06,
+      "loss": 1.0926,
+      "step": 365
+    },
+    {
+      "epoch": 0.6362854686156492,
+      "grad_norm": 0.39489033818244934,
+      "learning_rate": 1.967670903335148e-06,
+      "loss": 1.2244,
+      "step": 370
+    },
+    {
+      "epoch": 0.6448839208942391,
+      "grad_norm": 0.29848822951316833,
+      "learning_rate": 1.966219803385865e-06,
+      "loss": 1.1829,
+      "step": 375
+    },
+    {
+      "epoch": 0.6534823731728289,
+      "grad_norm": 0.31506502628326416,
+      "learning_rate": 1.9647374068947467e-06,
+      "loss": 1.1418,
+      "step": 380
+    },
+    {
+      "epoch": 0.6620808254514188,
+      "grad_norm": 0.39251357316970825,
+      "learning_rate": 1.963223761877662e-06,
+      "loss": 1.1543,
+      "step": 385
+    },
+    {
+      "epoch": 0.6706792777300086,
+      "grad_norm": 0.3380272686481476,
+      "learning_rate": 1.9616789173626418e-06,
+      "loss": 1.0856,
+      "step": 390
+    },
+    {
+      "epoch": 0.6792777300085985,
+      "grad_norm": 0.379609614610672,
+      "learning_rate": 1.960102923388291e-06,
+      "loss": 1.1562,
+      "step": 395
+    },
+    {
+      "epoch": 0.6878761822871883,
+      "grad_norm": 0.26444530487060547,
+      "learning_rate": 1.958495831002168e-06,
+      "loss": 1.1356,
+      "step": 400
+    },
+    {
+      "epoch": 0.6964746345657782,
+      "grad_norm": 0.4068126082420349,
+      "learning_rate": 1.9568576922591304e-06,
+      "loss": 1.2286,
+      "step": 405
+    },
+    {
+      "epoch": 0.705073086844368,
+      "grad_norm": 0.3288695514202118,
+      "learning_rate": 1.955188560219648e-06,
+      "loss": 1.206,
+      "step": 410
+    },
+    {
+      "epoch": 0.7136715391229579,
+      "grad_norm": 0.33392417430877686,
+      "learning_rate": 1.9534884889480876e-06,
+      "loss": 1.1944,
+      "step": 415
+    },
+    {
+      "epoch": 0.7222699914015477,
+      "grad_norm": 0.37305575609207153,
+      "learning_rate": 1.951757533510957e-06,
+      "loss": 1.204,
+      "step": 420
+    },
+    {
+      "epoch": 0.7308684436801376,
+      "grad_norm": 0.3695738911628723,
+      "learning_rate": 1.949995749975127e-06,
+      "loss": 1.1728,
+      "step": 425
+    },
+    {
+      "epoch": 0.7394668959587274,
+      "grad_norm": 0.37770697474479675,
+      "learning_rate": 1.948203195406009e-06,
+      "loss": 1.2171,
+      "step": 430
+    },
+    {
+      "epoch": 0.7480653482373173,
+      "grad_norm": 0.38213881850242615,
+      "learning_rate": 1.9463799278657124e-06,
+      "loss": 1.1948,
+      "step": 435
+    },
+    {
+      "epoch": 0.7566638005159071,
+      "grad_norm": 0.4096551239490509,
+      "learning_rate": 1.9445260064111607e-06,
+      "loss": 1.1774,
+      "step": 440
+    },
+    {
+      "epoch": 0.765262252794497,
+      "grad_norm": 0.41796615719795227,
+      "learning_rate": 1.9426414910921785e-06,
+      "loss": 1.1234,
+      "step": 445
+    },
+    {
+      "epoch": 0.7738607050730868,
+      "grad_norm": 0.3439182639122009,
+      "learning_rate": 1.9407264429495484e-06,
+      "loss": 1.1748,
+      "step": 450
+    },
+    {
+      "epoch": 0.7824591573516767,
+      "grad_norm": 0.34333643317222595,
+      "learning_rate": 1.938780924013032e-06,
+      "loss": 1.1441,
+      "step": 455
+    },
+    {
+      "epoch": 0.7910576096302665,
+      "grad_norm": 0.32014700770378113,
+      "learning_rate": 1.936804997299362e-06,
+      "loss": 1.1205,
+      "step": 460
+    },
+    {
+      "epoch": 0.7996560619088564,
+      "grad_norm": 0.3878689408302307,
+      "learning_rate": 1.9347987268101996e-06,
+      "loss": 1.1397,
+      "step": 465
+    },
+    {
+      "epoch": 0.8082545141874462,
+      "grad_norm": 0.3905511796474457,
+      "learning_rate": 1.9327621775300633e-06,
+      "loss": 1.1016,
+      "step": 470
+    },
+    {
+      "epoch": 0.8168529664660361,
+      "grad_norm": 0.3861243426799774,
+      "learning_rate": 1.9306954154242233e-06,
+      "loss": 1.0783,
+      "step": 475
+    },
+    {
+      "epoch": 0.825451418744626,
+      "grad_norm": 0.3234269618988037,
+      "learning_rate": 1.9285985074365627e-06,
+      "loss": 1.0694,
+      "step": 480
+    },
+    {
+      "epoch": 0.8340498710232158,
+      "grad_norm": 0.39914825558662415,
+      "learning_rate": 1.926471521487413e-06,
+      "loss": 1.2053,
+      "step": 485
+    },
+    {
+      "epoch": 0.8426483233018057,
+      "grad_norm": 0.39884087443351746,
+      "learning_rate": 1.924314526471351e-06,
+      "loss": 1.1663,
+      "step": 490
+    },
+    {
+      "epoch": 0.8512467755803955,
+      "grad_norm": 0.3793065547943115,
+      "learning_rate": 1.922127592254968e-06,
+      "loss": 1.083,
+      "step": 495
+    },
+    {
+      "epoch": 0.8598452278589854,
+      "grad_norm": 0.3194003403186798,
+      "learning_rate": 1.919910789674609e-06,
+      "loss": 1.0664,
+      "step": 500
+    },
+    {
+      "epoch": 0.8684436801375752,
+      "grad_norm": 0.30485859513282776,
+      "learning_rate": 1.917664190534075e-06,
+      "loss": 1.0426,
+      "step": 505
+    },
+    {
+      "epoch": 0.8770421324161651,
+      "grad_norm": 0.3151833117008209,
+      "learning_rate": 1.915387867602298e-06,
+      "loss": 1.157,
+      "step": 510
+    },
+    {
+      "epoch": 0.8856405846947549,
+      "grad_norm": 0.41206279397010803,
+      "learning_rate": 1.913081894610986e-06,
+      "loss": 1.1125,
+      "step": 515
+    },
+    {
+      "epoch": 0.8942390369733448,
+      "grad_norm": 0.42712071537971497,
+      "learning_rate": 1.9107463462522332e-06,
+      "loss": 0.9867,
+      "step": 520
+    },
+    {
+      "epoch": 0.9028374892519346,
+      "grad_norm": 0.45993658900260925,
+      "learning_rate": 1.9083812981760998e-06,
+      "loss": 1.2464,
+      "step": 525
+    },
+    {
+      "epoch": 0.9114359415305245,
+      "grad_norm": 0.34718605875968933,
+      "learning_rate": 1.9059868269881636e-06,
+      "loss": 1.0646,
+      "step": 530
+    },
+    {
+      "epoch": 0.9200343938091143,
+      "grad_norm": 0.4059743285179138,
+      "learning_rate": 1.9035630102470375e-06,
+      "loss": 1.08,
+      "step": 535
+    },
+    {
+      "epoch": 0.9286328460877042,
+      "grad_norm": 0.34420260787010193,
+      "learning_rate": 1.9011099264618573e-06,
+      "loss": 1.0843,
+      "step": 540
+    },
+    {
+      "epoch": 0.937231298366294,
+      "grad_norm": 0.36102572083473206,
+      "learning_rate": 1.89862765508974e-06,
+      "loss": 1.0874,
+      "step": 545
+    },
+    {
+      "epoch": 0.945829750644884,
+      "grad_norm": 0.4396149516105652,
+      "learning_rate": 1.896116276533208e-06,
+      "loss": 1.1614,
+      "step": 550
+    },
+    {
+      "epoch": 0.9544282029234737,
+      "grad_norm": 0.3992445170879364,
+      "learning_rate": 1.8935758721375862e-06,
+      "loss": 1.122,
+      "step": 555
+    },
+    {
+      "epoch": 0.9630266552020637,
+      "grad_norm": 0.3218730092048645,
+      "learning_rate": 1.8910065241883678e-06,
+      "loss": 0.9754,
+      "step": 560
+    },
+    {
+      "epoch": 0.9716251074806534,
+      "grad_norm": 0.3818008601665497,
+      "learning_rate": 1.8884083159085468e-06,
+      "loss": 1.1747,
+      "step": 565
+    },
+    {
+      "epoch": 0.9802235597592434,
+      "grad_norm": 0.407429575920105,
+      "learning_rate": 1.8857813314559254e-06,
+      "loss": 1.1048,
+      "step": 570
+    },
+    {
+      "epoch": 0.9888220120378332,
+      "grad_norm": 0.4260193407535553,
+      "learning_rate": 1.8831256559203843e-06,
+      "loss": 1.1217,
+      "step": 575
+    },
+    {
+      "epoch": 0.9974204643164231,
+      "grad_norm": 0.3930356204509735,
+      "learning_rate": 1.8804413753211304e-06,
+      "loss": 1.1259,
+      "step": 580
+    },
+    {
+      "epoch": 1.0051590713671539,
+      "grad_norm": 0.3557448983192444,
+      "learning_rate": 1.8777285766039075e-06,
+      "loss": 1.1636,
+      "step": 585
+    },
+    {
+      "epoch": 1.0137575236457437,
+      "grad_norm": 0.34454646706581116,
+      "learning_rate": 1.8749873476381826e-06,
+      "loss": 1.093,
+      "step": 590
+    },
+    {
+      "epoch": 1.0223559759243337,
+      "grad_norm": 0.4703282415866852,
+      "learning_rate": 1.8722177772142973e-06,
+      "loss": 1.0716,
+      "step": 595
+    },
+    {
+      "epoch": 1.0309544282029235,
+      "grad_norm": 0.418491929769516,
+      "learning_rate": 1.8694199550405942e-06,
+      "loss": 1.0626,
+      "step": 600
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 2905,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4341188063131075e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc08ca78b1b85a05419bb60ccf59304233558785fc7d6c7ac2b94b1b45d61f82
+size 8376

zero_to_fp32.py

@@ -0,0 +1,674 @@
+#!/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 json
+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)
+
+        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 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}"
+            )
+
+        # 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_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)
+
+    # 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")
+        state_dict_split = split_torch_state_dict_into_shards(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
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard = {tensor: state_dict[tensor].contiguous() for tensor in tensors}
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard, output_path)
+
+    # 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)