Add files using upload-large-folder tool

README.md

@@ -0,0 +1,209 @@
+---
+base_model: tacodevs/Behemoth-X-R1-123B
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:tacodevs/Behemoth-X-R1-123B
+- lora
+- sft
+- transformers
+- trl
+---
+
+# 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 0.18.1

adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "tacodevs/Behemoth-X-R1-123B",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 256,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "gate_proj",
+    "k_proj",
+    "o_proj",
+    "down_proj",
+    "v_proj",
+    "q_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d85da0f15060d1fd7624ba4f12ffde0f1bb001ad0e357f1938a4c724fe83c643
+size 4475494496

chat_template.jinja

@@ -0,0 +1 @@
+{{ bos_token }}{% for message in messages %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + '[/INST]' }}{% elif message['role'] == 'system' %}{{ '[SYSTEM_PROMPT] ' + message['content'] + '[/SYSTEM_PROMPT]' }}{% elif message['role'] == 'assistant' %}{{ ' ' + message['content'] + eos_token }}{% else %}{{ raise_exception('Only user, system and assistant roles are supported!') }}{% endif %}{% endfor %}

global_step199/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:180a2fb934980ea4cab4f3fa1b39af708330cdbc0bd5862aaabd01587b67d396
+size 3356495909

global_step199/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:e8a3072d4933c1e41819722b5f1f689cef43c7317159b9d3fc7cfa06fa3afa30
+size 3356495909

global_step199/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23f79a218cca1045b3f1644470ee01681541b05ecc954e1eaa88bd651b013d53
+size 3356495909

global_step199/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:c5e1adbd14dc62905e108928f2eb0daef90c342c4ee99f0f2c5482bd0a311818
+size 3356495909

global_step199/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b170f1e1dad3155f80e3875498ad7fd3189f4a4a9e2e9241e60870c48f7c23b
+size 3356495909

global_step199/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ca716667ef84634e2692d7964443c30843a2c3885becc48e6580095bc90ad85
+size 3356495909

global_step199/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7043d6b2d80e13076357c8c32099cfe7e9e6fb61dd473939f6f5497e58abe7a1
+size 3356495909

global_step199/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f074fb102c3953ef85219cd0cade34219025c783f2b6703fa5bd3587f6afbdbb
+size 3356495909

global_step199/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4862a402e4fea8d539cfd93701818e7ca259f3f8e497bdcc0e5293589dd3bb1a
+size 1219249

global_step199/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6bd503b3ef41f277d2fcd4155db9e2f5038041415dc9d6dd1493611b63603a3
+size 1219249

global_step199/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f2c902cd96d65ee71d0f8353845b63ff1a169427bd9af26aa69f5226be18b6e
+size 1219249

global_step199/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36c0e0cfd55042819e85df7b948d48f2577a55912ebb05d5cd776ca2d4524f48
+size 1219249

global_step199/zero_pp_rank_4_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dcdfa51d96bddddca749c993b8cc6fdd65b758ff3f9854a1bc41d4eee0b4d7a
+size 1219249

global_step199/zero_pp_rank_5_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13e7e282d43b883645a4b7f7407d8547409b860a65cce93440aad2212c3bcf3f
+size 1219249

global_step199/zero_pp_rank_6_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:189630ccd491f44a5ba12b25bc6d6e13c4bb7e35d02b24ee233509f3e8418f09
+size 1219249

global_step199/zero_pp_rank_7_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6dc66c6aa48f9c54d6e2e2d30e382111d008a2e928e19a158d83c997eb8cbe1c
+size 1219249

latest

@@ -0,0 +1 @@
+global_step199

rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5b29f519327a7187c0d997facd71e50e8c1f25b4c5c44d3d8e544c52cedc927
+size 16389

rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07214d5ba812a60b7c519b610c6cc74ab3e34da3fd21dc95e05cf3bd9c2f651f
+size 16389

rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a55a8ed7d787aff59b893a33baffa3273ab6b5ab9e7114dc1b4609828eceb6f
+size 16389

rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9efe94592e00a16a753ad52a6b48b0af4f2b8ecd5f8817fa842d57a73461f6de
+size 16389

rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2eaf65073b1d9cb631627c7d37e12d5d736eb87e14bae1b60ec25f6564ea2298
+size 16389

rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:483a4ecfdd2494917208f93996e07851cb5cc65362d7300bd19096346b88fa78
+size 16389

rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68ea28bb3fcc40689470726a82a29b9924d6c73a803800aff51704b4483dfe06
+size 16389

rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:889ccd6e5d452a61b7b76c9d1752553ce72c8ce1706f876bc03c0c0fd7e55272
+size 16389

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "add_prefix_space": null,
+  "backend": "tokenizers",
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "is_local": false,
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

trainer_state.json

@@ -0,0 +1,445 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 199,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.0732421875,
+      "epoch": 0.005044136191677175,
+      "grad_norm": 3.729904548791402,
+      "learning_rate": 0.0,
+      "loss": 1.5705682039260864,
+      "mean_token_accuracy": 0.5984444469213486,
+      "num_tokens": 397809.0,
+      "step": 1
+    },
+    {
+      "entropy": 1.140625,
+      "epoch": 0.025220680958385876,
+      "grad_norm": 3.537819920462687,
+      "learning_rate": 1.295370924755994e-05,
+      "loss": 1.569430947303772,
+      "mean_token_accuracy": 0.5931065641343594,
+      "num_tokens": 1792332.0,
+      "step": 5
+    },
+    {
+      "entropy": 1.28984375,
+      "epoch": 0.05044136191677175,
+      "grad_norm": 1.9953551967058947,
+      "learning_rate": 1.853256816058254e-05,
+      "loss": 1.4315099716186523,
+      "mean_token_accuracy": 0.6139604687690735,
+      "num_tokens": 3573398.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2671875,
+      "epoch": 0.07566204287515763,
+      "grad_norm": 2.0123345607376537,
+      "learning_rate": 1.989637305699482e-05,
+      "loss": 1.303097915649414,
+      "mean_token_accuracy": 0.6386384278535843,
+      "num_tokens": 5352795.0,
+      "step": 15
+    },
+    {
+      "entropy": 1.34296875,
+      "epoch": 0.1008827238335435,
+      "grad_norm": 2.3429686709929016,
+      "learning_rate": 1.9637305699481867e-05,
+      "loss": 1.2593023300170898,
+      "mean_token_accuracy": 0.646000188589096,
+      "num_tokens": 7129865.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.2890625,
+      "epoch": 0.12610340479192939,
+      "grad_norm": 2.093171715611346,
+      "learning_rate": 1.9378238341968913e-05,
+      "loss": 1.25695743560791,
+      "mean_token_accuracy": 0.6471466660499573,
+      "num_tokens": 8868732.0,
+      "step": 25
+    },
+    {
+      "entropy": 1.315234375,
+      "epoch": 0.15132408575031525,
+      "grad_norm": 1.5214039865876905,
+      "learning_rate": 1.9119170984455958e-05,
+      "loss": 1.2165451049804688,
+      "mean_token_accuracy": 0.6549518913030624,
+      "num_tokens": 10737138.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.282421875,
+      "epoch": 0.17654476670870115,
+      "grad_norm": 1.7720335086338008,
+      "learning_rate": 1.8860103626943007e-05,
+      "loss": 1.1993481636047363,
+      "mean_token_accuracy": 0.657388761639595,
+      "num_tokens": 12626963.0,
+      "step": 35
+    },
+    {
+      "entropy": 1.2640625,
+      "epoch": 0.201765447667087,
+      "grad_norm": 1.8352721153566052,
+      "learning_rate": 1.8601036269430053e-05,
+      "loss": 1.1846282958984375,
+      "mean_token_accuracy": 0.6627674549818039,
+      "num_tokens": 14433262.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.251953125,
+      "epoch": 0.22698612862547288,
+      "grad_norm": 1.8067651099435984,
+      "learning_rate": 1.83419689119171e-05,
+      "loss": 1.1756595611572265,
+      "mean_token_accuracy": 0.6638763636350632,
+      "num_tokens": 16326390.0,
+      "step": 45
+    },
+    {
+      "entropy": 1.273046875,
+      "epoch": 0.25220680958385877,
+      "grad_norm": 3.465628169730638,
+      "learning_rate": 1.8082901554404147e-05,
+      "loss": 1.1847558975219727,
+      "mean_token_accuracy": 0.6604940384626389,
+      "num_tokens": 18279034.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.25625,
+      "epoch": 0.27742749054224464,
+      "grad_norm": 1.4054861856544547,
+      "learning_rate": 1.7823834196891192e-05,
+      "loss": 1.1652128219604492,
+      "mean_token_accuracy": 0.6664551287889481,
+      "num_tokens": 20208031.0,
+      "step": 55
+    },
+    {
+      "entropy": 1.266796875,
+      "epoch": 0.3026481715006305,
+      "grad_norm": 3.4183832212008456,
+      "learning_rate": 1.756476683937824e-05,
+      "loss": 1.1770149230957032,
+      "mean_token_accuracy": 0.6627189248800278,
+      "num_tokens": 21908532.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.26484375,
+      "epoch": 0.32786885245901637,
+      "grad_norm": 3.1555109653499964,
+      "learning_rate": 1.7305699481865287e-05,
+      "loss": 1.1431950569152831,
+      "mean_token_accuracy": 0.6696279108524322,
+      "num_tokens": 23700105.0,
+      "step": 65
+    },
+    {
+      "entropy": 1.242578125,
+      "epoch": 0.3530895334174023,
+      "grad_norm": 2.3839995394597286,
+      "learning_rate": 1.7046632124352332e-05,
+      "loss": 1.1668883323669434,
+      "mean_token_accuracy": 0.6657638341188431,
+      "num_tokens": 25603944.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.24921875,
+      "epoch": 0.37831021437578816,
+      "grad_norm": 1.8426925035490107,
+      "learning_rate": 1.6787564766839378e-05,
+      "loss": 1.13353271484375,
+      "mean_token_accuracy": 0.6711880832910537,
+      "num_tokens": 27318273.0,
+      "step": 75
+    },
+    {
+      "entropy": 1.26171875,
+      "epoch": 0.403530895334174,
+      "grad_norm": 1.9994154023171613,
+      "learning_rate": 1.6528497409326427e-05,
+      "loss": 1.1364672660827637,
+      "mean_token_accuracy": 0.672755679488182,
+      "num_tokens": 29158308.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.249609375,
+      "epoch": 0.4287515762925599,
+      "grad_norm": 2.9275771398624597,
+      "learning_rate": 1.6269430051813472e-05,
+      "loss": 1.121041488647461,
+      "mean_token_accuracy": 0.6760148137807847,
+      "num_tokens": 30941822.0,
+      "step": 85
+    },
+    {
+      "entropy": 1.25390625,
+      "epoch": 0.45397225725094575,
+      "grad_norm": 1.5615516079216,
+      "learning_rate": 1.6010362694300518e-05,
+      "loss": 1.1162029266357423,
+      "mean_token_accuracy": 0.6768355071544647,
+      "num_tokens": 32787654.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.254296875,
+      "epoch": 0.4791929382093317,
+      "grad_norm": 2.8170519799515197,
+      "learning_rate": 1.5751295336787566e-05,
+      "loss": 1.1182659149169922,
+      "mean_token_accuracy": 0.6761424988508224,
+      "num_tokens": 34644366.0,
+      "step": 95
+    },
+    {
+      "entropy": 1.238671875,
+      "epoch": 0.5044136191677175,
+      "grad_norm": 2.805028653265598,
+      "learning_rate": 1.5492227979274612e-05,
+      "loss": 1.1245113372802735,
+      "mean_token_accuracy": 0.6737537115812302,
+      "num_tokens": 36483579.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.226953125,
+      "epoch": 0.5296343001261034,
+      "grad_norm": 1.438265666554789,
+      "learning_rate": 1.523316062176166e-05,
+      "loss": 1.1135278701782227,
+      "mean_token_accuracy": 0.6800242573022842,
+      "num_tokens": 38534892.0,
+      "step": 105
+    },
+    {
+      "entropy": 1.240625,
+      "epoch": 0.5548549810844893,
+      "grad_norm": 6.154593405469548,
+      "learning_rate": 1.4974093264248706e-05,
+      "loss": 1.127542495727539,
+      "mean_token_accuracy": 0.6712189227342605,
+      "num_tokens": 40255595.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.221875,
+      "epoch": 0.5800756620428752,
+      "grad_norm": 1.712858492865262,
+      "learning_rate": 1.4715025906735752e-05,
+      "loss": 1.0767477989196776,
+      "mean_token_accuracy": 0.6854143172502518,
+      "num_tokens": 42143795.0,
+      "step": 115
+    },
+    {
+      "entropy": 1.228125,
+      "epoch": 0.605296343001261,
+      "grad_norm": 1.7954082649002236,
+      "learning_rate": 1.4455958549222799e-05,
+      "loss": 1.1128243446350097,
+      "mean_token_accuracy": 0.6780534535646439,
+      "num_tokens": 44010920.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.25,
+      "epoch": 0.6305170239596469,
+      "grad_norm": 2.557128723303994,
+      "learning_rate": 1.4196891191709844e-05,
+      "loss": 1.1001087188720704,
+      "mean_token_accuracy": 0.6781179249286652,
+      "num_tokens": 45633080.0,
+      "step": 125
+    },
+    {
+      "entropy": 1.233984375,
+      "epoch": 0.6557377049180327,
+      "grad_norm": 1.6132918279446875,
+      "learning_rate": 1.3937823834196893e-05,
+      "loss": 1.0829249382019044,
+      "mean_token_accuracy": 0.6852418035268784,
+      "num_tokens": 47264233.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.2265625,
+      "epoch": 0.6809583858764187,
+      "grad_norm": 1.35683478183202,
+      "learning_rate": 1.3678756476683939e-05,
+      "loss": 1.0771641731262207,
+      "mean_token_accuracy": 0.6857132732868194,
+      "num_tokens": 49137662.0,
+      "step": 135
+    },
+    {
+      "entropy": 1.2171875,
+      "epoch": 0.7061790668348046,
+      "grad_norm": 1.5570774063045756,
+      "learning_rate": 1.3419689119170986e-05,
+      "loss": 1.0707455635070802,
+      "mean_token_accuracy": 0.6857165455818176,
+      "num_tokens": 50943509.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.248828125,
+      "epoch": 0.7313997477931904,
+      "grad_norm": 2.8786377018075457,
+      "learning_rate": 1.3160621761658033e-05,
+      "loss": 1.082810401916504,
+      "mean_token_accuracy": 0.6832135945558548,
+      "num_tokens": 52697400.0,
+      "step": 145
+    },
+    {
+      "entropy": 1.212109375,
+      "epoch": 0.7566204287515763,
+      "grad_norm": 1.4838869261745697,
+      "learning_rate": 1.2901554404145079e-05,
+      "loss": 1.055869960784912,
+      "mean_token_accuracy": 0.6903243094682694,
+      "num_tokens": 54550790.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.2265625,
+      "epoch": 0.7818411097099621,
+      "grad_norm": 1.322651287548672,
+      "learning_rate": 1.2642487046632126e-05,
+      "loss": 1.0742380142211914,
+      "mean_token_accuracy": 0.6852674335241318,
+      "num_tokens": 56273458.0,
+      "step": 155
+    },
+    {
+      "entropy": 1.2453125,
+      "epoch": 0.807061790668348,
+      "grad_norm": 8.55404246200832,
+      "learning_rate": 1.2383419689119171e-05,
+      "loss": 1.0786266326904297,
+      "mean_token_accuracy": 0.6855500996112823,
+      "num_tokens": 58121056.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.203125,
+      "epoch": 0.832282471626734,
+      "grad_norm": 2.0497153186806325,
+      "learning_rate": 1.212435233160622e-05,
+      "loss": 1.0652950286865235,
+      "mean_token_accuracy": 0.6877057552337646,
+      "num_tokens": 60048538.0,
+      "step": 165
+    },
+    {
+      "entropy": 1.208203125,
+      "epoch": 0.8575031525851198,
+      "grad_norm": 1.2643588159534247,
+      "learning_rate": 1.1865284974093266e-05,
+      "loss": 1.0611819267272948,
+      "mean_token_accuracy": 0.6904475122690201,
+      "num_tokens": 61835940.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.2171875,
+      "epoch": 0.8827238335435057,
+      "grad_norm": 1.204921091847234,
+      "learning_rate": 1.1606217616580311e-05,
+      "loss": 1.0413402557373046,
+      "mean_token_accuracy": 0.6954286068677902,
+      "num_tokens": 63573308.0,
+      "step": 175
+    },
+    {
+      "entropy": 1.20078125,
+      "epoch": 0.9079445145018915,
+      "grad_norm": 1.387043880574625,
+      "learning_rate": 1.1347150259067358e-05,
+      "loss": 1.049238681793213,
+      "mean_token_accuracy": 0.6919074177742004,
+      "num_tokens": 65497008.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.206640625,
+      "epoch": 0.9331651954602774,
+      "grad_norm": 1.279478089664614,
+      "learning_rate": 1.1088082901554404e-05,
+      "loss": 1.0584297180175781,
+      "mean_token_accuracy": 0.6892471343278885,
+      "num_tokens": 67254338.0,
+      "step": 185
+    },
+    {
+      "entropy": 1.2171875,
+      "epoch": 0.9583858764186634,
+      "grad_norm": 1.5719174306561159,
+      "learning_rate": 1.0829015544041453e-05,
+      "loss": 1.0455381393432617,
+      "mean_token_accuracy": 0.6926654905080796,
+      "num_tokens": 69176523.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.209375,
+      "epoch": 0.9836065573770492,
+      "grad_norm": 1.4906711673793527,
+      "learning_rate": 1.0569948186528498e-05,
+      "loss": 1.0241477966308594,
+      "mean_token_accuracy": 0.6976693719625473,
+      "num_tokens": 71048975.0,
+      "step": 195
+    },
+    {
+      "epoch": 1.0,
+      "eval_entropy": 1.2234002976190477,
+      "eval_loss": 1.0335501432418823,
+      "eval_mean_token_accuracy": 0.6941843855948675,
+      "eval_num_tokens": 72109037.0,
+      "eval_runtime": 533.6615,
+      "eval_samples_per_second": 0.624,
+      "eval_steps_per_second": 0.079,
+      "step": 199
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 398,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 940923320795136.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c27c47fb07591ee19b0ae602689db755bb34d6e5487af3dea5041a88ffdec13a
+size 7889

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 ZERO_STAGE not 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("Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info("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)