Upload 16 files

SMC_Random/README.md

@@ -0,0 +1,202 @@
+---
+base_model: /home/bingxing2/ailab/gaoben/models/Mistral-7B-Instruct/Mistral-7B-Instruct-v0.2/AI-ModelScope/Mistral-7B-Instruct-v0___2
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.15.1

SMC_Random/adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/bingxing2/ailab/gaoben/models/Mistral-7B-Instruct/Mistral-7B-Instruct-v0.2/AI-ModelScope/Mistral-7B-Instruct-v0___2",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "down_proj",
+    "o_proj",
+    "v_proj",
+    "k_proj",
+    "q_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

SMC_Random/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6faf811cebf1ae0e2f8639acc83a37862a5eac41ea6845ce05a333b84bf9986
+size 83946192

SMC_Random/latest

@@ -0,0 +1 @@
+global_step800

SMC_Random/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a81e3916b1392c4c49afb171dee5415c15f5a5a5af8749b28195fcfa0596699c
+size 15024

SMC_Random/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a781038dd714b87b8adb1aac8dbc8217ceb607428a992133954ad522365236e
+size 15024

SMC_Random/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9446c3db15f382a5546f13622787fc99392a5e0bc8a9ca2da1838de7ab621a37
+size 15024

SMC_Random/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f11e7a6b3faa884fc23044e3772ff9dd72c257f02e121665061e2a03d518bd9
+size 15024

SMC_Random/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:540c398d4b2cccd6c98de20baf64b1c0da7fda2ec2a996ff6b01eec2bb4dd770
+size 1064

SMC_Random/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

SMC_Random/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

SMC_Random/tokenizer_config.json

@@ -0,0 +1,47 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [],
+  "bos_token": "<s>",
+  "chat_template": "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ message['content'] + eos_token}}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "split_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

SMC_Random/trainer_state.json

@@ -0,0 +1,674 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 12.698412698412698,
+  "eval_steps": 500,
+  "global_step": 800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.15873015873015872,
+      "grad_norm": 0.42645490169525146,
+      "learning_rate": 4.999720254525684e-05,
+      "loss": 1.3067,
+      "num_input_tokens_seen": 269280,
+      "step": 10
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 0.10797163844108582,
+      "learning_rate": 4.9987533135093934e-05,
+      "loss": 0.2064,
+      "num_input_tokens_seen": 536656,
+      "step": 20
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.10832954943180084,
+      "learning_rate": 4.997095990396411e-05,
+      "loss": 0.2025,
+      "num_input_tokens_seen": 804720,
+      "step": 30
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.11103782057762146,
+      "learning_rate": 4.994748743089566e-05,
+      "loss": 0.2011,
+      "num_input_tokens_seen": 1073520,
+      "step": 40
+    },
+    {
+      "epoch": 0.7936507936507936,
+      "grad_norm": 0.09958792477846146,
+      "learning_rate": 4.9917122201112656e-05,
+      "loss": 0.2028,
+      "num_input_tokens_seen": 1341184,
+      "step": 50
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.6581681370735168,
+      "learning_rate": 4.9879872604243184e-05,
+      "loss": 0.1993,
+      "num_input_tokens_seen": 1609968,
+      "step": 60
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.2799030542373657,
+      "learning_rate": 4.983574893200139e-05,
+      "loss": 0.1979,
+      "num_input_tokens_seen": 1878240,
+      "step": 70
+    },
+    {
+      "epoch": 1.2698412698412698,
+      "grad_norm": 0.17586013674736023,
+      "learning_rate": 4.978476337534393e-05,
+      "loss": 0.1931,
+      "num_input_tokens_seen": 2146528,
+      "step": 80
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.24899278581142426,
+      "learning_rate": 4.972693002110176e-05,
+      "loss": 0.1931,
+      "num_input_tokens_seen": 2415696,
+      "step": 90
+    },
+    {
+      "epoch": 1.5873015873015874,
+      "grad_norm": 0.16181747615337372,
+      "learning_rate": 4.9662264848088034e-05,
+      "loss": 0.192,
+      "num_input_tokens_seen": 2683600,
+      "step": 100
+    },
+    {
+      "epoch": 1.746031746031746,
+      "grad_norm": 0.18402352929115295,
+      "learning_rate": 4.959078572268337e-05,
+      "loss": 0.1874,
+      "num_input_tokens_seen": 2950720,
+      "step": 110
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.2943824827671051,
+      "learning_rate": 4.951251239389948e-05,
+      "loss": 0.1871,
+      "num_input_tokens_seen": 3219792,
+      "step": 120
+    },
+    {
+      "epoch": 2.0634920634920633,
+      "grad_norm": 0.18450967967510223,
+      "learning_rate": 4.942746648792274e-05,
+      "loss": 0.1887,
+      "num_input_tokens_seen": 3488400,
+      "step": 130
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.2516356408596039,
+      "learning_rate": 4.9335671502139024e-05,
+      "loss": 0.1876,
+      "num_input_tokens_seen": 3757952,
+      "step": 140
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.23607608675956726,
+      "learning_rate": 4.9237152798641696e-05,
+      "loss": 0.1843,
+      "num_input_tokens_seen": 4025536,
+      "step": 150
+    },
+    {
+      "epoch": 2.5396825396825395,
+      "grad_norm": 0.1812293380498886,
+      "learning_rate": 4.9131937597224185e-05,
+      "loss": 0.1791,
+      "num_input_tokens_seen": 4294240,
+      "step": 160
+    },
+    {
+      "epoch": 2.6984126984126986,
+      "grad_norm": 0.1874535083770752,
+      "learning_rate": 4.902005496785951e-05,
+      "loss": 0.1851,
+      "num_input_tokens_seen": 4563376,
+      "step": 170
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.25721630454063416,
+      "learning_rate": 4.8901535822668446e-05,
+      "loss": 0.1836,
+      "num_input_tokens_seen": 4831168,
+      "step": 180
+    },
+    {
+      "epoch": 3.015873015873016,
+      "grad_norm": 0.22797122597694397,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.1834,
+      "num_input_tokens_seen": 5098496,
+      "step": 190
+    },
+    {
+      "epoch": 3.1746031746031744,
+      "grad_norm": 0.16337507963180542,
+      "learning_rate": 4.8644720792278264e-05,
+      "loss": 0.186,
+      "num_input_tokens_seen": 5368864,
+      "step": 200
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.17769697308540344,
+      "learning_rate": 4.850649586266255e-05,
+      "loss": 0.1803,
+      "num_input_tokens_seen": 5637456,
+      "step": 210
+    },
+    {
+      "epoch": 3.492063492063492,
+      "grad_norm": 0.2481444925069809,
+      "learning_rate": 4.836177630878289e-05,
+      "loss": 0.1798,
+      "num_input_tokens_seen": 5905104,
+      "step": 220
+    },
+    {
+      "epoch": 3.6507936507936507,
+      "grad_norm": 0.22745923697948456,
+      "learning_rate": 4.821060211529424e-05,
+      "loss": 0.1815,
+      "num_input_tokens_seen": 6174032,
+      "step": 230
+    },
+    {
+      "epoch": 3.8095238095238093,
+      "grad_norm": 0.16727988421916962,
+      "learning_rate": 4.8053015050207915e-05,
+      "loss": 0.1811,
+      "num_input_tokens_seen": 6442896,
+      "step": 240
+    },
+    {
+      "epoch": 3.9682539682539684,
+      "grad_norm": 0.3471706807613373,
+      "learning_rate": 4.7889058653351485e-05,
+      "loss": 0.1795,
+      "num_input_tokens_seen": 6710352,
+      "step": 250
+    },
+    {
+      "epoch": 4.1269841269841265,
+      "grad_norm": 0.23989547789096832,
+      "learning_rate": 4.771877822433911e-05,
+      "loss": 0.1769,
+      "num_input_tokens_seen": 6977744,
+      "step": 260
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.23704519867897034,
+      "learning_rate": 4.754222081005574e-05,
+      "loss": 0.174,
+      "num_input_tokens_seen": 7246272,
+      "step": 270
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.2272966355085373,
+      "learning_rate": 4.7359435191658425e-05,
+      "loss": 0.1716,
+      "num_input_tokens_seen": 7512592,
+      "step": 280
+    },
+    {
+      "epoch": 4.603174603174603,
+      "grad_norm": 0.23121878504753113,
+      "learning_rate": 4.717047187109861e-05,
+      "loss": 0.1804,
+      "num_input_tokens_seen": 7780144,
+      "step": 290
+    },
+    {
+      "epoch": 4.761904761904762,
+      "grad_norm": 0.31674066185951233,
+      "learning_rate": 4.697538305716885e-05,
+      "loss": 0.1784,
+      "num_input_tokens_seen": 8049392,
+      "step": 300
+    },
+    {
+      "epoch": 4.920634920634921,
+      "grad_norm": 0.2399132400751114,
+      "learning_rate": 4.6774222651078106e-05,
+      "loss": 0.1796,
+      "num_input_tokens_seen": 8316912,
+      "step": 310
+    },
+    {
+      "epoch": 5.079365079365079,
+      "grad_norm": 0.2677905261516571,
+      "learning_rate": 4.656704623155922e-05,
+      "loss": 0.1736,
+      "num_input_tokens_seen": 8586544,
+      "step": 320
+    },
+    {
+      "epoch": 5.238095238095238,
+      "grad_norm": 0.33959662914276123,
+      "learning_rate": 4.6353911039513145e-05,
+      "loss": 0.1766,
+      "num_input_tokens_seen": 8855680,
+      "step": 330
+    },
+    {
+      "epoch": 5.396825396825397,
+      "grad_norm": 0.26891693472862244,
+      "learning_rate": 4.613487596219376e-05,
+      "loss": 0.1724,
+      "num_input_tokens_seen": 9123808,
+      "step": 340
+    },
+    {
+      "epoch": 5.555555555555555,
+      "grad_norm": 0.2796987295150757,
+      "learning_rate": 4.591000151693789e-05,
+      "loss": 0.1721,
+      "num_input_tokens_seen": 9392560,
+      "step": 350
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.257348895072937,
+      "learning_rate": 4.567934983444495e-05,
+      "loss": 0.1718,
+      "num_input_tokens_seen": 9660480,
+      "step": 360
+    },
+    {
+      "epoch": 5.8730158730158735,
+      "grad_norm": 0.2910774052143097,
+      "learning_rate": 4.544298464161079e-05,
+      "loss": 0.1718,
+      "num_input_tokens_seen": 9927936,
+      "step": 370
+    },
+    {
+      "epoch": 6.031746031746032,
+      "grad_norm": 0.3452795445919037,
+      "learning_rate": 4.520097124392055e-05,
+      "loss": 0.1711,
+      "num_input_tokens_seen": 10197520,
+      "step": 380
+    },
+    {
+      "epoch": 6.190476190476191,
+      "grad_norm": 0.46368861198425293,
+      "learning_rate": 4.49533765074054e-05,
+      "loss": 0.1652,
+      "num_input_tokens_seen": 10466240,
+      "step": 390
+    },
+    {
+      "epoch": 6.349206349206349,
+      "grad_norm": 0.42205390334129333,
+      "learning_rate": 4.4700268840168045e-05,
+      "loss": 0.1677,
+      "num_input_tokens_seen": 10734496,
+      "step": 400
+    },
+    {
+      "epoch": 6.507936507936508,
+      "grad_norm": 0.25223520398139954,
+      "learning_rate": 4.444171817348225e-05,
+      "loss": 0.1684,
+      "num_input_tokens_seen": 11004416,
+      "step": 410
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 0.4380488991737366,
+      "learning_rate": 4.417779594247143e-05,
+      "loss": 0.1655,
+      "num_input_tokens_seen": 11272656,
+      "step": 420
+    },
+    {
+      "epoch": 6.825396825396825,
+      "grad_norm": 0.2701490819454193,
+      "learning_rate": 4.3908575066371835e-05,
+      "loss": 0.1722,
+      "num_input_tokens_seen": 11540112,
+      "step": 430
+    },
+    {
+      "epoch": 6.984126984126984,
+      "grad_norm": 0.3422671854496002,
+      "learning_rate": 4.363412992838566e-05,
+      "loss": 0.1676,
+      "num_input_tokens_seen": 11808816,
+      "step": 440
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.6143015623092651,
+      "learning_rate": 4.335453635512961e-05,
+      "loss": 0.1538,
+      "num_input_tokens_seen": 12077648,
+      "step": 450
+    },
+    {
+      "epoch": 7.301587301587301,
+      "grad_norm": 0.44244784116744995,
+      "learning_rate": 4.306987159568479e-05,
+      "loss": 0.1572,
+      "num_input_tokens_seen": 12346240,
+      "step": 460
+    },
+    {
+      "epoch": 7.4603174603174605,
+      "grad_norm": 0.441853404045105,
+      "learning_rate": 4.278021430025343e-05,
+      "loss": 0.1587,
+      "num_input_tokens_seen": 12614864,
+      "step": 470
+    },
+    {
+      "epoch": 7.619047619047619,
+      "grad_norm": 0.520702600479126,
+      "learning_rate": 4.248564449842864e-05,
+      "loss": 0.1616,
+      "num_input_tokens_seen": 12883088,
+      "step": 480
+    },
+    {
+      "epoch": 7.777777777777778,
+      "grad_norm": 0.473958283662796,
+      "learning_rate": 4.2186243577082954e-05,
+      "loss": 0.1602,
+      "num_input_tokens_seen": 13151264,
+      "step": 490
+    },
+    {
+      "epoch": 7.936507936507937,
+      "grad_norm": 0.4550235867500305,
+      "learning_rate": 4.1882094257881885e-05,
+      "loss": 0.1597,
+      "num_input_tokens_seen": 13419344,
+      "step": 500
+    },
+    {
+      "epoch": 8.095238095238095,
+      "grad_norm": 0.7338590025901794,
+      "learning_rate": 4.157328057442874e-05,
+      "loss": 0.1473,
+      "num_input_tokens_seen": 13686752,
+      "step": 510
+    },
+    {
+      "epoch": 8.253968253968253,
+      "grad_norm": 0.6510297060012817,
+      "learning_rate": 4.1259887849046906e-05,
+      "loss": 0.1363,
+      "num_input_tokens_seen": 13954352,
+      "step": 520
+    },
+    {
+      "epoch": 8.412698412698413,
+      "grad_norm": 0.767859160900116,
+      "learning_rate": 4.0942002669206085e-05,
+      "loss": 0.1408,
+      "num_input_tokens_seen": 14222352,
+      "step": 530
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 0.7285030484199524,
+      "learning_rate": 4.0619712863599e-05,
+      "loss": 0.1422,
+      "num_input_tokens_seen": 14491920,
+      "step": 540
+    },
+    {
+      "epoch": 8.73015873015873,
+      "grad_norm": 0.6987579464912415,
+      "learning_rate": 4.029310747787516e-05,
+      "loss": 0.1483,
+      "num_input_tokens_seen": 14760400,
+      "step": 550
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 0.7618018984794617,
+      "learning_rate": 3.996227675003834e-05,
+      "loss": 0.1437,
+      "num_input_tokens_seen": 15029280,
+      "step": 560
+    },
+    {
+      "epoch": 9.047619047619047,
+      "grad_norm": 0.7082319855690002,
+      "learning_rate": 3.962731208551474e-05,
+      "loss": 0.1386,
+      "num_input_tokens_seen": 15298416,
+      "step": 570
+    },
+    {
+      "epoch": 9.206349206349206,
+      "grad_norm": 0.9523563385009766,
+      "learning_rate": 3.928830603189844e-05,
+      "loss": 0.1034,
+      "num_input_tokens_seen": 15567104,
+      "step": 580
+    },
+    {
+      "epoch": 9.365079365079366,
+      "grad_norm": 1.1607928276062012,
+      "learning_rate": 3.894535225338143e-05,
+      "loss": 0.1073,
+      "num_input_tokens_seen": 15835952,
+      "step": 590
+    },
+    {
+      "epoch": 9.523809523809524,
+      "grad_norm": 1.0483174324035645,
+      "learning_rate": 3.859854550487506e-05,
+      "loss": 0.1124,
+      "num_input_tokens_seen": 16103648,
+      "step": 600
+    },
+    {
+      "epoch": 9.682539682539682,
+      "grad_norm": 0.9111513495445251,
+      "learning_rate": 3.824798160583012e-05,
+      "loss": 0.1202,
+      "num_input_tokens_seen": 16373888,
+      "step": 610
+    },
+    {
+      "epoch": 9.841269841269842,
+      "grad_norm": 1.031439185142517,
+      "learning_rate": 3.789375741376286e-05,
+      "loss": 0.1194,
+      "num_input_tokens_seen": 16642320,
+      "step": 620
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.9815431237220764,
+      "learning_rate": 3.7535970797494136e-05,
+      "loss": 0.117,
+      "num_input_tokens_seen": 16910032,
+      "step": 630
+    },
+    {
+      "epoch": 10.158730158730158,
+      "grad_norm": 1.4907585382461548,
+      "learning_rate": 3.717472061010918e-05,
+      "loss": 0.0739,
+      "num_input_tokens_seen": 17178576,
+      "step": 640
+    },
+    {
+      "epoch": 10.317460317460318,
+      "grad_norm": 1.1762831211090088,
+      "learning_rate": 3.681010666164546e-05,
+      "loss": 0.0704,
+      "num_input_tokens_seen": 17448288,
+      "step": 650
+    },
+    {
+      "epoch": 10.476190476190476,
+      "grad_norm": 1.2105902433395386,
+      "learning_rate": 3.644222969151605e-05,
+      "loss": 0.0735,
+      "num_input_tokens_seen": 17716784,
+      "step": 660
+    },
+    {
+      "epoch": 10.634920634920634,
+      "grad_norm": 1.1394544839859009,
+      "learning_rate": 3.607119134067629e-05,
+      "loss": 0.077,
+      "num_input_tokens_seen": 17984944,
+      "step": 670
+    },
+    {
+      "epoch": 10.793650793650794,
+      "grad_norm": 1.2243598699569702,
+      "learning_rate": 3.569709412354136e-05,
+      "loss": 0.0763,
+      "num_input_tokens_seen": 18252080,
+      "step": 680
+    },
+    {
+      "epoch": 10.952380952380953,
+      "grad_norm": 1.0364540815353394,
+      "learning_rate": 3.5320041399662494e-05,
+      "loss": 0.0762,
+      "num_input_tokens_seen": 18520464,
+      "step": 690
+    },
+    {
+      "epoch": 11.11111111111111,
+      "grad_norm": 1.0455269813537598,
+      "learning_rate": 3.494013734516971e-05,
+      "loss": 0.0514,
+      "num_input_tokens_seen": 18786528,
+      "step": 700
+    },
+    {
+      "epoch": 11.26984126984127,
+      "grad_norm": 1.2155787944793701,
+      "learning_rate": 3.4557486923988924e-05,
+      "loss": 0.0375,
+      "num_input_tokens_seen": 19055536,
+      "step": 710
+    },
+    {
+      "epoch": 11.428571428571429,
+      "grad_norm": 1.1954303979873657,
+      "learning_rate": 3.4172195858841404e-05,
+      "loss": 0.0389,
+      "num_input_tokens_seen": 19324304,
+      "step": 720
+    },
+    {
+      "epoch": 11.587301587301587,
+      "grad_norm": 1.1928291320800781,
+      "learning_rate": 3.378437060203357e-05,
+      "loss": 0.0374,
+      "num_input_tokens_seen": 19593552,
+      "step": 730
+    },
+    {
+      "epoch": 11.746031746031747,
+      "grad_norm": 1.192438006401062,
+      "learning_rate": 3.3394118306045217e-05,
+      "loss": 0.0426,
+      "num_input_tokens_seen": 19862784,
+      "step": 740
+    },
+    {
+      "epoch": 11.904761904761905,
+      "grad_norm": 1.1554771661758423,
+      "learning_rate": 3.3001546793924285e-05,
+      "loss": 0.0432,
+      "num_input_tokens_seen": 20131584,
+      "step": 750
+    },
+    {
+      "epoch": 12.063492063492063,
+      "grad_norm": 0.7850580215454102,
+      "learning_rate": 3.260676452949641e-05,
+      "loss": 0.0348,
+      "num_input_tokens_seen": 20401120,
+      "step": 760
+    },
+    {
+      "epoch": 12.222222222222221,
+      "grad_norm": 0.6133368611335754,
+      "learning_rate": 3.22098805873973e-05,
+      "loss": 0.0165,
+      "num_input_tokens_seen": 20670080,
+      "step": 770
+    },
+    {
+      "epoch": 12.380952380952381,
+      "grad_norm": 0.9954155087471008,
+      "learning_rate": 3.1811004622936525e-05,
+      "loss": 0.0192,
+      "num_input_tokens_seen": 20938000,
+      "step": 780
+    },
+    {
+      "epoch": 12.53968253968254,
+      "grad_norm": 0.9651346206665039,
+      "learning_rate": 3.141024684180071e-05,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 21206432,
+      "step": 790
+    },
+    {
+      "epoch": 12.698412698412698,
+      "grad_norm": 1.0618289709091187,
+      "learning_rate": 3.10077179696048e-05,
+      "loss": 0.0231,
+      "num_input_tokens_seen": 21476960,
+      "step": 800
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1890,
+  "num_input_tokens_seen": 21476960,
+  "num_train_epochs": 30,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.216970364477768e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

SMC_Random/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68e8b6ca583bc5f45b95ebaf4fbbb9946e991bda9430768ff877861d9594f8a8
+size 7480

SMC_Random/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)