Training in progress, step 73, checkpoint

checkpoint-73/README.md

@@ -0,0 +1,202 @@
+---
+base_model: unsloth/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 0.14.0

checkpoint-73/adapter_config.json

@@ -0,0 +1,37 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "unsloth/Mistral-Nemo-Instruct-2407",
+  "bias": "none",
+  "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": 64,
+  "lora_bias": false,
+  "lora_dropout": 0.25,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "up_proj",
+    "q_proj",
+    "k_proj",
+    "o_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-73/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c27f628f3ec276db7ecd620d1d79c656b1d32f8256610691cd07176e8ed4eed
+size 228141160

checkpoint-73/global_step73/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b621f730ba3794db6ea2fa5fda1c9214c07ac4824f09ffccddb60edf220039f
+size 201837552

checkpoint-73/global_step73/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:b052837a5acfd645de9910fdf6b2747e9ad1a55e4c632d49db2b5efcff75ba5c
+size 201837552

checkpoint-73/global_step73/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51eb4f7d8209292450a861ea4f3cb094491f6d9f56f2885f0add9b154f2afa7c
+size 201837552

checkpoint-73/global_step73/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:a618d161d4a1fa5e0a2a0ecad91120af5f26a1fdbbc1a194a59fbb5740f64451
+size 201837552

checkpoint-73/global_step73/zero_pp_rank_0_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a5825271b4ccb90a836ff878e3db6e471e8221dcb054d4b010f620478cd6f95
+size 174532150

checkpoint-73/global_step73/zero_pp_rank_1_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c02a18c5259f8a1cb3ce2e63033bc6efd3fe0eecb994e5109accb40841edd98
+size 174532150

checkpoint-73/global_step73/zero_pp_rank_2_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dac2b51e397da81f4247d7372e6a022d40dacbe7603498a5181618db6a14a06c
+size 174532150

checkpoint-73/global_step73/zero_pp_rank_3_mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c07cbae3df7f042284b0f4af2c7bea99bcdd95075e3a179692fe3f56a2a7eb6b
+size 174532150

checkpoint-73/latest

@@ -0,0 +1 @@
+global_step73

checkpoint-73/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee715e4064acdc7a3a45c8529fbaa7e82d094f87a1ca8730b3ad7bc3f9598ce0
+size 14960

checkpoint-73/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c670a9f6a27286fbb477fd213f5f4442e2bea77df66f6e4c73824e9b50cf416c
+size 14960

checkpoint-73/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:785c2bad33adbcbcc5d882101e7cf1059946cbd7d03093c049597a45c73321b6
+size 14960

checkpoint-73/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:743a243fdccf9b48bc7f842471ced0cd96bb100b4bf4d6c32eecd6cdb2d1cdde
+size 14960

checkpoint-73/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b27af4070ce6c921d02745a1ca1f7fa75b66620f984d9535746fda4dd3d265d0
+size 1064

checkpoint-73/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": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-73/tokenizer.json

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

checkpoint-73/trainer_state.json

@@ -0,0 +1,560 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.20221606648199447,
+  "eval_steps": 37,
+  "global_step": 73,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002770083102493075,
+      "grad_norm": 0.9076941366015471,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1945,
+      "step": 1
+    },
+    {
+      "epoch": 0.002770083102493075,
+      "eval_loss": 1.0296825170516968,
+      "eval_runtime": 104.9654,
+      "eval_samples_per_second": 0.953,
+      "eval_steps_per_second": 0.238,
+      "step": 1
+    },
+    {
+      "epoch": 0.00554016620498615,
+      "grad_norm": 1.0344272252302953,
+      "learning_rate": 5e-06,
+      "loss": 1.3385,
+      "step": 2
+    },
+    {
+      "epoch": 0.008310249307479225,
+      "grad_norm": 1.0977608192145218,
+      "learning_rate": 7.5e-06,
+      "loss": 1.3767,
+      "step": 3
+    },
+    {
+      "epoch": 0.0110803324099723,
+      "grad_norm": 0.8895382737806115,
+      "learning_rate": 1e-05,
+      "loss": 1.1881,
+      "step": 4
+    },
+    {
+      "epoch": 0.013850415512465374,
+      "grad_norm": 0.958710062908987,
+      "learning_rate": 1.25e-05,
+      "loss": 1.2936,
+      "step": 5
+    },
+    {
+      "epoch": 0.01662049861495845,
+      "grad_norm": 0.8447951280142822,
+      "learning_rate": 1.5e-05,
+      "loss": 1.1547,
+      "step": 6
+    },
+    {
+      "epoch": 0.019390581717451522,
+      "grad_norm": 0.8794765482094637,
+      "learning_rate": 1.75e-05,
+      "loss": 1.3444,
+      "step": 7
+    },
+    {
+      "epoch": 0.0221606648199446,
+      "grad_norm": 0.8437125997364887,
+      "learning_rate": 2e-05,
+      "loss": 1.2765,
+      "step": 8
+    },
+    {
+      "epoch": 0.024930747922437674,
+      "grad_norm": 0.7276844901691898,
+      "learning_rate": 2.25e-05,
+      "loss": 1.2031,
+      "step": 9
+    },
+    {
+      "epoch": 0.027700831024930747,
+      "grad_norm": 0.577477596526854,
+      "learning_rate": 2.5e-05,
+      "loss": 1.1725,
+      "step": 10
+    },
+    {
+      "epoch": 0.030470914127423823,
+      "grad_norm": 0.4618405348091677,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 1.1008,
+      "step": 11
+    },
+    {
+      "epoch": 0.0332409972299169,
+      "grad_norm": 0.3671423964001518,
+      "learning_rate": 3e-05,
+      "loss": 1.0963,
+      "step": 12
+    },
+    {
+      "epoch": 0.036011080332409975,
+      "grad_norm": 0.4139651359524478,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 1.2831,
+      "step": 13
+    },
+    {
+      "epoch": 0.038781163434903045,
+      "grad_norm": 0.4364185358505406,
+      "learning_rate": 3.5e-05,
+      "loss": 1.0524,
+      "step": 14
+    },
+    {
+      "epoch": 0.04155124653739612,
+      "grad_norm": 0.5116601253143136,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.1499,
+      "step": 15
+    },
+    {
+      "epoch": 0.0443213296398892,
+      "grad_norm": 0.5236113301724862,
+      "learning_rate": 4e-05,
+      "loss": 1.1466,
+      "step": 16
+    },
+    {
+      "epoch": 0.04709141274238227,
+      "grad_norm": 0.4606756443741647,
+      "learning_rate": 4.25e-05,
+      "loss": 1.1385,
+      "step": 17
+    },
+    {
+      "epoch": 0.04986149584487535,
+      "grad_norm": 0.3991457405979455,
+      "learning_rate": 4.5e-05,
+      "loss": 1.108,
+      "step": 18
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 1.8553548238184947,
+      "learning_rate": 4.75e-05,
+      "loss": 0.9794,
+      "step": 19
+    },
+    {
+      "epoch": 0.055401662049861494,
+      "grad_norm": 0.31912881369455715,
+      "learning_rate": 5e-05,
+      "loss": 1.1583,
+      "step": 20
+    },
+    {
+      "epoch": 0.05817174515235457,
+      "grad_norm": 0.3315962832679952,
+      "learning_rate": 4.999904513837887e-05,
+      "loss": 1.1927,
+      "step": 21
+    },
+    {
+      "epoch": 0.060941828254847646,
+      "grad_norm": 0.3081653717373631,
+      "learning_rate": 4.999618063456087e-05,
+      "loss": 1.1666,
+      "step": 22
+    },
+    {
+      "epoch": 0.06371191135734072,
+      "grad_norm": 0.2386753107269311,
+      "learning_rate": 4.999140673167533e-05,
+      "loss": 1.1517,
+      "step": 23
+    },
+    {
+      "epoch": 0.0664819944598338,
+      "grad_norm": 0.20460753812847132,
+      "learning_rate": 4.998472383491481e-05,
+      "loss": 1.1475,
+      "step": 24
+    },
+    {
+      "epoch": 0.06925207756232687,
+      "grad_norm": 0.19426749294122878,
+      "learning_rate": 4.997613251150082e-05,
+      "loss": 1.0427,
+      "step": 25
+    },
+    {
+      "epoch": 0.07202216066481995,
+      "grad_norm": 0.16821147486362656,
+      "learning_rate": 4.996563349063557e-05,
+      "loss": 1.1392,
+      "step": 26
+    },
+    {
+      "epoch": 0.07479224376731301,
+      "grad_norm": 0.16476251779742826,
+      "learning_rate": 4.995322766344013e-05,
+      "loss": 1.0624,
+      "step": 27
+    },
+    {
+      "epoch": 0.07756232686980609,
+      "grad_norm": 0.14974022119841934,
+      "learning_rate": 4.993891608287879e-05,
+      "loss": 1.0357,
+      "step": 28
+    },
+    {
+      "epoch": 0.08033240997229917,
+      "grad_norm": 0.18664906507712956,
+      "learning_rate": 4.992269996366969e-05,
+      "loss": 1.0724,
+      "step": 29
+    },
+    {
+      "epoch": 0.08310249307479224,
+      "grad_norm": 0.1850945977027583,
+      "learning_rate": 4.990458068218171e-05,
+      "loss": 1.1334,
+      "step": 30
+    },
+    {
+      "epoch": 0.08587257617728532,
+      "grad_norm": 0.18423955915331258,
+      "learning_rate": 4.9884559776317644e-05,
+      "loss": 1.1707,
+      "step": 31
+    },
+    {
+      "epoch": 0.0886426592797784,
+      "grad_norm": 0.17778624557198042,
+      "learning_rate": 4.9862638945383674e-05,
+      "loss": 1.0923,
+      "step": 32
+    },
+    {
+      "epoch": 0.09141274238227147,
+      "grad_norm": 0.18753159302589986,
+      "learning_rate": 4.983882004994517e-05,
+      "loss": 1.0743,
+      "step": 33
+    },
+    {
+      "epoch": 0.09418282548476455,
+      "grad_norm": 0.16951133250798278,
+      "learning_rate": 4.98131051116687e-05,
+      "loss": 1.0776,
+      "step": 34
+    },
+    {
+      "epoch": 0.09695290858725762,
+      "grad_norm": 0.15767321847790128,
+      "learning_rate": 4.9785496313150515e-05,
+      "loss": 1.0987,
+      "step": 35
+    },
+    {
+      "epoch": 0.0997229916897507,
+      "grad_norm": 0.162961453637477,
+      "learning_rate": 4.975599599773124e-05,
+      "loss": 1.2561,
+      "step": 36
+    },
+    {
+      "epoch": 0.10249307479224377,
+      "grad_norm": 0.15327784219657248,
+      "learning_rate": 4.972460666929699e-05,
+      "loss": 1.1456,
+      "step": 37
+    },
+    {
+      "epoch": 0.10249307479224377,
+      "eval_loss": 0.7954654693603516,
+      "eval_runtime": 105.8629,
+      "eval_samples_per_second": 0.945,
+      "eval_steps_per_second": 0.236,
+      "step": 37
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.17503484326795066,
+      "learning_rate": 4.969133099206691e-05,
+      "loss": 1.1647,
+      "step": 38
+    },
+    {
+      "epoch": 0.10803324099722991,
+      "grad_norm": 0.1616243718822839,
+      "learning_rate": 4.965617179036695e-05,
+      "loss": 1.0552,
+      "step": 39
+    },
+    {
+      "epoch": 0.11080332409972299,
+      "grad_norm": 0.14651727257725808,
+      "learning_rate": 4.96191320483902e-05,
+      "loss": 0.9286,
+      "step": 40
+    },
+    {
+      "epoch": 0.11357340720221606,
+      "grad_norm": 0.16761163260684342,
+      "learning_rate": 4.958021490994361e-05,
+      "loss": 1.0554,
+      "step": 41
+    },
+    {
+      "epoch": 0.11634349030470914,
+      "grad_norm": 0.18422943695103963,
+      "learning_rate": 4.953942367818113e-05,
+      "loss": 1.1106,
+      "step": 42
+    },
+    {
+      "epoch": 0.11911357340720222,
+      "grad_norm": 0.1473247570248768,
+      "learning_rate": 4.9496761815323344e-05,
+      "loss": 1.2059,
+      "step": 43
+    },
+    {
+      "epoch": 0.12188365650969529,
+      "grad_norm": 0.14787258143002216,
+      "learning_rate": 4.9452232942363634e-05,
+      "loss": 1.1916,
+      "step": 44
+    },
+    {
+      "epoch": 0.12465373961218837,
+      "grad_norm": 0.13340848225076712,
+      "learning_rate": 4.940584083876084e-05,
+      "loss": 1.0421,
+      "step": 45
+    },
+    {
+      "epoch": 0.12742382271468145,
+      "grad_norm": 0.15834348128625553,
+      "learning_rate": 4.935758944211844e-05,
+      "loss": 1.0782,
+      "step": 46
+    },
+    {
+      "epoch": 0.13019390581717452,
+      "grad_norm": 0.14100327469205548,
+      "learning_rate": 4.9307482847850374e-05,
+      "loss": 1.03,
+      "step": 47
+    },
+    {
+      "epoch": 0.1329639889196676,
+      "grad_norm": 0.18919634141556568,
+      "learning_rate": 4.925552530883343e-05,
+      "loss": 1.0422,
+      "step": 48
+    },
+    {
+      "epoch": 0.13573407202216067,
+      "grad_norm": 0.15121560590293795,
+      "learning_rate": 4.9201721235046274e-05,
+      "loss": 0.9852,
+      "step": 49
+    },
+    {
+      "epoch": 0.13850415512465375,
+      "grad_norm": 0.14628904242090288,
+      "learning_rate": 4.914607519319512e-05,
+      "loss": 1.1146,
+      "step": 50
+    },
+    {
+      "epoch": 0.14127423822714683,
+      "grad_norm": 0.13794163183547553,
+      "learning_rate": 4.90885919063262e-05,
+      "loss": 1.0787,
+      "step": 51
+    },
+    {
+      "epoch": 0.1440443213296399,
+      "grad_norm": 0.13302301409932443,
+      "learning_rate": 4.902927625342476e-05,
+      "loss": 1.0993,
+      "step": 52
+    },
+    {
+      "epoch": 0.14681440443213298,
+      "grad_norm": 0.14124480977262469,
+      "learning_rate": 4.8968133269001096e-05,
+      "loss": 1.0137,
+      "step": 53
+    },
+    {
+      "epoch": 0.14958448753462603,
+      "grad_norm": 0.14791504265910085,
+      "learning_rate": 4.890516814266314e-05,
+      "loss": 1.1663,
+      "step": 54
+    },
+    {
+      "epoch": 0.1523545706371191,
+      "grad_norm": 0.13506987447877156,
+      "learning_rate": 4.884038621867599e-05,
+      "loss": 0.9941,
+      "step": 55
+    },
+    {
+      "epoch": 0.15512465373961218,
+      "grad_norm": 0.14611198065865885,
+      "learning_rate": 4.877379299550838e-05,
+      "loss": 0.9958,
+      "step": 56
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.16481726483113354,
+      "learning_rate": 4.8705394125365884e-05,
+      "loss": 1.1329,
+      "step": 57
+    },
+    {
+      "epoch": 0.16066481994459833,
+      "grad_norm": 0.1498291481940627,
+      "learning_rate": 4.8635195413711286e-05,
+      "loss": 1.058,
+      "step": 58
+    },
+    {
+      "epoch": 0.1634349030470914,
+      "grad_norm": 0.15057001184388058,
+      "learning_rate": 4.8563202818771735e-05,
+      "loss": 1.1062,
+      "step": 59
+    },
+    {
+      "epoch": 0.16620498614958448,
+      "grad_norm": 0.12916512409429184,
+      "learning_rate": 4.84894224510331e-05,
+      "loss": 1.0769,
+      "step": 60
+    },
+    {
+      "epoch": 0.16897506925207756,
+      "grad_norm": 0.13775185913237034,
+      "learning_rate": 4.841386057272129e-05,
+      "loss": 0.9624,
+      "step": 61
+    },
+    {
+      "epoch": 0.17174515235457063,
+      "grad_norm": 0.1447222703266104,
+      "learning_rate": 4.8336523597270765e-05,
+      "loss": 0.9653,
+      "step": 62
+    },
+    {
+      "epoch": 0.1745152354570637,
+      "grad_norm": 0.20538147828946965,
+      "learning_rate": 4.825741808878017e-05,
+      "loss": 0.9674,
+      "step": 63
+    },
+    {
+      "epoch": 0.1772853185595568,
+      "grad_norm": 0.15455976822474854,
+      "learning_rate": 4.8176550761455195e-05,
+      "loss": 1.1488,
+      "step": 64
+    },
+    {
+      "epoch": 0.18005540166204986,
+      "grad_norm": 0.14123106112442296,
+      "learning_rate": 4.8093928479038704e-05,
+      "loss": 1.0781,
+      "step": 65
+    },
+    {
+      "epoch": 0.18282548476454294,
+      "grad_norm": 0.153605912426132,
+      "learning_rate": 4.8009558254228185e-05,
+      "loss": 1.0193,
+      "step": 66
+    },
+    {
+      "epoch": 0.18559556786703602,
+      "grad_norm": 0.170305518682767,
+      "learning_rate": 4.792344724808048e-05,
+      "loss": 1.1249,
+      "step": 67
+    },
+    {
+      "epoch": 0.1883656509695291,
+      "grad_norm": 0.15808161680481964,
+      "learning_rate": 4.783560276940402e-05,
+      "loss": 0.9227,
+      "step": 68
+    },
+    {
+      "epoch": 0.19113573407202217,
+      "grad_norm": 0.13584733997983103,
+      "learning_rate": 4.7746032274138494e-05,
+      "loss": 1.0205,
+      "step": 69
+    },
+    {
+      "epoch": 0.19390581717451524,
+      "grad_norm": 0.152754335402458,
+      "learning_rate": 4.765474336472197e-05,
+      "loss": 1.1152,
+      "step": 70
+    },
+    {
+      "epoch": 0.19667590027700832,
+      "grad_norm": 0.1559364062645901,
+      "learning_rate": 4.7561743789445644e-05,
+      "loss": 1.004,
+      "step": 71
+    },
+    {
+      "epoch": 0.1994459833795014,
+      "grad_norm": 0.1601485375057651,
+      "learning_rate": 4.7467041441796214e-05,
+      "loss": 1.0463,
+      "step": 72
+    },
+    {
+      "epoch": 0.20221606648199447,
+      "grad_norm": 0.14468865075681345,
+      "learning_rate": 4.737064435978586e-05,
+      "loss": 0.9536,
+      "step": 73
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 361,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 73,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 217880671027200.0,
+  "train_batch_size": 5,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-73/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4df7d820b95121961ca32b27affe58af9424ecdbe740efca6416d669c045219e
+size 11640

checkpoint-73/zero_to_fp32.py

@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+#   python zero_to_fp32.py . output_dir/
+#   or
+#   python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device, weights_only=False)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+    total_files = len(files)
+    state_dicts = []
+    for f in tqdm(files, desc='Loading checkpoint shards'):
+        state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+    """
+    A pseudo tensor that collects partitioned weights.
+    It is more memory efficient when there are multiple groups.
+    """
+
+    def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+        self.flat_groups = flat_groups
+        self.flat_groups_offset = flat_groups_offset
+        self.offset = offset
+        self.partitioned_numel = partitioned_numel
+        self.shape = shape
+        self.dtype = self.flat_groups[0][0].dtype
+
+    def contiguous(self):
+        """
+        Merge partitioned weights from flat_groups into a single tensor.
+        """
+        end_idx = self.offset + self.partitioned_numel
+        world_size = len(self.flat_groups)
+        pad_flat_param_chunks = []
+
+        for rank_i in range(world_size):
+            # for each rank, we need to collect weights from related group/groups
+            flat_groups_at_rank_i = self.flat_groups[rank_i]
+            start_group_id = None
+            end_group_id = None
+            for group_id in range(len(self.flat_groups_offset)):
+                if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+                    start_group_id = group_id
+                if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+                    end_group_id = group_id
+                    break
+            # collect weights from related group/groups
+            for group_id in range(start_group_id, end_group_id + 1):
+                flat_tensor = flat_groups_at_rank_i[group_id]
+                start_offset = self.offset - self.flat_groups_offset[group_id]
+                end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+                pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+        # collect weights from all ranks
+        pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+        param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+        return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+    for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # memory efficient tensor
+        tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+        state_dict[name] = tensor
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+    """
+    Convert state_dict of GatheredTensor to torch tensor
+    """
+    torch_state_dict = {}
+    converted_tensors = {}
+    for name, tensor in state_dict.items():
+        tensor_id = id(tensor)
+        if tensor_id in converted_tensors:  # shared tensors
+            shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+            torch_state_dict[name] = shared_tensor
+        else:
+            converted_tensors[tensor_id] = name
+            if return_empty_tensor:
+                torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+            else:
+                torch_state_dict[name] = tensor.contiguous()
+    return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                             tag=None,
+                                             exclude_frozen_parameters=False,
+                                             lazy_mode=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+        - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+          Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+    You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint. Or you can load state_dict in lazy mode ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+        for name, lazy_tensor in state_dict.item():
+            tensor = lazy_tensor.contiguous()  # to cpu
+            print(name, tensor)
+            # del tensor to release memory if it no longer in use
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+    if lazy_mode:
+        return state_dict
+    else:
+        return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+                                               output_dir,
+                                               max_shard_size="5GB",
+                                               safe_serialization=False,
+                                               tag=None,
+                                               exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_dir``: directory to the pytorch fp32 state_dict output files
+        - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+        - ``safe_serialization``:  whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    # Dependency pre-check
+    if safe_serialization:
+        try:
+            from safetensors.torch import save_file
+        except ImportError:
+            print('If you want to use `safe_serialization`, please `pip install safetensors`')
+            raise
+    if max_shard_size is not None:
+        try:
+            from huggingface_hub import split_torch_state_dict_into_shards
+        except ImportError:
+            print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+            raise
+
+    # Convert zero checkpoint to state_dict
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+                                                          tag,
+                                                          exclude_frozen_parameters,
+                                                          lazy_mode=True)
+
+    # Shard the model if it is too big.
+    weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+    if max_shard_size is not None:
+        filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+        # an memory-efficient approach for sharding
+        empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+        state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+                                                              filename_pattern=filename_pattern,
+                                                              max_shard_size=max_shard_size)
+    else:
+        from collections import namedtuple
+        StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+        state_dict_split = StateDictSplit(is_sharded=False,
+                                          filename_to_tensors={weights_name: list(state_dict.keys())})
+
+    # Save the model by shard
+    os.makedirs(output_dir, exist_ok=True)
+    filename_to_tensors = state_dict_split.filename_to_tensors.items()
+    for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+        shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+        shard_state_dict = to_torch_tensor(shard_state_dict)
+        output_path = os.path.join(output_dir, shard_file)
+        if safe_serialization:
+            save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+        else:
+            torch.save(shard_state_dict, output_path)
+        # release the memory of current shard
+        for tensor_name in list(shard_state_dict.keys()):
+            del state_dict[tensor_name]
+            del shard_state_dict[tensor_name]
+        del shard_state_dict
+        gc.collect()
+
+    # Save index if sharded
+    if state_dict_split.is_sharded:
+        index = {
+            "metadata": state_dict_split.metadata,
+            "weight_map": state_dict_split.tensor_to_filename,
+        }
+        save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+        save_index_file = os.path.join(output_dir, save_index_file)
+        with open(save_index_file, "w", encoding="utf-8") as f:
+            content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+            f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument("output_dir",
+                        type=str,
+                        help="directory to the pytorch fp32 state_dict output files"
+                        "(e.g. path/checkpoint-12-output/)")
+    parser.add_argument(
+        "--max_shard_size",
+        type=str,
+        default="5GB",
+        help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+        "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+        "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+        "without CPU OOM issues.")
+    parser.add_argument(
+        "--safe_serialization",
+        default=False,
+        action='store_true',
+        help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_dir,
+                                               max_shard_size=args.max_shard_size,
+                                               safe_serialization=args.safe_serialization,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)