Training in progress, step 100, checkpoint

last-checkpoint/README.md

@@ -0,0 +1,204 @@
+---
+library_name: peft
+base_model: scb10x/typhoon-7b
+---
+
+# 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 0.10.0

last-checkpoint/adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "scb10x/typhoon-7b",
+  "bias": "none",
+  "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_dropout": 0.1,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "k_proj",
+    "q_proj",
+    "down_proj",
+    "gate_proj",
+    "o_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

last-checkpoint/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbc779994213d871e1f4070d7c034d94cf7454b01af940bf22136e01d1a1a59e
+size 42002584

last-checkpoint/global_step100/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37db02300cf84f2628252409e2eec85a0a16ed24e9a1a956fda27fef6bd9d65f
+size 251710672

last-checkpoint/global_step100/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56b171df51745eee09fcc81a6fd53f574f41702e91c3d62fa11c7976f1389eaf
+size 153726841

last-checkpoint/latest

@@ -0,0 +1 @@
+global_step100

last-checkpoint/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c514d3eff35e36088fb0b96868a3f7a54f5fe77734350e00ca17637049d51d54
+size 14244

last-checkpoint/trainer_state.json

@@ -0,0 +1,721 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.10982976386600769,
+  "eval_steps": 1000,
+  "global_step": 100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.704396157183714,
+      "learning_rate": 0.0,
+      "loss": 2.6917,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.261265634920543,
+      "learning_rate": 6.020599913279623e-05,
+      "loss": 2.0455,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.252464584699219,
+      "learning_rate": 9.542425094393248e-05,
+      "loss": 2.1024,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.478072090774749,
+      "learning_rate": 0.00012041199826559246,
+      "loss": 1.9419,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.996421444202911,
+      "learning_rate": 0.00013979400086720374,
+      "loss": 1.2142,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.62925923224712,
+      "learning_rate": 0.00015563025007672872,
+      "loss": 1.4192,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9971192119861545,
+      "learning_rate": 0.00016901960800285134,
+      "loss": 1.1698,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.448410811680446,
+      "learning_rate": 0.00018061799739838867,
+      "loss": 1.0663,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.664012698560983,
+      "learning_rate": 0.00019084850188786495,
+      "loss": 1.0346,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9872198688383325,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 1.2899,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.492858411635369,
+      "learning_rate": 0.0002,
+      "loss": 0.733,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.117272508239465,
+      "learning_rate": 0.00019931271477663232,
+      "loss": 0.7529,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.57016142244955,
+      "learning_rate": 0.0001986254295532646,
+      "loss": 0.9221,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.106703686957636,
+      "learning_rate": 0.00019793814432989693,
+      "loss": 1.1292,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.4331559960790115,
+      "learning_rate": 0.00019725085910652924,
+      "loss": 0.9505,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.871898936371301,
+      "learning_rate": 0.0001965635738831615,
+      "loss": 1.0894,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.546128257668831,
+      "learning_rate": 0.00019587628865979381,
+      "loss": 1.0915,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.426135233570764,
+      "learning_rate": 0.00019518900343642613,
+      "loss": 1.2174,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7481567939197316,
+      "learning_rate": 0.00019450171821305842,
+      "loss": 0.8281,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.264747615275931,
+      "learning_rate": 0.00019381443298969073,
+      "loss": 1.1069,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.850720515264914,
+      "learning_rate": 0.00019312714776632305,
+      "loss": 1.1216,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.583277384098615,
+      "learning_rate": 0.00019243986254295533,
+      "loss": 1.0751,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.923174586611154,
+      "learning_rate": 0.00019175257731958765,
+      "loss": 1.1915,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.21422328394722,
+      "learning_rate": 0.00019106529209621996,
+      "loss": 1.1412,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9312652393524803,
+      "learning_rate": 0.00019037800687285222,
+      "loss": 0.9699,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.511161102577495,
+      "learning_rate": 0.00018969072164948454,
+      "loss": 1.1289,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9233152624949454,
+      "learning_rate": 0.00018900343642611685,
+      "loss": 0.8803,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.413459689226915,
+      "learning_rate": 0.00018831615120274914,
+      "loss": 1.0594,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2106489602848685,
+      "learning_rate": 0.00018762886597938145,
+      "loss": 1.1003,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.568263346040736,
+      "learning_rate": 0.00018694158075601377,
+      "loss": 1.0816,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.12202434010765,
+      "learning_rate": 0.00018625429553264605,
+      "loss": 0.9053,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.282623293516478,
+      "learning_rate": 0.00018556701030927837,
+      "loss": 1.206,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.45030819651884,
+      "learning_rate": 0.00018487972508591068,
+      "loss": 0.9049,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.871975082534186,
+      "learning_rate": 0.00018419243986254294,
+      "loss": 1.1436,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.715014058776802,
+      "learning_rate": 0.00018350515463917526,
+      "loss": 1.3847,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3808462582338707,
+      "learning_rate": 0.00018281786941580757,
+      "loss": 0.7817,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8976964456454866,
+      "learning_rate": 0.00018213058419243986,
+      "loss": 1.0001,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.355867444757705,
+      "learning_rate": 0.00018144329896907217,
+      "loss": 0.9259,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.178991855035252,
+      "learning_rate": 0.0001807560137457045,
+      "loss": 0.794,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.14441205427679,
+      "learning_rate": 0.00018006872852233677,
+      "loss": 1.0384,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7176836552561974,
+      "learning_rate": 0.0001793814432989691,
+      "loss": 0.75,
+      "step": 41
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.011588210132538,
+      "learning_rate": 0.0001786941580756014,
+      "loss": 1.3034,
+      "step": 42
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.111725152950483,
+      "learning_rate": 0.00017800687285223366,
+      "loss": 1.3023,
+      "step": 43
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.021959822322685,
+      "learning_rate": 0.00017731958762886598,
+      "loss": 0.9624,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.943541786297702,
+      "learning_rate": 0.0001766323024054983,
+      "loss": 0.6659,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.534002060279327,
+      "learning_rate": 0.00017594501718213058,
+      "loss": 1.1354,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.108426720583385,
+      "learning_rate": 0.0001752577319587629,
+      "loss": 1.4258,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.235810963301271,
+      "learning_rate": 0.0001745704467353952,
+      "loss": 1.3918,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.207791712185817,
+      "learning_rate": 0.0001738831615120275,
+      "loss": 1.2796,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.050243728273998,
+      "learning_rate": 0.0001731958762886598,
+      "loss": 1.244,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.657182228931627,
+      "learning_rate": 0.00017250859106529212,
+      "loss": 1.2068,
+      "step": 51
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5652582299228643,
+      "learning_rate": 0.00017182130584192438,
+      "loss": 0.8897,
+      "step": 52
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.529701968193304,
+      "learning_rate": 0.0001711340206185567,
+      "loss": 1.1582,
+      "step": 53
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.49586529984068,
+      "learning_rate": 0.000170446735395189,
+      "loss": 1.1379,
+      "step": 54
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.845994813972232,
+      "learning_rate": 0.0001697594501718213,
+      "loss": 1.4724,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.732950229095352,
+      "learning_rate": 0.00016907216494845361,
+      "loss": 0.886,
+      "step": 56
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8567901480339426,
+      "learning_rate": 0.00016838487972508593,
+      "loss": 0.7763,
+      "step": 57
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.931804800794497,
+      "learning_rate": 0.00016769759450171822,
+      "loss": 1.1761,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7294788904744625,
+      "learning_rate": 0.00016701030927835053,
+      "loss": 0.6722,
+      "step": 59
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.688890615204965,
+      "learning_rate": 0.00016632302405498285,
+      "loss": 0.9642,
+      "step": 60
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.022317141495382,
+      "learning_rate": 0.00016563573883161513,
+      "loss": 0.8865,
+      "step": 61
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.55573614414137,
+      "learning_rate": 0.00016494845360824742,
+      "loss": 1.217,
+      "step": 62
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.5890564068710615,
+      "learning_rate": 0.00016426116838487973,
+      "loss": 1.0791,
+      "step": 63
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.435288078022071,
+      "learning_rate": 0.00016357388316151202,
+      "loss": 1.1501,
+      "step": 64
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.965664463354534,
+      "learning_rate": 0.00016288659793814434,
+      "loss": 1.4109,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.483997056692371,
+      "learning_rate": 0.00016219931271477665,
+      "loss": 1.0735,
+      "step": 66
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.9923544603584915,
+      "learning_rate": 0.00016151202749140894,
+      "loss": 0.9927,
+      "step": 67
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.143678077389599,
+      "learning_rate": 0.00016082474226804125,
+      "loss": 1.1299,
+      "step": 68
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.420855401236021,
+      "learning_rate": 0.00016013745704467357,
+      "loss": 1.0263,
+      "step": 69
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.02485050719212,
+      "learning_rate": 0.00015945017182130585,
+      "loss": 1.1168,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.524608717382107,
+      "learning_rate": 0.00015876288659793814,
+      "loss": 1.1015,
+      "step": 71
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9028726855074054,
+      "learning_rate": 0.00015807560137457046,
+      "loss": 1.091,
+      "step": 72
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.271315148375007,
+      "learning_rate": 0.00015738831615120274,
+      "loss": 0.9397,
+      "step": 73
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.5227145076761515,
+      "learning_rate": 0.00015670103092783506,
+      "loss": 1.0062,
+      "step": 74
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.344752651586208,
+      "learning_rate": 0.00015601374570446737,
+      "loss": 1.0027,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.102353206893142,
+      "learning_rate": 0.00015532646048109966,
+      "loss": 1.3915,
+      "step": 76
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.722343844673861,
+      "learning_rate": 0.00015463917525773197,
+      "loss": 1.1678,
+      "step": 77
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.017428516681286,
+      "learning_rate": 0.0001539518900343643,
+      "loss": 1.329,
+      "step": 78
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7313811619387516,
+      "learning_rate": 0.00015326460481099657,
+      "loss": 0.887,
+      "step": 79
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7179924970759406,
+      "learning_rate": 0.00015257731958762886,
+      "loss": 1.0153,
+      "step": 80
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.499981774187266,
+      "learning_rate": 0.00015189003436426118,
+      "loss": 1.3911,
+      "step": 81
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.955286629774712,
+      "learning_rate": 0.00015120274914089346,
+      "loss": 1.215,
+      "step": 82
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5953610996772984,
+      "learning_rate": 0.00015051546391752578,
+      "loss": 0.9343,
+      "step": 83
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8040779769157687,
+      "learning_rate": 0.0001498281786941581,
+      "loss": 1.062,
+      "step": 84
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.755142565965664,
+      "learning_rate": 0.00014914089347079038,
+      "loss": 1.2539,
+      "step": 85
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7550494212874264,
+      "learning_rate": 0.0001484536082474227,
+      "loss": 0.8594,
+      "step": 86
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.7330159455584635,
+      "learning_rate": 0.000147766323024055,
+      "loss": 0.9157,
+      "step": 87
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.438479463179745,
+      "learning_rate": 0.0001470790378006873,
+      "loss": 1.1235,
+      "step": 88
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.87421384650703,
+      "learning_rate": 0.00014639175257731958,
+      "loss": 0.9763,
+      "step": 89
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4000323333151683,
+      "learning_rate": 0.0001457044673539519,
+      "loss": 0.5867,
+      "step": 90
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.866412115354472,
+      "learning_rate": 0.00014501718213058418,
+      "loss": 1.5134,
+      "step": 91
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.023945137404242,
+      "learning_rate": 0.0001443298969072165,
+      "loss": 0.7887,
+      "step": 92
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7174822786238435,
+      "learning_rate": 0.00014364261168384881,
+      "loss": 0.925,
+      "step": 93
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.007705895268358,
+      "learning_rate": 0.0001429553264604811,
+      "loss": 0.9114,
+      "step": 94
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.852559484855566,
+      "learning_rate": 0.00014226804123711342,
+      "loss": 1.2322,
+      "step": 95
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9331976985714006,
+      "learning_rate": 0.00014158075601374573,
+      "loss": 1.021,
+      "step": 96
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7472774795317676,
+      "learning_rate": 0.00014089347079037802,
+      "loss": 0.9222,
+      "step": 97
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.430655545449399,
+      "learning_rate": 0.0001402061855670103,
+      "loss": 1.0648,
+      "step": 98
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8259440534841365,
+      "learning_rate": 0.00013951890034364262,
+      "loss": 0.9769,
+      "step": 99
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.784794509604932,
+      "learning_rate": 0.0001388316151202749,
+      "loss": 1.0134,
+      "step": 100
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 301,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 25,
+  "total_flos": 1192784269148160.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

last-checkpoint/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae82d01f8f983f767da1b994efbf6cd066f8d5d97f5b71458226d055ebeb49a4
+size 6648

last-checkpoint/zero_to_fp32.py

@@ -0,0 +1,592 @@
+#!/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 . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir):
+    """
+    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)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states)
+
+
+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):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None):
+    """
+    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``
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None):
+    """
+    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_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``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``
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+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_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    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("-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_file, tag=args.tag)