Upload folder using huggingface_hub

1_128_e5_3e-5/checkpoint-1156/README.md

@@ -0,0 +1,202 @@
+---
+base_model: ibm-granite/granite-3.3-8b-base
+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 0.15.2

1_128_e5_3e-5/checkpoint-1156/adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
+  "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": 256,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "k_proj",
+    "v_proj",
+    "o_proj",
+    "up_proj",
+    "q_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

1_128_e5_3e-5/checkpoint-1156/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d35f34f3fdf4b68ea4b832edd0f45bd39b3042dc116401eb5bf24aeb9c68113
+size 791751704

1_128_e5_3e-5/checkpoint-1156/latest

@@ -0,0 +1 @@
+global_step1156

1_128_e5_3e-5/checkpoint-1156/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97b0bb118885a1fbdfd3417ff25fed5004837e520ab8d0324d677cc44aa73c99
+size 15920

1_128_e5_3e-5/checkpoint-1156/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7261be3bf6365ed920049a95718d9a46f5b056b7781e8d10fed4adfa7f75b24
+size 15920

1_128_e5_3e-5/checkpoint-1156/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9d2b01be5f31b5e5ff426e6903a467e9cd6cab5e3c917c58eb5f4b1eacbfe3c
+size 15920

1_128_e5_3e-5/checkpoint-1156/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c7af3e4b9ca978d09f8ed3a7797b5a84507b94ccda1b21b14d58134b85af607
+size 15920

1_128_e5_3e-5/checkpoint-1156/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f0fb307aa934b7ee31b40b149015b03b91eec9036b5b728d8a46246188196ab
+size 15920

1_128_e5_3e-5/checkpoint-1156/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9d8d8225ce0d023e57f59274034cf2f98de0ea580286b0781b7ded8d8e70201
+size 15920

1_128_e5_3e-5/checkpoint-1156/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e231a441129f26a406d20de4fcac7cfa290d2bd925ab31307246636cab216c2
+size 15920

1_128_e5_3e-5/checkpoint-1156/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc307e710c708fe978cb8c3e8e82eee48597d1cf05c86637276f320ee6d20f72
+size 15920

1_128_e5_3e-5/checkpoint-1156/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec27f0c1fcf7a0613792b524c7f165dcbdb06ef3868c81265472a28ff927819e
+size 1064

1_128_e5_3e-5/checkpoint-1156/special_tokens_map.json

@@ -0,0 +1,45 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<reponame>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

1_128_e5_3e-5/checkpoint-1156/tokenizer_config.json

@@ -0,0 +1,188 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<fim_prefix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<fim_middle>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<fim_suffix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<fim_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<filename>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "<gh_stars>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "7": {
+      "content": "<issue_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "8": {
+      "content": "<issue_comment>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "9": {
+      "content": "<issue_closed>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "10": {
+      "content": "<jupyter_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "11": {
+      "content": "<jupyter_text>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "12": {
+      "content": "<jupyter_code>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "13": {
+      "content": "<jupyter_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "14": {
+      "content": "<empty_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "15": {
+      "content": "<commit_before>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "16": {
+      "content": "<commit_msg>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "17": {
+      "content": "<commit_after>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "18": {
+      "content": "<reponame>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<reponame>",
+  "padding_side": "left",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

1_128_e5_3e-5/checkpoint-1156/trainer_state.json

@@ -0,0 +1,1651 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 1156,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01733102253032929,
+      "grad_norm": 1.1325727701187134,
+      "learning_rate": 1.6438356164383561e-06,
+      "loss": 1.3757,
+      "step": 5
+    },
+    {
+      "epoch": 0.03466204506065858,
+      "grad_norm": 1.044039249420166,
+      "learning_rate": 3.6986301369863014e-06,
+      "loss": 1.3112,
+      "step": 10
+    },
+    {
+      "epoch": 0.05199306759098787,
+      "grad_norm": 0.621895968914032,
+      "learning_rate": 5.753424657534246e-06,
+      "loss": 1.3603,
+      "step": 15
+    },
+    {
+      "epoch": 0.06932409012131716,
+      "grad_norm": 0.6180606484413147,
+      "learning_rate": 7.808219178082192e-06,
+      "loss": 1.2902,
+      "step": 20
+    },
+    {
+      "epoch": 0.08665511265164645,
+      "grad_norm": 0.5306586027145386,
+      "learning_rate": 9.863013698630136e-06,
+      "loss": 1.2892,
+      "step": 25
+    },
+    {
+      "epoch": 0.10398613518197573,
+      "grad_norm": 0.661157488822937,
+      "learning_rate": 1.1917808219178083e-05,
+      "loss": 1.2568,
+      "step": 30
+    },
+    {
+      "epoch": 0.12131715771230503,
+      "grad_norm": 0.666887104511261,
+      "learning_rate": 1.3972602739726027e-05,
+      "loss": 1.2041,
+      "step": 35
+    },
+    {
+      "epoch": 0.1386481802426343,
+      "grad_norm": 0.48998892307281494,
+      "learning_rate": 1.6027397260273974e-05,
+      "loss": 1.235,
+      "step": 40
+    },
+    {
+      "epoch": 0.1559792027729636,
+      "grad_norm": 0.4912819266319275,
+      "learning_rate": 1.8082191780821916e-05,
+      "loss": 1.2399,
+      "step": 45
+    },
+    {
+      "epoch": 0.1733102253032929,
+      "grad_norm": 0.48102739453315735,
+      "learning_rate": 2.0136986301369863e-05,
+      "loss": 1.2004,
+      "step": 50
+    },
+    {
+      "epoch": 0.19064124783362218,
+      "grad_norm": 0.5094467997550964,
+      "learning_rate": 2.219178082191781e-05,
+      "loss": 1.1671,
+      "step": 55
+    },
+    {
+      "epoch": 0.20797227036395147,
+      "grad_norm": 0.5597960352897644,
+      "learning_rate": 2.4246575342465755e-05,
+      "loss": 1.2149,
+      "step": 60
+    },
+    {
+      "epoch": 0.22530329289428075,
+      "grad_norm": 0.48657160997390747,
+      "learning_rate": 2.6301369863013698e-05,
+      "loss": 1.1888,
+      "step": 65
+    },
+    {
+      "epoch": 0.24263431542461006,
+      "grad_norm": 0.6103832721710205,
+      "learning_rate": 2.8356164383561644e-05,
+      "loss": 1.1243,
+      "step": 70
+    },
+    {
+      "epoch": 0.25996533795493937,
+      "grad_norm": 0.4703650176525116,
+      "learning_rate": 2.9999960676460984e-05,
+      "loss": 1.1172,
+      "step": 75
+    },
+    {
+      "epoch": 0.2772963604852686,
+      "grad_norm": 0.5571750998497009,
+      "learning_rate": 2.9998584374244097e-05,
+      "loss": 1.1198,
+      "step": 80
+    },
+    {
+      "epoch": 0.29462738301559793,
+      "grad_norm": 0.5368012189865112,
+      "learning_rate": 2.999524210125035e-05,
+      "loss": 1.1335,
+      "step": 85
+    },
+    {
+      "epoch": 0.3119584055459272,
+      "grad_norm": 0.6115505695343018,
+      "learning_rate": 2.9989934295575147e-05,
+      "loss": 1.1163,
+      "step": 90
+    },
+    {
+      "epoch": 0.3292894280762565,
+      "grad_norm": 0.5922970175743103,
+      "learning_rate": 2.998266165295021e-05,
+      "loss": 1.1322,
+      "step": 95
+    },
+    {
+      "epoch": 0.3466204506065858,
+      "grad_norm": 0.6367848515510559,
+      "learning_rate": 2.9973425126652373e-05,
+      "loss": 1.0499,
+      "step": 100
+    },
+    {
+      "epoch": 0.36395147313691506,
+      "grad_norm": 0.5297245383262634,
+      "learning_rate": 2.9962225927378597e-05,
+      "loss": 1.0052,
+      "step": 105
+    },
+    {
+      "epoch": 0.38128249566724437,
+      "grad_norm": 0.6090821623802185,
+      "learning_rate": 2.9949065523087333e-05,
+      "loss": 1.102,
+      "step": 110
+    },
+    {
+      "epoch": 0.3986135181975737,
+      "grad_norm": 0.698509156703949,
+      "learning_rate": 2.9933945638806056e-05,
+      "loss": 1.0093,
+      "step": 115
+    },
+    {
+      "epoch": 0.41594454072790293,
+      "grad_norm": 0.706832766532898,
+      "learning_rate": 2.9916868256405185e-05,
+      "loss": 1.0042,
+      "step": 120
+    },
+    {
+      "epoch": 0.43327556325823224,
+      "grad_norm": 0.6957172155380249,
+      "learning_rate": 2.9897835614338295e-05,
+      "loss": 0.9757,
+      "step": 125
+    },
+    {
+      "epoch": 0.4506065857885615,
+      "grad_norm": 0.6912302374839783,
+      "learning_rate": 2.987685020734869e-05,
+      "loss": 0.9567,
+      "step": 130
+    },
+    {
+      "epoch": 0.4679376083188908,
+      "grad_norm": 0.7200044393539429,
+      "learning_rate": 2.985391478614244e-05,
+      "loss": 0.9602,
+      "step": 135
+    },
+    {
+      "epoch": 0.4852686308492201,
+      "grad_norm": 0.6456925272941589,
+      "learning_rate": 2.982903235702778e-05,
+      "loss": 0.9608,
+      "step": 140
+    },
+    {
+      "epoch": 0.5025996533795494,
+      "grad_norm": 0.7961540222167969,
+      "learning_rate": 2.9802206181521086e-05,
+      "loss": 0.8843,
+      "step": 145
+    },
+    {
+      "epoch": 0.5199306759098787,
+      "grad_norm": 0.6846944093704224,
+      "learning_rate": 2.9773439775919343e-05,
+      "loss": 0.9035,
+      "step": 150
+    },
+    {
+      "epoch": 0.537261698440208,
+      "grad_norm": 0.8010955452919006,
+      "learning_rate": 2.974273691083926e-05,
+      "loss": 0.9036,
+      "step": 155
+    },
+    {
+      "epoch": 0.5545927209705372,
+      "grad_norm": 0.8240249752998352,
+      "learning_rate": 2.971010161072301e-05,
+      "loss": 0.9011,
+      "step": 160
+    },
+    {
+      "epoch": 0.5719237435008665,
+      "grad_norm": 0.8053805828094482,
+      "learning_rate": 2.9675538153310732e-05,
+      "loss": 0.8949,
+      "step": 165
+    },
+    {
+      "epoch": 0.5892547660311959,
+      "grad_norm": 0.8074398040771484,
+      "learning_rate": 2.9639051069079794e-05,
+      "loss": 0.8769,
+      "step": 170
+    },
+    {
+      "epoch": 0.6065857885615251,
+      "grad_norm": 0.8516232371330261,
+      "learning_rate": 2.9600645140650985e-05,
+      "loss": 0.8376,
+      "step": 175
+    },
+    {
+      "epoch": 0.6239168110918544,
+      "grad_norm": 0.7514265179634094,
+      "learning_rate": 2.9560325402161598e-05,
+      "loss": 0.8566,
+      "step": 180
+    },
+    {
+      "epoch": 0.6412478336221837,
+      "grad_norm": 0.8322745561599731,
+      "learning_rate": 2.9518097138605574e-05,
+      "loss": 0.819,
+      "step": 185
+    },
+    {
+      "epoch": 0.658578856152513,
+      "grad_norm": 0.8268292546272278,
+      "learning_rate": 2.9473965885140774e-05,
+      "loss": 0.8155,
+      "step": 190
+    },
+    {
+      "epoch": 0.6759098786828422,
+      "grad_norm": 0.769992470741272,
+      "learning_rate": 2.9427937426363424e-05,
+      "loss": 0.7585,
+      "step": 195
+    },
+    {
+      "epoch": 0.6932409012131716,
+      "grad_norm": 0.9443396329879761,
+      "learning_rate": 2.938001779554991e-05,
+      "loss": 0.8124,
+      "step": 200
+    },
+    {
+      "epoch": 0.7105719237435009,
+      "grad_norm": 0.8200739026069641,
+      "learning_rate": 2.9330213273865936e-05,
+      "loss": 0.8034,
+      "step": 205
+    },
+    {
+      "epoch": 0.7279029462738301,
+      "grad_norm": 0.90153568983078,
+      "learning_rate": 2.927853038954322e-05,
+      "loss": 0.7209,
+      "step": 210
+    },
+    {
+      "epoch": 0.7452339688041595,
+      "grad_norm": 0.9012041687965393,
+      "learning_rate": 2.9224975917023778e-05,
+      "loss": 0.8571,
+      "step": 215
+    },
+    {
+      "epoch": 0.7625649913344887,
+      "grad_norm": 0.8709543943405151,
+      "learning_rate": 2.9169556876071967e-05,
+      "loss": 0.8261,
+      "step": 220
+    },
+    {
+      "epoch": 0.779896013864818,
+      "grad_norm": 0.9728206396102905,
+      "learning_rate": 2.911228053085434e-05,
+      "loss": 0.7364,
+      "step": 225
+    },
+    {
+      "epoch": 0.7972270363951474,
+      "grad_norm": 1.0054287910461426,
+      "learning_rate": 2.9053154388987493e-05,
+      "loss": 0.7678,
+      "step": 230
+    },
+    {
+      "epoch": 0.8145580589254766,
+      "grad_norm": 0.8341831564903259,
+      "learning_rate": 2.8992186200553975e-05,
+      "loss": 0.7069,
+      "step": 235
+    },
+    {
+      "epoch": 0.8318890814558059,
+      "grad_norm": 0.9587536454200745,
+      "learning_rate": 2.892938395708644e-05,
+      "loss": 0.7277,
+      "step": 240
+    },
+    {
+      "epoch": 0.8492201039861352,
+      "grad_norm": 0.9497514963150024,
+      "learning_rate": 2.886475589052013e-05,
+      "loss": 0.7346,
+      "step": 245
+    },
+    {
+      "epoch": 0.8665511265164645,
+      "grad_norm": 0.9413470029830933,
+      "learning_rate": 2.8798310472113877e-05,
+      "loss": 0.7299,
+      "step": 250
+    },
+    {
+      "epoch": 0.8838821490467937,
+      "grad_norm": 1.027370810508728,
+      "learning_rate": 2.8730056411339695e-05,
+      "loss": 0.6685,
+      "step": 255
+    },
+    {
+      "epoch": 0.901213171577123,
+      "grad_norm": 0.9176335334777832,
+      "learning_rate": 2.866000265474117e-05,
+      "loss": 0.7323,
+      "step": 260
+    },
+    {
+      "epoch": 0.9185441941074524,
+      "grad_norm": 0.9519228339195251,
+      "learning_rate": 2.858815838476078e-05,
+      "loss": 0.684,
+      "step": 265
+    },
+    {
+      "epoch": 0.9358752166377816,
+      "grad_norm": 1.0398930311203003,
+      "learning_rate": 2.8514533018536286e-05,
+      "loss": 0.7218,
+      "step": 270
+    },
+    {
+      "epoch": 0.9532062391681109,
+      "grad_norm": 1.0150400400161743,
+      "learning_rate": 2.8439136206666365e-05,
+      "loss": 0.6864,
+      "step": 275
+    },
+    {
+      "epoch": 0.9705372616984402,
+      "grad_norm": 1.4193578958511353,
+      "learning_rate": 2.8361977831945614e-05,
+      "loss": 0.66,
+      "step": 280
+    },
+    {
+      "epoch": 0.9878682842287695,
+      "grad_norm": 0.9296590089797974,
+      "learning_rate": 2.8283068008069188e-05,
+      "loss": 0.6661,
+      "step": 285
+    },
+    {
+      "epoch": 1.0034662045060658,
+      "grad_norm": 1.0106604099273682,
+      "learning_rate": 2.820241707830707e-05,
+      "loss": 0.6015,
+      "step": 290
+    },
+    {
+      "epoch": 1.0207972270363952,
+      "grad_norm": 1.0099163055419922,
+      "learning_rate": 2.8120035614148358e-05,
+      "loss": 0.5811,
+      "step": 295
+    },
+    {
+      "epoch": 1.0381282495667243,
+      "grad_norm": 1.0148612260818481,
+      "learning_rate": 2.803593441391555e-05,
+      "loss": 0.5606,
+      "step": 300
+    },
+    {
+      "epoch": 1.0554592720970537,
+      "grad_norm": 1.1285183429718018,
+      "learning_rate": 2.795012450134913e-05,
+      "loss": 0.593,
+      "step": 305
+    },
+    {
+      "epoch": 1.072790294627383,
+      "grad_norm": 1.0806018114089966,
+      "learning_rate": 2.7862617124162643e-05,
+      "loss": 0.5492,
+      "step": 310
+    },
+    {
+      "epoch": 1.0901213171577122,
+      "grad_norm": 1.0284126996994019,
+      "learning_rate": 2.7773423752568347e-05,
+      "loss": 0.5228,
+      "step": 315
+    },
+    {
+      "epoch": 1.1074523396880416,
+      "grad_norm": 1.3091356754302979,
+      "learning_rate": 2.768255607777373e-05,
+      "loss": 0.5503,
+      "step": 320
+    },
+    {
+      "epoch": 1.124783362218371,
+      "grad_norm": 0.9635022282600403,
+      "learning_rate": 2.7590026010449076e-05,
+      "loss": 0.5203,
+      "step": 325
+    },
+    {
+      "epoch": 1.1421143847487,
+      "grad_norm": 0.998070478439331,
+      "learning_rate": 2.7495845679166252e-05,
+      "loss": 0.5471,
+      "step": 330
+    },
+    {
+      "epoch": 1.1594454072790294,
+      "grad_norm": 1.2872036695480347,
+      "learning_rate": 2.7400027428808897e-05,
+      "loss": 0.505,
+      "step": 335
+    },
+    {
+      "epoch": 1.1767764298093588,
+      "grad_norm": 1.0663659572601318,
+      "learning_rate": 2.730258381895434e-05,
+      "loss": 0.5486,
+      "step": 340
+    },
+    {
+      "epoch": 1.194107452339688,
+      "grad_norm": 1.1138153076171875,
+      "learning_rate": 2.720352762222728e-05,
+      "loss": 0.5617,
+      "step": 345
+    },
+    {
+      "epoch": 1.2114384748700173,
+      "grad_norm": 1.1818647384643555,
+      "learning_rate": 2.7102871822625627e-05,
+      "loss": 0.5401,
+      "step": 350
+    },
+    {
+      "epoch": 1.2287694974003467,
+      "grad_norm": 1.0879274606704712,
+      "learning_rate": 2.700062961381856e-05,
+      "loss": 0.571,
+      "step": 355
+    },
+    {
+      "epoch": 1.2461005199306758,
+      "grad_norm": 1.346692442893982,
+      "learning_rate": 2.6896814397417174e-05,
+      "loss": 0.5403,
+      "step": 360
+    },
+    {
+      "epoch": 1.2634315424610052,
+      "grad_norm": 1.0724241733551025,
+      "learning_rate": 2.6791439781217815e-05,
+      "loss": 0.4927,
+      "step": 365
+    },
+    {
+      "epoch": 1.2807625649913346,
+      "grad_norm": 1.144758701324463,
+      "learning_rate": 2.6684519577418417e-05,
+      "loss": 0.5394,
+      "step": 370
+    },
+    {
+      "epoch": 1.2980935875216637,
+      "grad_norm": 1.071765661239624,
+      "learning_rate": 2.6576067800808026e-05,
+      "loss": 0.4743,
+      "step": 375
+    },
+    {
+      "epoch": 1.315424610051993,
+      "grad_norm": 1.0622762441635132,
+      "learning_rate": 2.646609866692981e-05,
+      "loss": 0.4881,
+      "step": 380
+    },
+    {
+      "epoch": 1.3327556325823224,
+      "grad_norm": 1.4975517988204956,
+      "learning_rate": 2.6354626590217705e-05,
+      "loss": 0.4954,
+      "step": 385
+    },
+    {
+      "epoch": 1.3500866551126516,
+      "grad_norm": 1.0805246829986572,
+      "learning_rate": 2.624166618210701e-05,
+      "loss": 0.458,
+      "step": 390
+    },
+    {
+      "epoch": 1.367417677642981,
+      "grad_norm": 1.0548033714294434,
+      "learning_rate": 2.6127232249119177e-05,
+      "loss": 0.4765,
+      "step": 395
+    },
+    {
+      "epoch": 1.38474870017331,
+      "grad_norm": 1.2031217813491821,
+      "learning_rate": 2.6011339790921012e-05,
+      "loss": 0.5447,
+      "step": 400
+    },
+    {
+      "epoch": 1.4020797227036395,
+      "grad_norm": 1.2651617527008057,
+      "learning_rate": 2.5894003998358553e-05,
+      "loss": 0.4935,
+      "step": 405
+    },
+    {
+      "epoch": 1.4194107452339688,
+      "grad_norm": 1.282373070716858,
+      "learning_rate": 2.5775240251465914e-05,
+      "loss": 0.4586,
+      "step": 410
+    },
+    {
+      "epoch": 1.436741767764298,
+      "grad_norm": 1.0972130298614502,
+      "learning_rate": 2.56550641174493e-05,
+      "loss": 0.4681,
+      "step": 415
+    },
+    {
+      "epoch": 1.4540727902946273,
+      "grad_norm": 1.1157981157302856,
+      "learning_rate": 2.5533491348646504e-05,
+      "loss": 0.4499,
+      "step": 420
+    },
+    {
+      "epoch": 1.4714038128249567,
+      "grad_norm": 1.1279178857803345,
+      "learning_rate": 2.541053788046215e-05,
+      "loss": 0.4645,
+      "step": 425
+    },
+    {
+      "epoch": 1.4887348353552858,
+      "grad_norm": 1.1415854692459106,
+      "learning_rate": 2.5286219829278914e-05,
+      "loss": 0.4214,
+      "step": 430
+    },
+    {
+      "epoch": 1.5060658578856152,
+      "grad_norm": 1.1717807054519653,
+      "learning_rate": 2.5160553490345038e-05,
+      "loss": 0.4397,
+      "step": 435
+    },
+    {
+      "epoch": 1.5233968804159446,
+      "grad_norm": 1.2159512042999268,
+      "learning_rate": 2.5033555335638388e-05,
+      "loss": 0.4619,
+      "step": 440
+    },
+    {
+      "epoch": 1.5407279029462737,
+      "grad_norm": 1.2196030616760254,
+      "learning_rate": 2.490524201170739e-05,
+      "loss": 0.4245,
+      "step": 445
+    },
+    {
+      "epoch": 1.558058925476603,
+      "grad_norm": 1.108749508857727,
+      "learning_rate": 2.4775630337489e-05,
+      "loss": 0.4329,
+      "step": 450
+    },
+    {
+      "epoch": 1.5753899480069324,
+      "grad_norm": 1.1284644603729248,
+      "learning_rate": 2.4644737302104156e-05,
+      "loss": 0.4155,
+      "step": 455
+    },
+    {
+      "epoch": 1.5927209705372616,
+      "grad_norm": 1.0621742010116577,
+      "learning_rate": 2.451258006263089e-05,
+      "loss": 0.4377,
+      "step": 460
+    },
+    {
+      "epoch": 1.610051993067591,
+      "grad_norm": 1.178806185722351,
+      "learning_rate": 2.4379175941855423e-05,
+      "loss": 0.4428,
+      "step": 465
+    },
+    {
+      "epoch": 1.6273830155979203,
+      "grad_norm": 1.216556191444397,
+      "learning_rate": 2.424454242600155e-05,
+      "loss": 0.4196,
+      "step": 470
+    },
+    {
+      "epoch": 1.6447140381282495,
+      "grad_norm": 1.2306348085403442,
+      "learning_rate": 2.4108697162438595e-05,
+      "loss": 0.4011,
+      "step": 475
+    },
+    {
+      "epoch": 1.6620450606585788,
+      "grad_norm": 1.5304279327392578,
+      "learning_rate": 2.397165795736824e-05,
+      "loss": 0.4211,
+      "step": 480
+    },
+    {
+      "epoch": 1.6793760831889082,
+      "grad_norm": 1.033577561378479,
+      "learning_rate": 2.3833442773490544e-05,
+      "loss": 0.3988,
+      "step": 485
+    },
+    {
+      "epoch": 1.6967071057192373,
+      "grad_norm": 1.2301174402236938,
+      "learning_rate": 2.369406972764945e-05,
+      "loss": 0.381,
+      "step": 490
+    },
+    {
+      "epoch": 1.7140381282495667,
+      "grad_norm": 1.1185941696166992,
+      "learning_rate": 2.3553557088458077e-05,
+      "loss": 0.3582,
+      "step": 495
+    },
+    {
+      "epoch": 1.731369150779896,
+      "grad_norm": 1.2030754089355469,
+      "learning_rate": 2.3411923273904104e-05,
+      "loss": 0.3738,
+      "step": 500
+    },
+    {
+      "epoch": 1.7487001733102252,
+      "grad_norm": 1.1300477981567383,
+      "learning_rate": 2.326918684893564e-05,
+      "loss": 0.3944,
+      "step": 505
+    },
+    {
+      "epoch": 1.7660311958405546,
+      "grad_norm": 1.1240997314453125,
+      "learning_rate": 2.312536652302774e-05,
+      "loss": 0.3737,
+      "step": 510
+    },
+    {
+      "epoch": 1.783362218370884,
+      "grad_norm": 1.1487401723861694,
+      "learning_rate": 2.298048114773005e-05,
+      "loss": 0.4144,
+      "step": 515
+    },
+    {
+      "epoch": 1.800693240901213,
+      "grad_norm": 1.1210594177246094,
+      "learning_rate": 2.2834549714195772e-05,
+      "loss": 0.3697,
+      "step": 520
+    },
+    {
+      "epoch": 1.8180242634315424,
+      "grad_norm": 1.1202471256256104,
+      "learning_rate": 2.26875913506924e-05,
+      "loss": 0.4059,
+      "step": 525
+    },
+    {
+      "epoch": 1.8353552859618718,
+      "grad_norm": 1.058485507965088,
+      "learning_rate": 2.2539625320094396e-05,
+      "loss": 0.4164,
+      "step": 530
+    },
+    {
+      "epoch": 1.852686308492201,
+      "grad_norm": 1.0952345132827759,
+      "learning_rate": 2.2390671017358307e-05,
+      "loss": 0.3537,
+      "step": 535
+    },
+    {
+      "epoch": 1.8700173310225303,
+      "grad_norm": 1.035676121711731,
+      "learning_rate": 2.2240747966980508e-05,
+      "loss": 0.3455,
+      "step": 540
+    },
+    {
+      "epoch": 1.8873483535528597,
+      "grad_norm": 1.5411173105239868,
+      "learning_rate": 2.2089875820437995e-05,
+      "loss": 0.3657,
+      "step": 545
+    },
+    {
+      "epoch": 1.9046793760831888,
+      "grad_norm": 1.1188682317733765,
+      "learning_rate": 2.1938074353612533e-05,
+      "loss": 0.3422,
+      "step": 550
+    },
+    {
+      "epoch": 1.9220103986135182,
+      "grad_norm": 1.0884416103363037,
+      "learning_rate": 2.178536346419847e-05,
+      "loss": 0.3412,
+      "step": 555
+    },
+    {
+      "epoch": 1.9393414211438476,
+      "grad_norm": 1.2015438079833984,
+      "learning_rate": 2.1631763169094628e-05,
+      "loss": 0.3559,
+      "step": 560
+    },
+    {
+      "epoch": 1.9566724436741767,
+      "grad_norm": 1.431586503982544,
+      "learning_rate": 2.1477293601780554e-05,
+      "loss": 0.3611,
+      "step": 565
+    },
+    {
+      "epoch": 1.974003466204506,
+      "grad_norm": 1.111899971961975,
+      "learning_rate": 2.132197500967743e-05,
+      "loss": 0.3566,
+      "step": 570
+    },
+    {
+      "epoch": 1.9913344887348354,
+      "grad_norm": 1.2778851985931396,
+      "learning_rate": 2.116582775149417e-05,
+      "loss": 0.3464,
+      "step": 575
+    },
+    {
+      "epoch": 2.0069324090121317,
+      "grad_norm": 1.0723752975463867,
+      "learning_rate": 2.1008872294558802e-05,
+      "loss": 0.2894,
+      "step": 580
+    },
+    {
+      "epoch": 2.024263431542461,
+      "grad_norm": 1.194329857826233,
+      "learning_rate": 2.0851129212135702e-05,
+      "loss": 0.2778,
+      "step": 585
+    },
+    {
+      "epoch": 2.0415944540727904,
+      "grad_norm": 1.1894909143447876,
+      "learning_rate": 2.0692619180728914e-05,
+      "loss": 0.2422,
+      "step": 590
+    },
+    {
+      "epoch": 2.0589254766031195,
+      "grad_norm": 1.8235598802566528,
+      "learning_rate": 2.0533362977371892e-05,
+      "loss": 0.2683,
+      "step": 595
+    },
+    {
+      "epoch": 2.0762564991334487,
+      "grad_norm": 1.3136845827102661,
+      "learning_rate": 2.0373381476904173e-05,
+      "loss": 0.2878,
+      "step": 600
+    },
+    {
+      "epoch": 2.0935875216637783,
+      "grad_norm": 1.1483287811279297,
+      "learning_rate": 2.02126956492351e-05,
+      "loss": 0.2553,
+      "step": 605
+    },
+    {
+      "epoch": 2.1109185441941074,
+      "grad_norm": 1.1638771295547485,
+      "learning_rate": 2.0051326556595192e-05,
+      "loss": 0.276,
+      "step": 610
+    },
+    {
+      "epoch": 2.1282495667244365,
+      "grad_norm": 1.2393336296081543,
+      "learning_rate": 1.9889295350775343e-05,
+      "loss": 0.2562,
+      "step": 615
+    },
+    {
+      "epoch": 2.145580589254766,
+      "grad_norm": 1.1414188146591187,
+      "learning_rate": 1.9726623270354314e-05,
+      "loss": 0.2365,
+      "step": 620
+    },
+    {
+      "epoch": 2.1629116117850953,
+      "grad_norm": 0.9609736204147339,
+      "learning_rate": 1.956333163791485e-05,
+      "loss": 0.2367,
+      "step": 625
+    },
+    {
+      "epoch": 2.1802426343154244,
+      "grad_norm": 1.1595170497894287,
+      "learning_rate": 1.9399441857248756e-05,
+      "loss": 0.2572,
+      "step": 630
+    },
+    {
+      "epoch": 2.197573656845754,
+      "grad_norm": 1.1949115991592407,
+      "learning_rate": 1.9234975410551397e-05,
+      "loss": 0.3066,
+      "step": 635
+    },
+    {
+      "epoch": 2.214904679376083,
+      "grad_norm": 1.422609806060791,
+      "learning_rate": 1.9069953855605805e-05,
+      "loss": 0.2792,
+      "step": 640
+    },
+    {
+      "epoch": 2.2322357019064123,
+      "grad_norm": 1.1198484897613525,
+      "learning_rate": 1.8904398822957007e-05,
+      "loss": 0.2647,
+      "step": 645
+    },
+    {
+      "epoch": 2.249566724436742,
+      "grad_norm": 1.0784331560134888,
+      "learning_rate": 1.873833201307673e-05,
+      "loss": 0.2449,
+      "step": 650
+    },
+    {
+      "epoch": 2.266897746967071,
+      "grad_norm": 1.2669086456298828,
+      "learning_rate": 1.8571775193518974e-05,
+      "loss": 0.2581,
+      "step": 655
+    },
+    {
+      "epoch": 2.2842287694974,
+      "grad_norm": 1.20986008644104,
+      "learning_rate": 1.840475019606677e-05,
+      "loss": 0.2313,
+      "step": 660
+    },
+    {
+      "epoch": 2.3015597920277298,
+      "grad_norm": 1.016635775566101,
+      "learning_rate": 1.8237278913870558e-05,
+      "loss": 0.2332,
+      "step": 665
+    },
+    {
+      "epoch": 2.318890814558059,
+      "grad_norm": 1.004381775856018,
+      "learning_rate": 1.8069383298578474e-05,
+      "loss": 0.2391,
+      "step": 670
+    },
+    {
+      "epoch": 2.336221837088388,
+      "grad_norm": 1.2095917463302612,
+      "learning_rate": 1.7901085357459e-05,
+      "loss": 0.2363,
+      "step": 675
+    },
+    {
+      "epoch": 2.3535528596187176,
+      "grad_norm": 1.021548867225647,
+      "learning_rate": 1.7732407150516324e-05,
+      "loss": 0.2394,
+      "step": 680
+    },
+    {
+      "epoch": 2.3708838821490468,
+      "grad_norm": 1.1039049625396729,
+      "learning_rate": 1.7563370787598766e-05,
+      "loss": 0.2595,
+      "step": 685
+    },
+    {
+      "epoch": 2.388214904679376,
+      "grad_norm": 1.1957190036773682,
+      "learning_rate": 1.739399842550069e-05,
+      "loss": 0.239,
+      "step": 690
+    },
+    {
+      "epoch": 2.4055459272097055,
+      "grad_norm": 1.2246172428131104,
+      "learning_rate": 1.7224312265058252e-05,
+      "loss": 0.2448,
+      "step": 695
+    },
+    {
+      "epoch": 2.4228769497400346,
+      "grad_norm": 0.9511550664901733,
+      "learning_rate": 1.7054334548239385e-05,
+      "loss": 0.2014,
+      "step": 700
+    },
+    {
+      "epoch": 2.440207972270364,
+      "grad_norm": 1.2386733293533325,
+      "learning_rate": 1.6884087555228387e-05,
+      "loss": 0.242,
+      "step": 705
+    },
+    {
+      "epoch": 2.4575389948006934,
+      "grad_norm": 1.0456082820892334,
+      "learning_rate": 1.6713593601505478e-05,
+      "loss": 0.2701,
+      "step": 710
+    },
+    {
+      "epoch": 2.4748700173310225,
+      "grad_norm": 1.1309462785720825,
+      "learning_rate": 1.6542875034921788e-05,
+      "loss": 0.2251,
+      "step": 715
+    },
+    {
+      "epoch": 2.4922010398613517,
+      "grad_norm": 1.0494320392608643,
+      "learning_rate": 1.6371954232770037e-05,
+      "loss": 0.1934,
+      "step": 720
+    },
+    {
+      "epoch": 2.5095320623916813,
+      "grad_norm": 1.5740057229995728,
+      "learning_rate": 1.62008535988514e-05,
+      "loss": 0.2481,
+      "step": 725
+    },
+    {
+      "epoch": 2.5268630849220104,
+      "grad_norm": 1.1560115814208984,
+      "learning_rate": 1.602959556053888e-05,
+      "loss": 0.2317,
+      "step": 730
+    },
+    {
+      "epoch": 2.5441941074523395,
+      "grad_norm": 1.0618700981140137,
+      "learning_rate": 1.5858202565837567e-05,
+      "loss": 0.2075,
+      "step": 735
+    },
+    {
+      "epoch": 2.561525129982669,
+      "grad_norm": 1.1966997385025024,
+      "learning_rate": 1.568669708044227e-05,
+      "loss": 0.2258,
+      "step": 740
+    },
+    {
+      "epoch": 2.5788561525129983,
+      "grad_norm": 1.2444030046463013,
+      "learning_rate": 1.5515101584792742e-05,
+      "loss": 0.2189,
+      "step": 745
+    },
+    {
+      "epoch": 2.5961871750433274,
+      "grad_norm": 1.0499557256698608,
+      "learning_rate": 1.5343438571127008e-05,
+      "loss": 0.2071,
+      "step": 750
+    },
+    {
+      "epoch": 2.613518197573657,
+      "grad_norm": 1.1848242282867432,
+      "learning_rate": 1.5171730540533165e-05,
+      "loss": 0.2309,
+      "step": 755
+    },
+    {
+      "epoch": 2.630849220103986,
+      "grad_norm": 1.2427995204925537,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2186,
+      "step": 760
+    },
+    {
+      "epoch": 2.6481802426343153,
+      "grad_norm": 1.1202366352081299,
+      "learning_rate": 1.4828269459466837e-05,
+      "loss": 0.2109,
+      "step": 765
+    },
+    {
+      "epoch": 2.665511265164645,
+      "grad_norm": 1.1815730333328247,
+      "learning_rate": 1.4656561428872996e-05,
+      "loss": 0.2234,
+      "step": 770
+    },
+    {
+      "epoch": 2.682842287694974,
+      "grad_norm": 1.2846264839172363,
+      "learning_rate": 1.4484898415207257e-05,
+      "loss": 0.1996,
+      "step": 775
+    },
+    {
+      "epoch": 2.700173310225303,
+      "grad_norm": 1.0805020332336426,
+      "learning_rate": 1.4313302919557727e-05,
+      "loss": 0.2134,
+      "step": 780
+    },
+    {
+      "epoch": 2.7175043327556327,
+      "grad_norm": 1.5587838888168335,
+      "learning_rate": 1.4141797434162437e-05,
+      "loss": 0.218,
+      "step": 785
+    },
+    {
+      "epoch": 2.734835355285962,
+      "grad_norm": 1.1907397508621216,
+      "learning_rate": 1.3970404439461129e-05,
+      "loss": 0.2024,
+      "step": 790
+    },
+    {
+      "epoch": 2.752166377816291,
+      "grad_norm": 1.0635807514190674,
+      "learning_rate": 1.3799146401148602e-05,
+      "loss": 0.1973,
+      "step": 795
+    },
+    {
+      "epoch": 2.76949740034662,
+      "grad_norm": 1.032384991645813,
+      "learning_rate": 1.3628045767229967e-05,
+      "loss": 0.1847,
+      "step": 800
+    },
+    {
+      "epoch": 2.7868284228769498,
+      "grad_norm": 1.5816798210144043,
+      "learning_rate": 1.3457124965078214e-05,
+      "loss": 0.2017,
+      "step": 805
+    },
+    {
+      "epoch": 2.804159445407279,
+      "grad_norm": 1.0850601196289062,
+      "learning_rate": 1.3286406398494524e-05,
+      "loss": 0.1986,
+      "step": 810
+    },
+    {
+      "epoch": 2.8214904679376085,
+      "grad_norm": 1.2348600625991821,
+      "learning_rate": 1.3115912444771617e-05,
+      "loss": 0.1711,
+      "step": 815
+    },
+    {
+      "epoch": 2.8388214904679376,
+      "grad_norm": 1.021596074104309,
+      "learning_rate": 1.2945665451760616e-05,
+      "loss": 0.1912,
+      "step": 820
+    },
+    {
+      "epoch": 2.856152512998267,
+      "grad_norm": 1.0768331289291382,
+      "learning_rate": 1.277568773494175e-05,
+      "loss": 0.1929,
+      "step": 825
+    },
+    {
+      "epoch": 2.873483535528596,
+      "grad_norm": 1.1603977680206299,
+      "learning_rate": 1.2606001574499316e-05,
+      "loss": 0.1789,
+      "step": 830
+    },
+    {
+      "epoch": 2.8908145580589255,
+      "grad_norm": 1.0426552295684814,
+      "learning_rate": 1.243662921240124e-05,
+      "loss": 0.179,
+      "step": 835
+    },
+    {
+      "epoch": 2.9081455805892547,
+      "grad_norm": 1.1170494556427002,
+      "learning_rate": 1.226759284948368e-05,
+      "loss": 0.1931,
+      "step": 840
+    },
+    {
+      "epoch": 2.9254766031195842,
+      "grad_norm": 1.1928642988204956,
+      "learning_rate": 1.2098914642541005e-05,
+      "loss": 0.2001,
+      "step": 845
+    },
+    {
+      "epoch": 2.9428076256499134,
+      "grad_norm": 1.1335448026657104,
+      "learning_rate": 1.1930616701421532e-05,
+      "loss": 0.1852,
+      "step": 850
+    },
+    {
+      "epoch": 2.9601386481802425,
+      "grad_norm": 1.200695514678955,
+      "learning_rate": 1.1762721086129447e-05,
+      "loss": 0.1795,
+      "step": 855
+    },
+    {
+      "epoch": 2.9774696707105717,
+      "grad_norm": 1.2488305568695068,
+      "learning_rate": 1.1595249803933232e-05,
+      "loss": 0.1796,
+      "step": 860
+    },
+    {
+      "epoch": 2.9948006932409013,
+      "grad_norm": 1.0621708631515503,
+      "learning_rate": 1.1428224806481025e-05,
+      "loss": 0.1793,
+      "step": 865
+    },
+    {
+      "epoch": 3.0103986135181975,
+      "grad_norm": 1.0290765762329102,
+      "learning_rate": 1.126166798692327e-05,
+      "loss": 0.1344,
+      "step": 870
+    },
+    {
+      "epoch": 3.027729636048527,
+      "grad_norm": 1.0009404420852661,
+      "learning_rate": 1.1095601177042995e-05,
+      "loss": 0.1426,
+      "step": 875
+    },
+    {
+      "epoch": 3.045060658578856,
+      "grad_norm": 1.0544887781143188,
+      "learning_rate": 1.09300461443942e-05,
+      "loss": 0.1417,
+      "step": 880
+    },
+    {
+      "epoch": 3.0623916811091854,
+      "grad_norm": 1.1883243322372437,
+      "learning_rate": 1.076502458944861e-05,
+      "loss": 0.1355,
+      "step": 885
+    },
+    {
+      "epoch": 3.079722703639515,
+      "grad_norm": 1.0606168508529663,
+      "learning_rate": 1.0600558142751245e-05,
+      "loss": 0.1391,
+      "step": 890
+    },
+    {
+      "epoch": 3.097053726169844,
+      "grad_norm": 1.047018051147461,
+      "learning_rate": 1.0436668362085157e-05,
+      "loss": 0.1248,
+      "step": 895
+    },
+    {
+      "epoch": 3.1143847487001732,
+      "grad_norm": 1.11692476272583,
+      "learning_rate": 1.027337672964569e-05,
+      "loss": 0.1177,
+      "step": 900
+    },
+    {
+      "epoch": 3.1317157712305024,
+      "grad_norm": 1.155957579612732,
+      "learning_rate": 1.0110704649224661e-05,
+      "loss": 0.1164,
+      "step": 905
+    },
+    {
+      "epoch": 3.149046793760832,
+      "grad_norm": 1.0759484767913818,
+      "learning_rate": 9.948673443404809e-06,
+      "loss": 0.1292,
+      "step": 910
+    },
+    {
+      "epoch": 3.166377816291161,
+      "grad_norm": 1.1999250650405884,
+      "learning_rate": 9.7873043507649e-06,
+      "loss": 0.1327,
+      "step": 915
+    },
+    {
+      "epoch": 3.1837088388214907,
+      "grad_norm": 1.1638423204421997,
+      "learning_rate": 9.626618523095825e-06,
+      "loss": 0.1432,
+      "step": 920
+    },
+    {
+      "epoch": 3.20103986135182,
+      "grad_norm": 0.9528873562812805,
+      "learning_rate": 9.46663702262811e-06,
+      "loss": 0.1212,
+      "step": 925
+    },
+    {
+      "epoch": 3.218370883882149,
+      "grad_norm": 1.1454589366912842,
+      "learning_rate": 9.307380819271092e-06,
+      "loss": 0.1292,
+      "step": 930
+    },
+    {
+      "epoch": 3.235701906412478,
+      "grad_norm": 1.0722419023513794,
+      "learning_rate": 9.148870787864297e-06,
+      "loss": 0.1336,
+      "step": 935
+    },
+    {
+      "epoch": 3.2530329289428077,
+      "grad_norm": 1.2213457822799683,
+      "learning_rate": 8.991127705441202e-06,
+      "loss": 0.1312,
+      "step": 940
+    },
+    {
+      "epoch": 3.270363951473137,
+      "grad_norm": 1.0836328268051147,
+      "learning_rate": 8.834172248505834e-06,
+      "loss": 0.1246,
+      "step": 945
+    },
+    {
+      "epoch": 3.2876949740034664,
+      "grad_norm": 1.0646871328353882,
+      "learning_rate": 8.678024990322571e-06,
+      "loss": 0.1327,
+      "step": 950
+    },
+    {
+      "epoch": 3.3050259965337956,
+      "grad_norm": 1.1547240018844604,
+      "learning_rate": 8.522706398219447e-06,
+      "loss": 0.1408,
+      "step": 955
+    },
+    {
+      "epoch": 3.3223570190641247,
+      "grad_norm": 1.2322415113449097,
+      "learning_rate": 8.368236830905371e-06,
+      "loss": 0.1312,
+      "step": 960
+    },
+    {
+      "epoch": 3.339688041594454,
+      "grad_norm": 0.8942275643348694,
+      "learning_rate": 8.214636535801532e-06,
+      "loss": 0.1102,
+      "step": 965
+    },
+    {
+      "epoch": 3.3570190641247835,
+      "grad_norm": 1.0620125532150269,
+      "learning_rate": 8.061925646387474e-06,
+      "loss": 0.1368,
+      "step": 970
+    },
+    {
+      "epoch": 3.3743500866551126,
+      "grad_norm": 1.198222041130066,
+      "learning_rate": 7.910124179562004e-06,
+      "loss": 0.1336,
+      "step": 975
+    },
+    {
+      "epoch": 3.391681109185442,
+      "grad_norm": 0.9789800047874451,
+      "learning_rate": 7.759252033019497e-06,
+      "loss": 0.1156,
+      "step": 980
+    },
+    {
+      "epoch": 3.4090121317157713,
+      "grad_norm": 0.8720641136169434,
+      "learning_rate": 7.609328982641693e-06,
+      "loss": 0.127,
+      "step": 985
+    },
+    {
+      "epoch": 3.4263431542461005,
+      "grad_norm": 1.1355503797531128,
+      "learning_rate": 7.460374679905608e-06,
+      "loss": 0.1388,
+      "step": 990
+    },
+    {
+      "epoch": 3.4436741767764296,
+      "grad_norm": 1.0707637071609497,
+      "learning_rate": 7.312408649307602e-06,
+      "loss": 0.114,
+      "step": 995
+    },
+    {
+      "epoch": 3.461005199306759,
+      "grad_norm": 0.910365641117096,
+      "learning_rate": 7.165450285804223e-06,
+      "loss": 0.1135,
+      "step": 1000
+    },
+    {
+      "epoch": 3.4783362218370883,
+      "grad_norm": 0.9218733310699463,
+      "learning_rate": 7.019518852269953e-06,
+      "loss": 0.1126,
+      "step": 1005
+    },
+    {
+      "epoch": 3.4956672443674175,
+      "grad_norm": 0.9174326658248901,
+      "learning_rate": 6.8746334769722576e-06,
+      "loss": 0.1193,
+      "step": 1010
+    },
+    {
+      "epoch": 3.512998266897747,
+      "grad_norm": 1.1110209226608276,
+      "learning_rate": 6.730813151064363e-06,
+      "loss": 0.1109,
+      "step": 1015
+    },
+    {
+      "epoch": 3.5303292894280762,
+      "grad_norm": 1.018364429473877,
+      "learning_rate": 6.588076726095902e-06,
+      "loss": 0.1406,
+      "step": 1020
+    },
+    {
+      "epoch": 3.5476603119584054,
+      "grad_norm": 0.9070491194725037,
+      "learning_rate": 6.44644291154193e-06,
+      "loss": 0.1283,
+      "step": 1025
+    },
+    {
+      "epoch": 3.564991334488735,
+      "grad_norm": 0.9356378316879272,
+      "learning_rate": 6.305930272350549e-06,
+      "loss": 0.1273,
+      "step": 1030
+    },
+    {
+      "epoch": 3.582322357019064,
+      "grad_norm": 0.9019298553466797,
+      "learning_rate": 6.1665572265094565e-06,
+      "loss": 0.1157,
+      "step": 1035
+    },
+    {
+      "epoch": 3.5996533795493937,
+      "grad_norm": 0.8685014247894287,
+      "learning_rate": 6.0283420426317606e-06,
+      "loss": 0.1059,
+      "step": 1040
+    },
+    {
+      "epoch": 3.616984402079723,
+      "grad_norm": 0.9012817740440369,
+      "learning_rate": 5.891302837561407e-06,
+      "loss": 0.1168,
+      "step": 1045
+    },
+    {
+      "epoch": 3.634315424610052,
+      "grad_norm": 0.9666245579719543,
+      "learning_rate": 5.755457573998451e-06,
+      "loss": 0.1137,
+      "step": 1050
+    },
+    {
+      "epoch": 3.651646447140381,
+      "grad_norm": 0.886913001537323,
+      "learning_rate": 5.620824058144576e-06,
+      "loss": 0.1121,
+      "step": 1055
+    },
+    {
+      "epoch": 3.6689774696707107,
+      "grad_norm": 0.8900812864303589,
+      "learning_rate": 5.487419937369112e-06,
+      "loss": 0.1131,
+      "step": 1060
+    },
+    {
+      "epoch": 3.68630849220104,
+      "grad_norm": 0.9007474780082703,
+      "learning_rate": 5.35526269789585e-06,
+      "loss": 0.1168,
+      "step": 1065
+    },
+    {
+      "epoch": 3.703639514731369,
+      "grad_norm": 0.9596942663192749,
+      "learning_rate": 5.224369662511003e-06,
+      "loss": 0.0933,
+      "step": 1070
+    },
+    {
+      "epoch": 3.7209705372616986,
+      "grad_norm": 1.0141007900238037,
+      "learning_rate": 5.094757988292613e-06,
+      "loss": 0.1073,
+      "step": 1075
+    },
+    {
+      "epoch": 3.7383015597920277,
+      "grad_norm": 0.990217387676239,
+      "learning_rate": 4.966444664361611e-06,
+      "loss": 0.1167,
+      "step": 1080
+    },
+    {
+      "epoch": 3.755632582322357,
+      "grad_norm": 0.8267380595207214,
+      "learning_rate": 4.839446509654968e-06,
+      "loss": 0.1099,
+      "step": 1085
+    },
+    {
+      "epoch": 3.7729636048526864,
+      "grad_norm": 0.762187659740448,
+      "learning_rate": 4.713780170721089e-06,
+      "loss": 0.1005,
+      "step": 1090
+    },
+    {
+      "epoch": 3.7902946273830156,
+      "grad_norm": 0.9725874662399292,
+      "learning_rate": 4.589462119537849e-06,
+      "loss": 0.123,
+      "step": 1095
+    },
+    {
+      "epoch": 3.8076256499133447,
+      "grad_norm": 0.9700238704681396,
+      "learning_rate": 4.466508651353498e-06,
+      "loss": 0.0971,
+      "step": 1100
+    },
+    {
+      "epoch": 3.8249566724436743,
+      "grad_norm": 1.0050092935562134,
+      "learning_rate": 4.344935882550701e-06,
+      "loss": 0.1115,
+      "step": 1105
+    },
+    {
+      "epoch": 3.8422876949740035,
+      "grad_norm": 0.904208242893219,
+      "learning_rate": 4.224759748534088e-06,
+      "loss": 0.1021,
+      "step": 1110
+    },
+    {
+      "epoch": 3.8596187175043326,
+      "grad_norm": 0.9154079556465149,
+      "learning_rate": 4.105996001641452e-06,
+      "loss": 0.1033,
+      "step": 1115
+    },
+    {
+      "epoch": 3.876949740034662,
+      "grad_norm": 0.9821785688400269,
+      "learning_rate": 3.988660209078994e-06,
+      "loss": 0.1152,
+      "step": 1120
+    },
+    {
+      "epoch": 3.8942807625649913,
+      "grad_norm": 0.8482122421264648,
+      "learning_rate": 3.872767750880825e-06,
+      "loss": 0.1009,
+      "step": 1125
+    },
+    {
+      "epoch": 3.9116117850953205,
+      "grad_norm": 0.8988242149353027,
+      "learning_rate": 3.7583338178929953e-06,
+      "loss": 0.1047,
+      "step": 1130
+    },
+    {
+      "epoch": 3.92894280762565,
+      "grad_norm": 1.079358458518982,
+      "learning_rate": 3.6453734097822966e-06,
+      "loss": 0.0981,
+      "step": 1135
+    },
+    {
+      "epoch": 3.946273830155979,
+      "grad_norm": 0.9375408291816711,
+      "learning_rate": 3.533901333070191e-06,
+      "loss": 0.1075,
+      "step": 1140
+    },
+    {
+      "epoch": 3.9636048526863084,
+      "grad_norm": 0.8924766778945923,
+      "learning_rate": 3.423932199191973e-06,
+      "loss": 0.107,
+      "step": 1145
+    },
+    {
+      "epoch": 3.980935875216638,
+      "grad_norm": 0.7494178414344788,
+      "learning_rate": 3.315480422581583e-06,
+      "loss": 0.0989,
+      "step": 1150
+    },
+    {
+      "epoch": 3.998266897746967,
+      "grad_norm": 0.9619389772415161,
+      "learning_rate": 3.2085602187821854e-06,
+      "loss": 0.108,
+      "step": 1155
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1445,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6397934526128456e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

1_128_e5_3e-5/checkpoint-1156/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b529bde561655fa6f1efd722a181ca5b0d63578ce8f7cd71cb25200956f56df
+size 7736

1_128_e5_3e-5/checkpoint-1156/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/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, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in 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,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, 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_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``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    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("--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_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

1_128_e5_3e-5/checkpoint-1445/README.md

@@ -0,0 +1,202 @@
+---
+base_model: ibm-granite/granite-3.3-8b-base
+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 0.15.2

1_128_e5_3e-5/checkpoint-1445/adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
+  "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": 256,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "k_proj",
+    "v_proj",
+    "o_proj",
+    "up_proj",
+    "q_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

1_128_e5_3e-5/checkpoint-1445/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:581a432e0ead1e80e94e3b6516239c56704c8b1586345eeeb3e471ee3a31de27
+size 791751704

1_128_e5_3e-5/checkpoint-1445/latest

@@ -0,0 +1 @@
+global_step1445

1_128_e5_3e-5/checkpoint-1445/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7e35fc86260776cca3c0529df988a5a1ea5c62130f5467351ee0b91a1792cc4
+size 15920

1_128_e5_3e-5/checkpoint-1445/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0626c8f97ca644989caf577a6c98e13e1c2282ceafa4c2b874c35cd0dcf15265
+size 15920

1_128_e5_3e-5/checkpoint-1445/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:379224bf50d3c42dfc99049254a050669bb2523aeaa1fe7cbb443174041c6a72
+size 15920

1_128_e5_3e-5/checkpoint-1445/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2af1805ec61f24fe369cb57d9344423d558d25f6752b830627b6989263ca4384
+size 15920

1_128_e5_3e-5/checkpoint-1445/rng_state_4.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e87a5587610b5844fb6a63082f0dcb651e91f8197d7ab7bb493edd6618dd0fb
+size 15920

1_128_e5_3e-5/checkpoint-1445/rng_state_5.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:703d9549bf81deb55701f0d7e392e323ea5c1056c471acff857353fed5503762
+size 15920

1_128_e5_3e-5/checkpoint-1445/rng_state_6.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3cc310e14ca8ed3bd40def00cdf73c09ba7221636febcde942b123e3c12a37e
+size 15920

1_128_e5_3e-5/checkpoint-1445/rng_state_7.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a4bc4141de8bcaa0f28d03f74e4fd7da7996a7db56a153e310cf8a71b0cb7d8
+size 15920

1_128_e5_3e-5/checkpoint-1445/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:181155ce868c0e0e7fbedfefe7b7a25fa1768b171b6db7e69387945472bcbae6
+size 1064

1_128_e5_3e-5/checkpoint-1445/special_tokens_map.json

@@ -0,0 +1,45 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<reponame>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

1_128_e5_3e-5/checkpoint-1445/tokenizer_config.json

@@ -0,0 +1,188 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<fim_prefix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<fim_middle>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<fim_suffix>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<fim_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<filename>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "<gh_stars>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "7": {
+      "content": "<issue_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "8": {
+      "content": "<issue_comment>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "9": {
+      "content": "<issue_closed>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "10": {
+      "content": "<jupyter_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "11": {
+      "content": "<jupyter_text>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "12": {
+      "content": "<jupyter_code>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "13": {
+      "content": "<jupyter_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "14": {
+      "content": "<empty_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "15": {
+      "content": "<commit_before>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "16": {
+      "content": "<commit_msg>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "17": {
+      "content": "<commit_after>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "18": {
+      "content": "<reponame>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<empty_output>",
+    "<commit_before>",
+    "<commit_msg>",
+    "<commit_after>",
+    "<reponame>"
+  ],
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "pad_token": "<reponame>",
+  "padding_side": "left",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

1_128_e5_3e-5/checkpoint-1445/trainer_state.json

@@ -0,0 +1,2057 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1445,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01733102253032929,
+      "grad_norm": 1.1325727701187134,
+      "learning_rate": 1.6438356164383561e-06,
+      "loss": 1.3757,
+      "step": 5
+    },
+    {
+      "epoch": 0.03466204506065858,
+      "grad_norm": 1.044039249420166,
+      "learning_rate": 3.6986301369863014e-06,
+      "loss": 1.3112,
+      "step": 10
+    },
+    {
+      "epoch": 0.05199306759098787,
+      "grad_norm": 0.621895968914032,
+      "learning_rate": 5.753424657534246e-06,
+      "loss": 1.3603,
+      "step": 15
+    },
+    {
+      "epoch": 0.06932409012131716,
+      "grad_norm": 0.6180606484413147,
+      "learning_rate": 7.808219178082192e-06,
+      "loss": 1.2902,
+      "step": 20
+    },
+    {
+      "epoch": 0.08665511265164645,
+      "grad_norm": 0.5306586027145386,
+      "learning_rate": 9.863013698630136e-06,
+      "loss": 1.2892,
+      "step": 25
+    },
+    {
+      "epoch": 0.10398613518197573,
+      "grad_norm": 0.661157488822937,
+      "learning_rate": 1.1917808219178083e-05,
+      "loss": 1.2568,
+      "step": 30
+    },
+    {
+      "epoch": 0.12131715771230503,
+      "grad_norm": 0.666887104511261,
+      "learning_rate": 1.3972602739726027e-05,
+      "loss": 1.2041,
+      "step": 35
+    },
+    {
+      "epoch": 0.1386481802426343,
+      "grad_norm": 0.48998892307281494,
+      "learning_rate": 1.6027397260273974e-05,
+      "loss": 1.235,
+      "step": 40
+    },
+    {
+      "epoch": 0.1559792027729636,
+      "grad_norm": 0.4912819266319275,
+      "learning_rate": 1.8082191780821916e-05,
+      "loss": 1.2399,
+      "step": 45
+    },
+    {
+      "epoch": 0.1733102253032929,
+      "grad_norm": 0.48102739453315735,
+      "learning_rate": 2.0136986301369863e-05,
+      "loss": 1.2004,
+      "step": 50
+    },
+    {
+      "epoch": 0.19064124783362218,
+      "grad_norm": 0.5094467997550964,
+      "learning_rate": 2.219178082191781e-05,
+      "loss": 1.1671,
+      "step": 55
+    },
+    {
+      "epoch": 0.20797227036395147,
+      "grad_norm": 0.5597960352897644,
+      "learning_rate": 2.4246575342465755e-05,
+      "loss": 1.2149,
+      "step": 60
+    },
+    {
+      "epoch": 0.22530329289428075,
+      "grad_norm": 0.48657160997390747,
+      "learning_rate": 2.6301369863013698e-05,
+      "loss": 1.1888,
+      "step": 65
+    },
+    {
+      "epoch": 0.24263431542461006,
+      "grad_norm": 0.6103832721710205,
+      "learning_rate": 2.8356164383561644e-05,
+      "loss": 1.1243,
+      "step": 70
+    },
+    {
+      "epoch": 0.25996533795493937,
+      "grad_norm": 0.4703650176525116,
+      "learning_rate": 2.9999960676460984e-05,
+      "loss": 1.1172,
+      "step": 75
+    },
+    {
+      "epoch": 0.2772963604852686,
+      "grad_norm": 0.5571750998497009,
+      "learning_rate": 2.9998584374244097e-05,
+      "loss": 1.1198,
+      "step": 80
+    },
+    {
+      "epoch": 0.29462738301559793,
+      "grad_norm": 0.5368012189865112,
+      "learning_rate": 2.999524210125035e-05,
+      "loss": 1.1335,
+      "step": 85
+    },
+    {
+      "epoch": 0.3119584055459272,
+      "grad_norm": 0.6115505695343018,
+      "learning_rate": 2.9989934295575147e-05,
+      "loss": 1.1163,
+      "step": 90
+    },
+    {
+      "epoch": 0.3292894280762565,
+      "grad_norm": 0.5922970175743103,
+      "learning_rate": 2.998266165295021e-05,
+      "loss": 1.1322,
+      "step": 95
+    },
+    {
+      "epoch": 0.3466204506065858,
+      "grad_norm": 0.6367848515510559,
+      "learning_rate": 2.9973425126652373e-05,
+      "loss": 1.0499,
+      "step": 100
+    },
+    {
+      "epoch": 0.36395147313691506,
+      "grad_norm": 0.5297245383262634,
+      "learning_rate": 2.9962225927378597e-05,
+      "loss": 1.0052,
+      "step": 105
+    },
+    {
+      "epoch": 0.38128249566724437,
+      "grad_norm": 0.6090821623802185,
+      "learning_rate": 2.9949065523087333e-05,
+      "loss": 1.102,
+      "step": 110
+    },
+    {
+      "epoch": 0.3986135181975737,
+      "grad_norm": 0.698509156703949,
+      "learning_rate": 2.9933945638806056e-05,
+      "loss": 1.0093,
+      "step": 115
+    },
+    {
+      "epoch": 0.41594454072790293,
+      "grad_norm": 0.706832766532898,
+      "learning_rate": 2.9916868256405185e-05,
+      "loss": 1.0042,
+      "step": 120
+    },
+    {
+      "epoch": 0.43327556325823224,
+      "grad_norm": 0.6957172155380249,
+      "learning_rate": 2.9897835614338295e-05,
+      "loss": 0.9757,
+      "step": 125
+    },
+    {
+      "epoch": 0.4506065857885615,
+      "grad_norm": 0.6912302374839783,
+      "learning_rate": 2.987685020734869e-05,
+      "loss": 0.9567,
+      "step": 130
+    },
+    {
+      "epoch": 0.4679376083188908,
+      "grad_norm": 0.7200044393539429,
+      "learning_rate": 2.985391478614244e-05,
+      "loss": 0.9602,
+      "step": 135
+    },
+    {
+      "epoch": 0.4852686308492201,
+      "grad_norm": 0.6456925272941589,
+      "learning_rate": 2.982903235702778e-05,
+      "loss": 0.9608,
+      "step": 140
+    },
+    {
+      "epoch": 0.5025996533795494,
+      "grad_norm": 0.7961540222167969,
+      "learning_rate": 2.9802206181521086e-05,
+      "loss": 0.8843,
+      "step": 145
+    },
+    {
+      "epoch": 0.5199306759098787,
+      "grad_norm": 0.6846944093704224,
+      "learning_rate": 2.9773439775919343e-05,
+      "loss": 0.9035,
+      "step": 150
+    },
+    {
+      "epoch": 0.537261698440208,
+      "grad_norm": 0.8010955452919006,
+      "learning_rate": 2.974273691083926e-05,
+      "loss": 0.9036,
+      "step": 155
+    },
+    {
+      "epoch": 0.5545927209705372,
+      "grad_norm": 0.8240249752998352,
+      "learning_rate": 2.971010161072301e-05,
+      "loss": 0.9011,
+      "step": 160
+    },
+    {
+      "epoch": 0.5719237435008665,
+      "grad_norm": 0.8053805828094482,
+      "learning_rate": 2.9675538153310732e-05,
+      "loss": 0.8949,
+      "step": 165
+    },
+    {
+      "epoch": 0.5892547660311959,
+      "grad_norm": 0.8074398040771484,
+      "learning_rate": 2.9639051069079794e-05,
+      "loss": 0.8769,
+      "step": 170
+    },
+    {
+      "epoch": 0.6065857885615251,
+      "grad_norm": 0.8516232371330261,
+      "learning_rate": 2.9600645140650985e-05,
+      "loss": 0.8376,
+      "step": 175
+    },
+    {
+      "epoch": 0.6239168110918544,
+      "grad_norm": 0.7514265179634094,
+      "learning_rate": 2.9560325402161598e-05,
+      "loss": 0.8566,
+      "step": 180
+    },
+    {
+      "epoch": 0.6412478336221837,
+      "grad_norm": 0.8322745561599731,
+      "learning_rate": 2.9518097138605574e-05,
+      "loss": 0.819,
+      "step": 185
+    },
+    {
+      "epoch": 0.658578856152513,
+      "grad_norm": 0.8268292546272278,
+      "learning_rate": 2.9473965885140774e-05,
+      "loss": 0.8155,
+      "step": 190
+    },
+    {
+      "epoch": 0.6759098786828422,
+      "grad_norm": 0.769992470741272,
+      "learning_rate": 2.9427937426363424e-05,
+      "loss": 0.7585,
+      "step": 195
+    },
+    {
+      "epoch": 0.6932409012131716,
+      "grad_norm": 0.9443396329879761,
+      "learning_rate": 2.938001779554991e-05,
+      "loss": 0.8124,
+      "step": 200
+    },
+    {
+      "epoch": 0.7105719237435009,
+      "grad_norm": 0.8200739026069641,
+      "learning_rate": 2.9330213273865936e-05,
+      "loss": 0.8034,
+      "step": 205
+    },
+    {
+      "epoch": 0.7279029462738301,
+      "grad_norm": 0.90153568983078,
+      "learning_rate": 2.927853038954322e-05,
+      "loss": 0.7209,
+      "step": 210
+    },
+    {
+      "epoch": 0.7452339688041595,
+      "grad_norm": 0.9012041687965393,
+      "learning_rate": 2.9224975917023778e-05,
+      "loss": 0.8571,
+      "step": 215
+    },
+    {
+      "epoch": 0.7625649913344887,
+      "grad_norm": 0.8709543943405151,
+      "learning_rate": 2.9169556876071967e-05,
+      "loss": 0.8261,
+      "step": 220
+    },
+    {
+      "epoch": 0.779896013864818,
+      "grad_norm": 0.9728206396102905,
+      "learning_rate": 2.911228053085434e-05,
+      "loss": 0.7364,
+      "step": 225
+    },
+    {
+      "epoch": 0.7972270363951474,
+      "grad_norm": 1.0054287910461426,
+      "learning_rate": 2.9053154388987493e-05,
+      "loss": 0.7678,
+      "step": 230
+    },
+    {
+      "epoch": 0.8145580589254766,
+      "grad_norm": 0.8341831564903259,
+      "learning_rate": 2.8992186200553975e-05,
+      "loss": 0.7069,
+      "step": 235
+    },
+    {
+      "epoch": 0.8318890814558059,
+      "grad_norm": 0.9587536454200745,
+      "learning_rate": 2.892938395708644e-05,
+      "loss": 0.7277,
+      "step": 240
+    },
+    {
+      "epoch": 0.8492201039861352,
+      "grad_norm": 0.9497514963150024,
+      "learning_rate": 2.886475589052013e-05,
+      "loss": 0.7346,
+      "step": 245
+    },
+    {
+      "epoch": 0.8665511265164645,
+      "grad_norm": 0.9413470029830933,
+      "learning_rate": 2.8798310472113877e-05,
+      "loss": 0.7299,
+      "step": 250
+    },
+    {
+      "epoch": 0.8838821490467937,
+      "grad_norm": 1.027370810508728,
+      "learning_rate": 2.8730056411339695e-05,
+      "loss": 0.6685,
+      "step": 255
+    },
+    {
+      "epoch": 0.901213171577123,
+      "grad_norm": 0.9176335334777832,
+      "learning_rate": 2.866000265474117e-05,
+      "loss": 0.7323,
+      "step": 260
+    },
+    {
+      "epoch": 0.9185441941074524,
+      "grad_norm": 0.9519228339195251,
+      "learning_rate": 2.858815838476078e-05,
+      "loss": 0.684,
+      "step": 265
+    },
+    {
+      "epoch": 0.9358752166377816,
+      "grad_norm": 1.0398930311203003,
+      "learning_rate": 2.8514533018536286e-05,
+      "loss": 0.7218,
+      "step": 270
+    },
+    {
+      "epoch": 0.9532062391681109,
+      "grad_norm": 1.0150400400161743,
+      "learning_rate": 2.8439136206666365e-05,
+      "loss": 0.6864,
+      "step": 275
+    },
+    {
+      "epoch": 0.9705372616984402,
+      "grad_norm": 1.4193578958511353,
+      "learning_rate": 2.8361977831945614e-05,
+      "loss": 0.66,
+      "step": 280
+    },
+    {
+      "epoch": 0.9878682842287695,
+      "grad_norm": 0.9296590089797974,
+      "learning_rate": 2.8283068008069188e-05,
+      "loss": 0.6661,
+      "step": 285
+    },
+    {
+      "epoch": 1.0034662045060658,
+      "grad_norm": 1.0106604099273682,
+      "learning_rate": 2.820241707830707e-05,
+      "loss": 0.6015,
+      "step": 290
+    },
+    {
+      "epoch": 1.0207972270363952,
+      "grad_norm": 1.0099163055419922,
+      "learning_rate": 2.8120035614148358e-05,
+      "loss": 0.5811,
+      "step": 295
+    },
+    {
+      "epoch": 1.0381282495667243,
+      "grad_norm": 1.0148612260818481,
+      "learning_rate": 2.803593441391555e-05,
+      "loss": 0.5606,
+      "step": 300
+    },
+    {
+      "epoch": 1.0554592720970537,
+      "grad_norm": 1.1285183429718018,
+      "learning_rate": 2.795012450134913e-05,
+      "loss": 0.593,
+      "step": 305
+    },
+    {
+      "epoch": 1.072790294627383,
+      "grad_norm": 1.0806018114089966,
+      "learning_rate": 2.7862617124162643e-05,
+      "loss": 0.5492,
+      "step": 310
+    },
+    {
+      "epoch": 1.0901213171577122,
+      "grad_norm": 1.0284126996994019,
+      "learning_rate": 2.7773423752568347e-05,
+      "loss": 0.5228,
+      "step": 315
+    },
+    {
+      "epoch": 1.1074523396880416,
+      "grad_norm": 1.3091356754302979,
+      "learning_rate": 2.768255607777373e-05,
+      "loss": 0.5503,
+      "step": 320
+    },
+    {
+      "epoch": 1.124783362218371,
+      "grad_norm": 0.9635022282600403,
+      "learning_rate": 2.7590026010449076e-05,
+      "loss": 0.5203,
+      "step": 325
+    },
+    {
+      "epoch": 1.1421143847487,
+      "grad_norm": 0.998070478439331,
+      "learning_rate": 2.7495845679166252e-05,
+      "loss": 0.5471,
+      "step": 330
+    },
+    {
+      "epoch": 1.1594454072790294,
+      "grad_norm": 1.2872036695480347,
+      "learning_rate": 2.7400027428808897e-05,
+      "loss": 0.505,
+      "step": 335
+    },
+    {
+      "epoch": 1.1767764298093588,
+      "grad_norm": 1.0663659572601318,
+      "learning_rate": 2.730258381895434e-05,
+      "loss": 0.5486,
+      "step": 340
+    },
+    {
+      "epoch": 1.194107452339688,
+      "grad_norm": 1.1138153076171875,
+      "learning_rate": 2.720352762222728e-05,
+      "loss": 0.5617,
+      "step": 345
+    },
+    {
+      "epoch": 1.2114384748700173,
+      "grad_norm": 1.1818647384643555,
+      "learning_rate": 2.7102871822625627e-05,
+      "loss": 0.5401,
+      "step": 350
+    },
+    {
+      "epoch": 1.2287694974003467,
+      "grad_norm": 1.0879274606704712,
+      "learning_rate": 2.700062961381856e-05,
+      "loss": 0.571,
+      "step": 355
+    },
+    {
+      "epoch": 1.2461005199306758,
+      "grad_norm": 1.346692442893982,
+      "learning_rate": 2.6896814397417174e-05,
+      "loss": 0.5403,
+      "step": 360
+    },
+    {
+      "epoch": 1.2634315424610052,
+      "grad_norm": 1.0724241733551025,
+      "learning_rate": 2.6791439781217815e-05,
+      "loss": 0.4927,
+      "step": 365
+    },
+    {
+      "epoch": 1.2807625649913346,
+      "grad_norm": 1.144758701324463,
+      "learning_rate": 2.6684519577418417e-05,
+      "loss": 0.5394,
+      "step": 370
+    },
+    {
+      "epoch": 1.2980935875216637,
+      "grad_norm": 1.071765661239624,
+      "learning_rate": 2.6576067800808026e-05,
+      "loss": 0.4743,
+      "step": 375
+    },
+    {
+      "epoch": 1.315424610051993,
+      "grad_norm": 1.0622762441635132,
+      "learning_rate": 2.646609866692981e-05,
+      "loss": 0.4881,
+      "step": 380
+    },
+    {
+      "epoch": 1.3327556325823224,
+      "grad_norm": 1.4975517988204956,
+      "learning_rate": 2.6354626590217705e-05,
+      "loss": 0.4954,
+      "step": 385
+    },
+    {
+      "epoch": 1.3500866551126516,
+      "grad_norm": 1.0805246829986572,
+      "learning_rate": 2.624166618210701e-05,
+      "loss": 0.458,
+      "step": 390
+    },
+    {
+      "epoch": 1.367417677642981,
+      "grad_norm": 1.0548033714294434,
+      "learning_rate": 2.6127232249119177e-05,
+      "loss": 0.4765,
+      "step": 395
+    },
+    {
+      "epoch": 1.38474870017331,
+      "grad_norm": 1.2031217813491821,
+      "learning_rate": 2.6011339790921012e-05,
+      "loss": 0.5447,
+      "step": 400
+    },
+    {
+      "epoch": 1.4020797227036395,
+      "grad_norm": 1.2651617527008057,
+      "learning_rate": 2.5894003998358553e-05,
+      "loss": 0.4935,
+      "step": 405
+    },
+    {
+      "epoch": 1.4194107452339688,
+      "grad_norm": 1.282373070716858,
+      "learning_rate": 2.5775240251465914e-05,
+      "loss": 0.4586,
+      "step": 410
+    },
+    {
+      "epoch": 1.436741767764298,
+      "grad_norm": 1.0972130298614502,
+      "learning_rate": 2.56550641174493e-05,
+      "loss": 0.4681,
+      "step": 415
+    },
+    {
+      "epoch": 1.4540727902946273,
+      "grad_norm": 1.1157981157302856,
+      "learning_rate": 2.5533491348646504e-05,
+      "loss": 0.4499,
+      "step": 420
+    },
+    {
+      "epoch": 1.4714038128249567,
+      "grad_norm": 1.1279178857803345,
+      "learning_rate": 2.541053788046215e-05,
+      "loss": 0.4645,
+      "step": 425
+    },
+    {
+      "epoch": 1.4887348353552858,
+      "grad_norm": 1.1415854692459106,
+      "learning_rate": 2.5286219829278914e-05,
+      "loss": 0.4214,
+      "step": 430
+    },
+    {
+      "epoch": 1.5060658578856152,
+      "grad_norm": 1.1717807054519653,
+      "learning_rate": 2.5160553490345038e-05,
+      "loss": 0.4397,
+      "step": 435
+    },
+    {
+      "epoch": 1.5233968804159446,
+      "grad_norm": 1.2159512042999268,
+      "learning_rate": 2.5033555335638388e-05,
+      "loss": 0.4619,
+      "step": 440
+    },
+    {
+      "epoch": 1.5407279029462737,
+      "grad_norm": 1.2196030616760254,
+      "learning_rate": 2.490524201170739e-05,
+      "loss": 0.4245,
+      "step": 445
+    },
+    {
+      "epoch": 1.558058925476603,
+      "grad_norm": 1.108749508857727,
+      "learning_rate": 2.4775630337489e-05,
+      "loss": 0.4329,
+      "step": 450
+    },
+    {
+      "epoch": 1.5753899480069324,
+      "grad_norm": 1.1284644603729248,
+      "learning_rate": 2.4644737302104156e-05,
+      "loss": 0.4155,
+      "step": 455
+    },
+    {
+      "epoch": 1.5927209705372616,
+      "grad_norm": 1.0621742010116577,
+      "learning_rate": 2.451258006263089e-05,
+      "loss": 0.4377,
+      "step": 460
+    },
+    {
+      "epoch": 1.610051993067591,
+      "grad_norm": 1.178806185722351,
+      "learning_rate": 2.4379175941855423e-05,
+      "loss": 0.4428,
+      "step": 465
+    },
+    {
+      "epoch": 1.6273830155979203,
+      "grad_norm": 1.216556191444397,
+      "learning_rate": 2.424454242600155e-05,
+      "loss": 0.4196,
+      "step": 470
+    },
+    {
+      "epoch": 1.6447140381282495,
+      "grad_norm": 1.2306348085403442,
+      "learning_rate": 2.4108697162438595e-05,
+      "loss": 0.4011,
+      "step": 475
+    },
+    {
+      "epoch": 1.6620450606585788,
+      "grad_norm": 1.5304279327392578,
+      "learning_rate": 2.397165795736824e-05,
+      "loss": 0.4211,
+      "step": 480
+    },
+    {
+      "epoch": 1.6793760831889082,
+      "grad_norm": 1.033577561378479,
+      "learning_rate": 2.3833442773490544e-05,
+      "loss": 0.3988,
+      "step": 485
+    },
+    {
+      "epoch": 1.6967071057192373,
+      "grad_norm": 1.2301174402236938,
+      "learning_rate": 2.369406972764945e-05,
+      "loss": 0.381,
+      "step": 490
+    },
+    {
+      "epoch": 1.7140381282495667,
+      "grad_norm": 1.1185941696166992,
+      "learning_rate": 2.3553557088458077e-05,
+      "loss": 0.3582,
+      "step": 495
+    },
+    {
+      "epoch": 1.731369150779896,
+      "grad_norm": 1.2030754089355469,
+      "learning_rate": 2.3411923273904104e-05,
+      "loss": 0.3738,
+      "step": 500
+    },
+    {
+      "epoch": 1.7487001733102252,
+      "grad_norm": 1.1300477981567383,
+      "learning_rate": 2.326918684893564e-05,
+      "loss": 0.3944,
+      "step": 505
+    },
+    {
+      "epoch": 1.7660311958405546,
+      "grad_norm": 1.1240997314453125,
+      "learning_rate": 2.312536652302774e-05,
+      "loss": 0.3737,
+      "step": 510
+    },
+    {
+      "epoch": 1.783362218370884,
+      "grad_norm": 1.1487401723861694,
+      "learning_rate": 2.298048114773005e-05,
+      "loss": 0.4144,
+      "step": 515
+    },
+    {
+      "epoch": 1.800693240901213,
+      "grad_norm": 1.1210594177246094,
+      "learning_rate": 2.2834549714195772e-05,
+      "loss": 0.3697,
+      "step": 520
+    },
+    {
+      "epoch": 1.8180242634315424,
+      "grad_norm": 1.1202471256256104,
+      "learning_rate": 2.26875913506924e-05,
+      "loss": 0.4059,
+      "step": 525
+    },
+    {
+      "epoch": 1.8353552859618718,
+      "grad_norm": 1.058485507965088,
+      "learning_rate": 2.2539625320094396e-05,
+      "loss": 0.4164,
+      "step": 530
+    },
+    {
+      "epoch": 1.852686308492201,
+      "grad_norm": 1.0952345132827759,
+      "learning_rate": 2.2390671017358307e-05,
+      "loss": 0.3537,
+      "step": 535
+    },
+    {
+      "epoch": 1.8700173310225303,
+      "grad_norm": 1.035676121711731,
+      "learning_rate": 2.2240747966980508e-05,
+      "loss": 0.3455,
+      "step": 540
+    },
+    {
+      "epoch": 1.8873483535528597,
+      "grad_norm": 1.5411173105239868,
+      "learning_rate": 2.2089875820437995e-05,
+      "loss": 0.3657,
+      "step": 545
+    },
+    {
+      "epoch": 1.9046793760831888,
+      "grad_norm": 1.1188682317733765,
+      "learning_rate": 2.1938074353612533e-05,
+      "loss": 0.3422,
+      "step": 550
+    },
+    {
+      "epoch": 1.9220103986135182,
+      "grad_norm": 1.0884416103363037,
+      "learning_rate": 2.178536346419847e-05,
+      "loss": 0.3412,
+      "step": 555
+    },
+    {
+      "epoch": 1.9393414211438476,
+      "grad_norm": 1.2015438079833984,
+      "learning_rate": 2.1631763169094628e-05,
+      "loss": 0.3559,
+      "step": 560
+    },
+    {
+      "epoch": 1.9566724436741767,
+      "grad_norm": 1.431586503982544,
+      "learning_rate": 2.1477293601780554e-05,
+      "loss": 0.3611,
+      "step": 565
+    },
+    {
+      "epoch": 1.974003466204506,
+      "grad_norm": 1.111899971961975,
+      "learning_rate": 2.132197500967743e-05,
+      "loss": 0.3566,
+      "step": 570
+    },
+    {
+      "epoch": 1.9913344887348354,
+      "grad_norm": 1.2778851985931396,
+      "learning_rate": 2.116582775149417e-05,
+      "loss": 0.3464,
+      "step": 575
+    },
+    {
+      "epoch": 2.0069324090121317,
+      "grad_norm": 1.0723752975463867,
+      "learning_rate": 2.1008872294558802e-05,
+      "loss": 0.2894,
+      "step": 580
+    },
+    {
+      "epoch": 2.024263431542461,
+      "grad_norm": 1.194329857826233,
+      "learning_rate": 2.0851129212135702e-05,
+      "loss": 0.2778,
+      "step": 585
+    },
+    {
+      "epoch": 2.0415944540727904,
+      "grad_norm": 1.1894909143447876,
+      "learning_rate": 2.0692619180728914e-05,
+      "loss": 0.2422,
+      "step": 590
+    },
+    {
+      "epoch": 2.0589254766031195,
+      "grad_norm": 1.8235598802566528,
+      "learning_rate": 2.0533362977371892e-05,
+      "loss": 0.2683,
+      "step": 595
+    },
+    {
+      "epoch": 2.0762564991334487,
+      "grad_norm": 1.3136845827102661,
+      "learning_rate": 2.0373381476904173e-05,
+      "loss": 0.2878,
+      "step": 600
+    },
+    {
+      "epoch": 2.0935875216637783,
+      "grad_norm": 1.1483287811279297,
+      "learning_rate": 2.02126956492351e-05,
+      "loss": 0.2553,
+      "step": 605
+    },
+    {
+      "epoch": 2.1109185441941074,
+      "grad_norm": 1.1638771295547485,
+      "learning_rate": 2.0051326556595192e-05,
+      "loss": 0.276,
+      "step": 610
+    },
+    {
+      "epoch": 2.1282495667244365,
+      "grad_norm": 1.2393336296081543,
+      "learning_rate": 1.9889295350775343e-05,
+      "loss": 0.2562,
+      "step": 615
+    },
+    {
+      "epoch": 2.145580589254766,
+      "grad_norm": 1.1414188146591187,
+      "learning_rate": 1.9726623270354314e-05,
+      "loss": 0.2365,
+      "step": 620
+    },
+    {
+      "epoch": 2.1629116117850953,
+      "grad_norm": 0.9609736204147339,
+      "learning_rate": 1.956333163791485e-05,
+      "loss": 0.2367,
+      "step": 625
+    },
+    {
+      "epoch": 2.1802426343154244,
+      "grad_norm": 1.1595170497894287,
+      "learning_rate": 1.9399441857248756e-05,
+      "loss": 0.2572,
+      "step": 630
+    },
+    {
+      "epoch": 2.197573656845754,
+      "grad_norm": 1.1949115991592407,
+      "learning_rate": 1.9234975410551397e-05,
+      "loss": 0.3066,
+      "step": 635
+    },
+    {
+      "epoch": 2.214904679376083,
+      "grad_norm": 1.422609806060791,
+      "learning_rate": 1.9069953855605805e-05,
+      "loss": 0.2792,
+      "step": 640
+    },
+    {
+      "epoch": 2.2322357019064123,
+      "grad_norm": 1.1198484897613525,
+      "learning_rate": 1.8904398822957007e-05,
+      "loss": 0.2647,
+      "step": 645
+    },
+    {
+      "epoch": 2.249566724436742,
+      "grad_norm": 1.0784331560134888,
+      "learning_rate": 1.873833201307673e-05,
+      "loss": 0.2449,
+      "step": 650
+    },
+    {
+      "epoch": 2.266897746967071,
+      "grad_norm": 1.2669086456298828,
+      "learning_rate": 1.8571775193518974e-05,
+      "loss": 0.2581,
+      "step": 655
+    },
+    {
+      "epoch": 2.2842287694974,
+      "grad_norm": 1.20986008644104,
+      "learning_rate": 1.840475019606677e-05,
+      "loss": 0.2313,
+      "step": 660
+    },
+    {
+      "epoch": 2.3015597920277298,
+      "grad_norm": 1.016635775566101,
+      "learning_rate": 1.8237278913870558e-05,
+      "loss": 0.2332,
+      "step": 665
+    },
+    {
+      "epoch": 2.318890814558059,
+      "grad_norm": 1.004381775856018,
+      "learning_rate": 1.8069383298578474e-05,
+      "loss": 0.2391,
+      "step": 670
+    },
+    {
+      "epoch": 2.336221837088388,
+      "grad_norm": 1.2095917463302612,
+      "learning_rate": 1.7901085357459e-05,
+      "loss": 0.2363,
+      "step": 675
+    },
+    {
+      "epoch": 2.3535528596187176,
+      "grad_norm": 1.021548867225647,
+      "learning_rate": 1.7732407150516324e-05,
+      "loss": 0.2394,
+      "step": 680
+    },
+    {
+      "epoch": 2.3708838821490468,
+      "grad_norm": 1.1039049625396729,
+      "learning_rate": 1.7563370787598766e-05,
+      "loss": 0.2595,
+      "step": 685
+    },
+    {
+      "epoch": 2.388214904679376,
+      "grad_norm": 1.1957190036773682,
+      "learning_rate": 1.739399842550069e-05,
+      "loss": 0.239,
+      "step": 690
+    },
+    {
+      "epoch": 2.4055459272097055,
+      "grad_norm": 1.2246172428131104,
+      "learning_rate": 1.7224312265058252e-05,
+      "loss": 0.2448,
+      "step": 695
+    },
+    {
+      "epoch": 2.4228769497400346,
+      "grad_norm": 0.9511550664901733,
+      "learning_rate": 1.7054334548239385e-05,
+      "loss": 0.2014,
+      "step": 700
+    },
+    {
+      "epoch": 2.440207972270364,
+      "grad_norm": 1.2386733293533325,
+      "learning_rate": 1.6884087555228387e-05,
+      "loss": 0.242,
+      "step": 705
+    },
+    {
+      "epoch": 2.4575389948006934,
+      "grad_norm": 1.0456082820892334,
+      "learning_rate": 1.6713593601505478e-05,
+      "loss": 0.2701,
+      "step": 710
+    },
+    {
+      "epoch": 2.4748700173310225,
+      "grad_norm": 1.1309462785720825,
+      "learning_rate": 1.6542875034921788e-05,
+      "loss": 0.2251,
+      "step": 715
+    },
+    {
+      "epoch": 2.4922010398613517,
+      "grad_norm": 1.0494320392608643,
+      "learning_rate": 1.6371954232770037e-05,
+      "loss": 0.1934,
+      "step": 720
+    },
+    {
+      "epoch": 2.5095320623916813,
+      "grad_norm": 1.5740057229995728,
+      "learning_rate": 1.62008535988514e-05,
+      "loss": 0.2481,
+      "step": 725
+    },
+    {
+      "epoch": 2.5268630849220104,
+      "grad_norm": 1.1560115814208984,
+      "learning_rate": 1.602959556053888e-05,
+      "loss": 0.2317,
+      "step": 730
+    },
+    {
+      "epoch": 2.5441941074523395,
+      "grad_norm": 1.0618700981140137,
+      "learning_rate": 1.5858202565837567e-05,
+      "loss": 0.2075,
+      "step": 735
+    },
+    {
+      "epoch": 2.561525129982669,
+      "grad_norm": 1.1966997385025024,
+      "learning_rate": 1.568669708044227e-05,
+      "loss": 0.2258,
+      "step": 740
+    },
+    {
+      "epoch": 2.5788561525129983,
+      "grad_norm": 1.2444030046463013,
+      "learning_rate": 1.5515101584792742e-05,
+      "loss": 0.2189,
+      "step": 745
+    },
+    {
+      "epoch": 2.5961871750433274,
+      "grad_norm": 1.0499557256698608,
+      "learning_rate": 1.5343438571127008e-05,
+      "loss": 0.2071,
+      "step": 750
+    },
+    {
+      "epoch": 2.613518197573657,
+      "grad_norm": 1.1848242282867432,
+      "learning_rate": 1.5171730540533165e-05,
+      "loss": 0.2309,
+      "step": 755
+    },
+    {
+      "epoch": 2.630849220103986,
+      "grad_norm": 1.2427995204925537,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2186,
+      "step": 760
+    },
+    {
+      "epoch": 2.6481802426343153,
+      "grad_norm": 1.1202366352081299,
+      "learning_rate": 1.4828269459466837e-05,
+      "loss": 0.2109,
+      "step": 765
+    },
+    {
+      "epoch": 2.665511265164645,
+      "grad_norm": 1.1815730333328247,
+      "learning_rate": 1.4656561428872996e-05,
+      "loss": 0.2234,
+      "step": 770
+    },
+    {
+      "epoch": 2.682842287694974,
+      "grad_norm": 1.2846264839172363,
+      "learning_rate": 1.4484898415207257e-05,
+      "loss": 0.1996,
+      "step": 775
+    },
+    {
+      "epoch": 2.700173310225303,
+      "grad_norm": 1.0805020332336426,
+      "learning_rate": 1.4313302919557727e-05,
+      "loss": 0.2134,
+      "step": 780
+    },
+    {
+      "epoch": 2.7175043327556327,
+      "grad_norm": 1.5587838888168335,
+      "learning_rate": 1.4141797434162437e-05,
+      "loss": 0.218,
+      "step": 785
+    },
+    {
+      "epoch": 2.734835355285962,
+      "grad_norm": 1.1907397508621216,
+      "learning_rate": 1.3970404439461129e-05,
+      "loss": 0.2024,
+      "step": 790
+    },
+    {
+      "epoch": 2.752166377816291,
+      "grad_norm": 1.0635807514190674,
+      "learning_rate": 1.3799146401148602e-05,
+      "loss": 0.1973,
+      "step": 795
+    },
+    {
+      "epoch": 2.76949740034662,
+      "grad_norm": 1.032384991645813,
+      "learning_rate": 1.3628045767229967e-05,
+      "loss": 0.1847,
+      "step": 800
+    },
+    {
+      "epoch": 2.7868284228769498,
+      "grad_norm": 1.5816798210144043,
+      "learning_rate": 1.3457124965078214e-05,
+      "loss": 0.2017,
+      "step": 805
+    },
+    {
+      "epoch": 2.804159445407279,
+      "grad_norm": 1.0850601196289062,
+      "learning_rate": 1.3286406398494524e-05,
+      "loss": 0.1986,
+      "step": 810
+    },
+    {
+      "epoch": 2.8214904679376085,
+      "grad_norm": 1.2348600625991821,
+      "learning_rate": 1.3115912444771617e-05,
+      "loss": 0.1711,
+      "step": 815
+    },
+    {
+      "epoch": 2.8388214904679376,
+      "grad_norm": 1.021596074104309,
+      "learning_rate": 1.2945665451760616e-05,
+      "loss": 0.1912,
+      "step": 820
+    },
+    {
+      "epoch": 2.856152512998267,
+      "grad_norm": 1.0768331289291382,
+      "learning_rate": 1.277568773494175e-05,
+      "loss": 0.1929,
+      "step": 825
+    },
+    {
+      "epoch": 2.873483535528596,
+      "grad_norm": 1.1603977680206299,
+      "learning_rate": 1.2606001574499316e-05,
+      "loss": 0.1789,
+      "step": 830
+    },
+    {
+      "epoch": 2.8908145580589255,
+      "grad_norm": 1.0426552295684814,
+      "learning_rate": 1.243662921240124e-05,
+      "loss": 0.179,
+      "step": 835
+    },
+    {
+      "epoch": 2.9081455805892547,
+      "grad_norm": 1.1170494556427002,
+      "learning_rate": 1.226759284948368e-05,
+      "loss": 0.1931,
+      "step": 840
+    },
+    {
+      "epoch": 2.9254766031195842,
+      "grad_norm": 1.1928642988204956,
+      "learning_rate": 1.2098914642541005e-05,
+      "loss": 0.2001,
+      "step": 845
+    },
+    {
+      "epoch": 2.9428076256499134,
+      "grad_norm": 1.1335448026657104,
+      "learning_rate": 1.1930616701421532e-05,
+      "loss": 0.1852,
+      "step": 850
+    },
+    {
+      "epoch": 2.9601386481802425,
+      "grad_norm": 1.200695514678955,
+      "learning_rate": 1.1762721086129447e-05,
+      "loss": 0.1795,
+      "step": 855
+    },
+    {
+      "epoch": 2.9774696707105717,
+      "grad_norm": 1.2488305568695068,
+      "learning_rate": 1.1595249803933232e-05,
+      "loss": 0.1796,
+      "step": 860
+    },
+    {
+      "epoch": 2.9948006932409013,
+      "grad_norm": 1.0621708631515503,
+      "learning_rate": 1.1428224806481025e-05,
+      "loss": 0.1793,
+      "step": 865
+    },
+    {
+      "epoch": 3.0103986135181975,
+      "grad_norm": 1.0290765762329102,
+      "learning_rate": 1.126166798692327e-05,
+      "loss": 0.1344,
+      "step": 870
+    },
+    {
+      "epoch": 3.027729636048527,
+      "grad_norm": 1.0009404420852661,
+      "learning_rate": 1.1095601177042995e-05,
+      "loss": 0.1426,
+      "step": 875
+    },
+    {
+      "epoch": 3.045060658578856,
+      "grad_norm": 1.0544887781143188,
+      "learning_rate": 1.09300461443942e-05,
+      "loss": 0.1417,
+      "step": 880
+    },
+    {
+      "epoch": 3.0623916811091854,
+      "grad_norm": 1.1883243322372437,
+      "learning_rate": 1.076502458944861e-05,
+      "loss": 0.1355,
+      "step": 885
+    },
+    {
+      "epoch": 3.079722703639515,
+      "grad_norm": 1.0606168508529663,
+      "learning_rate": 1.0600558142751245e-05,
+      "loss": 0.1391,
+      "step": 890
+    },
+    {
+      "epoch": 3.097053726169844,
+      "grad_norm": 1.047018051147461,
+      "learning_rate": 1.0436668362085157e-05,
+      "loss": 0.1248,
+      "step": 895
+    },
+    {
+      "epoch": 3.1143847487001732,
+      "grad_norm": 1.11692476272583,
+      "learning_rate": 1.027337672964569e-05,
+      "loss": 0.1177,
+      "step": 900
+    },
+    {
+      "epoch": 3.1317157712305024,
+      "grad_norm": 1.155957579612732,
+      "learning_rate": 1.0110704649224661e-05,
+      "loss": 0.1164,
+      "step": 905
+    },
+    {
+      "epoch": 3.149046793760832,
+      "grad_norm": 1.0759484767913818,
+      "learning_rate": 9.948673443404809e-06,
+      "loss": 0.1292,
+      "step": 910
+    },
+    {
+      "epoch": 3.166377816291161,
+      "grad_norm": 1.1999250650405884,
+      "learning_rate": 9.7873043507649e-06,
+      "loss": 0.1327,
+      "step": 915
+    },
+    {
+      "epoch": 3.1837088388214907,
+      "grad_norm": 1.1638423204421997,
+      "learning_rate": 9.626618523095825e-06,
+      "loss": 0.1432,
+      "step": 920
+    },
+    {
+      "epoch": 3.20103986135182,
+      "grad_norm": 0.9528873562812805,
+      "learning_rate": 9.46663702262811e-06,
+      "loss": 0.1212,
+      "step": 925
+    },
+    {
+      "epoch": 3.218370883882149,
+      "grad_norm": 1.1454589366912842,
+      "learning_rate": 9.307380819271092e-06,
+      "loss": 0.1292,
+      "step": 930
+    },
+    {
+      "epoch": 3.235701906412478,
+      "grad_norm": 1.0722419023513794,
+      "learning_rate": 9.148870787864297e-06,
+      "loss": 0.1336,
+      "step": 935
+    },
+    {
+      "epoch": 3.2530329289428077,
+      "grad_norm": 1.2213457822799683,
+      "learning_rate": 8.991127705441202e-06,
+      "loss": 0.1312,
+      "step": 940
+    },
+    {
+      "epoch": 3.270363951473137,
+      "grad_norm": 1.0836328268051147,
+      "learning_rate": 8.834172248505834e-06,
+      "loss": 0.1246,
+      "step": 945
+    },
+    {
+      "epoch": 3.2876949740034664,
+      "grad_norm": 1.0646871328353882,
+      "learning_rate": 8.678024990322571e-06,
+      "loss": 0.1327,
+      "step": 950
+    },
+    {
+      "epoch": 3.3050259965337956,
+      "grad_norm": 1.1547240018844604,
+      "learning_rate": 8.522706398219447e-06,
+      "loss": 0.1408,
+      "step": 955
+    },
+    {
+      "epoch": 3.3223570190641247,
+      "grad_norm": 1.2322415113449097,
+      "learning_rate": 8.368236830905371e-06,
+      "loss": 0.1312,
+      "step": 960
+    },
+    {
+      "epoch": 3.339688041594454,
+      "grad_norm": 0.8942275643348694,
+      "learning_rate": 8.214636535801532e-06,
+      "loss": 0.1102,
+      "step": 965
+    },
+    {
+      "epoch": 3.3570190641247835,
+      "grad_norm": 1.0620125532150269,
+      "learning_rate": 8.061925646387474e-06,
+      "loss": 0.1368,
+      "step": 970
+    },
+    {
+      "epoch": 3.3743500866551126,
+      "grad_norm": 1.198222041130066,
+      "learning_rate": 7.910124179562004e-06,
+      "loss": 0.1336,
+      "step": 975
+    },
+    {
+      "epoch": 3.391681109185442,
+      "grad_norm": 0.9789800047874451,
+      "learning_rate": 7.759252033019497e-06,
+      "loss": 0.1156,
+      "step": 980
+    },
+    {
+      "epoch": 3.4090121317157713,
+      "grad_norm": 0.8720641136169434,
+      "learning_rate": 7.609328982641693e-06,
+      "loss": 0.127,
+      "step": 985
+    },
+    {
+      "epoch": 3.4263431542461005,
+      "grad_norm": 1.1355503797531128,
+      "learning_rate": 7.460374679905608e-06,
+      "loss": 0.1388,
+      "step": 990
+    },
+    {
+      "epoch": 3.4436741767764296,
+      "grad_norm": 1.0707637071609497,
+      "learning_rate": 7.312408649307602e-06,
+      "loss": 0.114,
+      "step": 995
+    },
+    {
+      "epoch": 3.461005199306759,
+      "grad_norm": 0.910365641117096,
+      "learning_rate": 7.165450285804223e-06,
+      "loss": 0.1135,
+      "step": 1000
+    },
+    {
+      "epoch": 3.4783362218370883,
+      "grad_norm": 0.9218733310699463,
+      "learning_rate": 7.019518852269953e-06,
+      "loss": 0.1126,
+      "step": 1005
+    },
+    {
+      "epoch": 3.4956672443674175,
+      "grad_norm": 0.9174326658248901,
+      "learning_rate": 6.8746334769722576e-06,
+      "loss": 0.1193,
+      "step": 1010
+    },
+    {
+      "epoch": 3.512998266897747,
+      "grad_norm": 1.1110209226608276,
+      "learning_rate": 6.730813151064363e-06,
+      "loss": 0.1109,
+      "step": 1015
+    },
+    {
+      "epoch": 3.5303292894280762,
+      "grad_norm": 1.018364429473877,
+      "learning_rate": 6.588076726095902e-06,
+      "loss": 0.1406,
+      "step": 1020
+    },
+    {
+      "epoch": 3.5476603119584054,
+      "grad_norm": 0.9070491194725037,
+      "learning_rate": 6.44644291154193e-06,
+      "loss": 0.1283,
+      "step": 1025
+    },
+    {
+      "epoch": 3.564991334488735,
+      "grad_norm": 0.9356378316879272,
+      "learning_rate": 6.305930272350549e-06,
+      "loss": 0.1273,
+      "step": 1030
+    },
+    {
+      "epoch": 3.582322357019064,
+      "grad_norm": 0.9019298553466797,
+      "learning_rate": 6.1665572265094565e-06,
+      "loss": 0.1157,
+      "step": 1035
+    },
+    {
+      "epoch": 3.5996533795493937,
+      "grad_norm": 0.8685014247894287,
+      "learning_rate": 6.0283420426317606e-06,
+      "loss": 0.1059,
+      "step": 1040
+    },
+    {
+      "epoch": 3.616984402079723,
+      "grad_norm": 0.9012817740440369,
+      "learning_rate": 5.891302837561407e-06,
+      "loss": 0.1168,
+      "step": 1045
+    },
+    {
+      "epoch": 3.634315424610052,
+      "grad_norm": 0.9666245579719543,
+      "learning_rate": 5.755457573998451e-06,
+      "loss": 0.1137,
+      "step": 1050
+    },
+    {
+      "epoch": 3.651646447140381,
+      "grad_norm": 0.886913001537323,
+      "learning_rate": 5.620824058144576e-06,
+      "loss": 0.1121,
+      "step": 1055
+    },
+    {
+      "epoch": 3.6689774696707107,
+      "grad_norm": 0.8900812864303589,
+      "learning_rate": 5.487419937369112e-06,
+      "loss": 0.1131,
+      "step": 1060
+    },
+    {
+      "epoch": 3.68630849220104,
+      "grad_norm": 0.9007474780082703,
+      "learning_rate": 5.35526269789585e-06,
+      "loss": 0.1168,
+      "step": 1065
+    },
+    {
+      "epoch": 3.703639514731369,
+      "grad_norm": 0.9596942663192749,
+      "learning_rate": 5.224369662511003e-06,
+      "loss": 0.0933,
+      "step": 1070
+    },
+    {
+      "epoch": 3.7209705372616986,
+      "grad_norm": 1.0141007900238037,
+      "learning_rate": 5.094757988292613e-06,
+      "loss": 0.1073,
+      "step": 1075
+    },
+    {
+      "epoch": 3.7383015597920277,
+      "grad_norm": 0.990217387676239,
+      "learning_rate": 4.966444664361611e-06,
+      "loss": 0.1167,
+      "step": 1080
+    },
+    {
+      "epoch": 3.755632582322357,
+      "grad_norm": 0.8267380595207214,
+      "learning_rate": 4.839446509654968e-06,
+      "loss": 0.1099,
+      "step": 1085
+    },
+    {
+      "epoch": 3.7729636048526864,
+      "grad_norm": 0.762187659740448,
+      "learning_rate": 4.713780170721089e-06,
+      "loss": 0.1005,
+      "step": 1090
+    },
+    {
+      "epoch": 3.7902946273830156,
+      "grad_norm": 0.9725874662399292,
+      "learning_rate": 4.589462119537849e-06,
+      "loss": 0.123,
+      "step": 1095
+    },
+    {
+      "epoch": 3.8076256499133447,
+      "grad_norm": 0.9700238704681396,
+      "learning_rate": 4.466508651353498e-06,
+      "loss": 0.0971,
+      "step": 1100
+    },
+    {
+      "epoch": 3.8249566724436743,
+      "grad_norm": 1.0050092935562134,
+      "learning_rate": 4.344935882550701e-06,
+      "loss": 0.1115,
+      "step": 1105
+    },
+    {
+      "epoch": 3.8422876949740035,
+      "grad_norm": 0.904208242893219,
+      "learning_rate": 4.224759748534088e-06,
+      "loss": 0.1021,
+      "step": 1110
+    },
+    {
+      "epoch": 3.8596187175043326,
+      "grad_norm": 0.9154079556465149,
+      "learning_rate": 4.105996001641452e-06,
+      "loss": 0.1033,
+      "step": 1115
+    },
+    {
+      "epoch": 3.876949740034662,
+      "grad_norm": 0.9821785688400269,
+      "learning_rate": 3.988660209078994e-06,
+      "loss": 0.1152,
+      "step": 1120
+    },
+    {
+      "epoch": 3.8942807625649913,
+      "grad_norm": 0.8482122421264648,
+      "learning_rate": 3.872767750880825e-06,
+      "loss": 0.1009,
+      "step": 1125
+    },
+    {
+      "epoch": 3.9116117850953205,
+      "grad_norm": 0.8988242149353027,
+      "learning_rate": 3.7583338178929953e-06,
+      "loss": 0.1047,
+      "step": 1130
+    },
+    {
+      "epoch": 3.92894280762565,
+      "grad_norm": 1.079358458518982,
+      "learning_rate": 3.6453734097822966e-06,
+      "loss": 0.0981,
+      "step": 1135
+    },
+    {
+      "epoch": 3.946273830155979,
+      "grad_norm": 0.9375408291816711,
+      "learning_rate": 3.533901333070191e-06,
+      "loss": 0.1075,
+      "step": 1140
+    },
+    {
+      "epoch": 3.9636048526863084,
+      "grad_norm": 0.8924766778945923,
+      "learning_rate": 3.423932199191973e-06,
+      "loss": 0.107,
+      "step": 1145
+    },
+    {
+      "epoch": 3.980935875216638,
+      "grad_norm": 0.7494178414344788,
+      "learning_rate": 3.315480422581583e-06,
+      "loss": 0.0989,
+      "step": 1150
+    },
+    {
+      "epoch": 3.998266897746967,
+      "grad_norm": 0.9619389772415161,
+      "learning_rate": 3.2085602187821854e-06,
+      "loss": 0.108,
+      "step": 1155
+    },
+    {
+      "epoch": 4.013864818024263,
+      "grad_norm": 0.8392796516418457,
+      "learning_rate": 3.1031856025828293e-06,
+      "loss": 0.0871,
+      "step": 1160
+    },
+    {
+      "epoch": 4.0311958405545925,
+      "grad_norm": 0.8619660139083862,
+      "learning_rate": 2.999370386181443e-06,
+      "loss": 0.0908,
+      "step": 1165
+    },
+    {
+      "epoch": 4.048526863084922,
+      "grad_norm": 0.9306987524032593,
+      "learning_rate": 2.8971281773743786e-06,
+      "loss": 0.0826,
+      "step": 1170
+    },
+    {
+      "epoch": 4.065857885615252,
+      "grad_norm": 0.8258434534072876,
+      "learning_rate": 2.7964723777727214e-06,
+      "loss": 0.0909,
+      "step": 1175
+    },
+    {
+      "epoch": 4.083188908145581,
+      "grad_norm": 0.7605803608894348,
+      "learning_rate": 2.6974161810456605e-06,
+      "loss": 0.0805,
+      "step": 1180
+    },
+    {
+      "epoch": 4.10051993067591,
+      "grad_norm": 0.8231697082519531,
+      "learning_rate": 2.5999725711911044e-06,
+      "loss": 0.0837,
+      "step": 1185
+    },
+    {
+      "epoch": 4.117850953206239,
+      "grad_norm": 0.8096256852149963,
+      "learning_rate": 2.504154320833751e-06,
+      "loss": 0.0802,
+      "step": 1190
+    },
+    {
+      "epoch": 4.135181975736568,
+      "grad_norm": 1.033969759941101,
+      "learning_rate": 2.409973989550925e-06,
+      "loss": 0.0811,
+      "step": 1195
+    },
+    {
+      "epoch": 4.152512998266897,
+      "grad_norm": 0.7165153622627258,
+      "learning_rate": 2.3174439222262716e-06,
+      "loss": 0.0743,
+      "step": 1200
+    },
+    {
+      "epoch": 4.169844020797227,
+      "grad_norm": 0.7980844378471375,
+      "learning_rate": 2.2265762474316564e-06,
+      "loss": 0.0831,
+      "step": 1205
+    },
+    {
+      "epoch": 4.1871750433275565,
+      "grad_norm": 0.8102248907089233,
+      "learning_rate": 2.1373828758373576e-06,
+      "loss": 0.0705,
+      "step": 1210
+    },
+    {
+      "epoch": 4.204506065857886,
+      "grad_norm": 0.830589234828949,
+      "learning_rate": 2.049875498650874e-06,
+      "loss": 0.0888,
+      "step": 1215
+    },
+    {
+      "epoch": 4.221837088388215,
+      "grad_norm": 0.7752622961997986,
+      "learning_rate": 1.9640655860844554e-06,
+      "loss": 0.0844,
+      "step": 1220
+    },
+    {
+      "epoch": 4.239168110918544,
+      "grad_norm": 0.8469278216362,
+      "learning_rate": 1.879964385851644e-06,
+      "loss": 0.0852,
+      "step": 1225
+    },
+    {
+      "epoch": 4.256499133448873,
+      "grad_norm": 0.7288994193077087,
+      "learning_rate": 1.797582921692929e-06,
+      "loss": 0.0958,
+      "step": 1230
+    },
+    {
+      "epoch": 4.273830155979203,
+      "grad_norm": 0.8898922801017761,
+      "learning_rate": 1.716931991930813e-06,
+      "loss": 0.0739,
+      "step": 1235
+    },
+    {
+      "epoch": 4.291161178509532,
+      "grad_norm": 0.842789351940155,
+      "learning_rate": 1.638022168054385e-06,
+      "loss": 0.0758,
+      "step": 1240
+    },
+    {
+      "epoch": 4.308492201039861,
+      "grad_norm": 0.7576664090156555,
+      "learning_rate": 1.5608637933336372e-06,
+      "loss": 0.0825,
+      "step": 1245
+    },
+    {
+      "epoch": 4.325823223570191,
+      "grad_norm": 0.7729648947715759,
+      "learning_rate": 1.4854669814637145e-06,
+      "loss": 0.0714,
+      "step": 1250
+    },
+    {
+      "epoch": 4.34315424610052,
+      "grad_norm": 0.8112753033638,
+      "learning_rate": 1.4118416152392227e-06,
+      "loss": 0.0776,
+      "step": 1255
+    },
+    {
+      "epoch": 4.360485268630849,
+      "grad_norm": 0.7901812791824341,
+      "learning_rate": 1.3399973452588322e-06,
+      "loss": 0.0846,
+      "step": 1260
+    },
+    {
+      "epoch": 4.377816291161179,
+      "grad_norm": 0.8140977621078491,
+      "learning_rate": 1.269943588660304e-06,
+      "loss": 0.0765,
+      "step": 1265
+    },
+    {
+      "epoch": 4.395147313691508,
+      "grad_norm": 0.9248760938644409,
+      "learning_rate": 1.2016895278861228e-06,
+      "loss": 0.08,
+      "step": 1270
+    },
+    {
+      "epoch": 4.412478336221837,
+      "grad_norm": 0.7612517476081848,
+      "learning_rate": 1.13524410947987e-06,
+      "loss": 0.0844,
+      "step": 1275
+    },
+    {
+      "epoch": 4.429809358752166,
+      "grad_norm": 0.6932576298713684,
+      "learning_rate": 1.0706160429135652e-06,
+      "loss": 0.0837,
+      "step": 1280
+    },
+    {
+      "epoch": 4.4471403812824954,
+      "grad_norm": 0.7995143532752991,
+      "learning_rate": 1.0078137994460284e-06,
+      "loss": 0.0675,
+      "step": 1285
+    },
+    {
+      "epoch": 4.464471403812825,
+      "grad_norm": 1.0186465978622437,
+      "learning_rate": 9.468456110125079e-07,
+      "loss": 0.0715,
+      "step": 1290
+    },
+    {
+      "epoch": 4.481802426343155,
+      "grad_norm": 0.8060331344604492,
+      "learning_rate": 8.877194691456624e-07,
+      "loss": 0.0738,
+      "step": 1295
+    },
+    {
+      "epoch": 4.499133448873484,
+      "grad_norm": 0.8241632580757141,
+      "learning_rate": 8.304431239280364e-07,
+      "loss": 0.0769,
+      "step": 1300
+    },
+    {
+      "epoch": 4.516464471403813,
+      "grad_norm": 0.6952365636825562,
+      "learning_rate": 7.750240829762245e-07,
+      "loss": 0.0798,
+      "step": 1305
+    },
+    {
+      "epoch": 4.533795493934142,
+      "grad_norm": 0.7716186046600342,
+      "learning_rate": 7.214696104567836e-07,
+      "loss": 0.0752,
+      "step": 1310
+    },
+    {
+      "epoch": 4.551126516464471,
+      "grad_norm": 0.8005116581916809,
+      "learning_rate": 6.697867261340651e-07,
+      "loss": 0.0716,
+      "step": 1315
+    },
+    {
+      "epoch": 4.5684575389948,
+      "grad_norm": 0.7815414667129517,
+      "learning_rate": 6.199822044500913e-07,
+      "loss": 0.0847,
+      "step": 1320
+    },
+    {
+      "epoch": 4.58578856152513,
+      "grad_norm": 0.7633384466171265,
+      "learning_rate": 5.720625736365775e-07,
+      "loss": 0.0785,
+      "step": 1325
+    },
+    {
+      "epoch": 4.6031195840554595,
+      "grad_norm": 0.7474676370620728,
+      "learning_rate": 5.260341148592279e-07,
+      "loss": 0.0833,
+      "step": 1330
+    },
+    {
+      "epoch": 4.620450606585789,
+      "grad_norm": 0.7778022885322571,
+      "learning_rate": 4.819028613944277e-07,
+      "loss": 0.0859,
+      "step": 1335
+    },
+    {
+      "epoch": 4.637781629116118,
+      "grad_norm": 0.7165287137031555,
+      "learning_rate": 4.3967459783840535e-07,
+      "loss": 0.0828,
+      "step": 1340
+    },
+    {
+      "epoch": 4.655112651646447,
+      "grad_norm": 0.7916089296340942,
+      "learning_rate": 3.993548593490165e-07,
+      "loss": 0.0788,
+      "step": 1345
+    },
+    {
+      "epoch": 4.672443674176776,
+      "grad_norm": 0.8870761394500732,
+      "learning_rate": 3.6094893092020854e-07,
+      "loss": 0.0861,
+      "step": 1350
+    },
+    {
+      "epoch": 4.689774696707106,
+      "grad_norm": 0.745093584060669,
+      "learning_rate": 3.2446184668927016e-07,
+      "loss": 0.0845,
+      "step": 1355
+    },
+    {
+      "epoch": 4.707105719237435,
+      "grad_norm": 0.7193507552146912,
+      "learning_rate": 2.898983892769874e-07,
+      "loss": 0.0801,
+      "step": 1360
+    },
+    {
+      "epoch": 4.724436741767764,
+      "grad_norm": 0.7658344507217407,
+      "learning_rate": 2.572630891607403e-07,
+      "loss": 0.0775,
+      "step": 1365
+    },
+    {
+      "epoch": 4.7417677642980935,
+      "grad_norm": 0.8309527039527893,
+      "learning_rate": 2.2656022408065747e-07,
+      "loss": 0.0757,
+      "step": 1370
+    },
+    {
+      "epoch": 4.759098786828423,
+      "grad_norm": 0.7888296842575073,
+      "learning_rate": 1.9779381847891852e-07,
+      "loss": 0.077,
+      "step": 1375
+    },
+    {
+      "epoch": 4.776429809358752,
+      "grad_norm": 0.8518990278244019,
+      "learning_rate": 1.7096764297222068e-07,
+      "loss": 0.0833,
+      "step": 1380
+    },
+    {
+      "epoch": 4.793760831889081,
+      "grad_norm": 0.673783540725708,
+      "learning_rate": 1.460852138575608e-07,
+      "loss": 0.0798,
+      "step": 1385
+    },
+    {
+      "epoch": 4.811091854419411,
+      "grad_norm": 0.6517499685287476,
+      "learning_rate": 1.231497926513092e-07,
+      "loss": 0.0739,
+      "step": 1390
+    },
+    {
+      "epoch": 4.82842287694974,
+      "grad_norm": 0.7029514908790588,
+      "learning_rate": 1.0216438566170827e-07,
+      "loss": 0.0807,
+      "step": 1395
+    },
+    {
+      "epoch": 4.845753899480069,
+      "grad_norm": 0.8831659555435181,
+      "learning_rate": 8.313174359481424e-08,
+      "loss": 0.0823,
+      "step": 1400
+    },
+    {
+      "epoch": 4.863084922010398,
+      "grad_norm": 0.6417370438575745,
+      "learning_rate": 6.605436119394404e-08,
+      "loss": 0.0691,
+      "step": 1405
+    },
+    {
+      "epoch": 4.880415944540728,
+      "grad_norm": 0.6646305918693542,
+      "learning_rate": 5.09344769126685e-08,
+      "loss": 0.0762,
+      "step": 1410
+    },
+    {
+      "epoch": 4.897746967071058,
+      "grad_norm": 0.7420198917388916,
+      "learning_rate": 3.777407262140209e-08,
+      "loss": 0.0778,
+      "step": 1415
+    },
+    {
+      "epoch": 4.915077989601387,
+      "grad_norm": 0.7527610659599304,
+      "learning_rate": 2.657487334762898e-08,
+      "loss": 0.0715,
+      "step": 1420
+    },
+    {
+      "epoch": 4.932409012131716,
+      "grad_norm": 0.8601207733154297,
+      "learning_rate": 1.7338347049787317e-08,
+      "loss": 0.0834,
+      "step": 1425
+    },
+    {
+      "epoch": 4.949740034662045,
+      "grad_norm": 0.7180348634719849,
+      "learning_rate": 1.0065704424854771e-08,
+      "loss": 0.079,
+      "step": 1430
+    },
+    {
+      "epoch": 4.967071057192374,
+      "grad_norm": 0.7777925133705139,
+      "learning_rate": 4.757898749653822e-09,
+      "loss": 0.078,
+      "step": 1435
+    },
+    {
+      "epoch": 4.984402079722703,
+      "grad_norm": 0.6979928612709045,
+      "learning_rate": 1.4156257559033714e-09,
+      "loss": 0.079,
+      "step": 1440
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.058016300201416,
+      "learning_rate": 3.932353901503483e-11,
+      "loss": 0.0881,
+      "step": 1445
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 1445,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.049922280594604e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

1_128_e5_3e-5/checkpoint-1445/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b529bde561655fa6f1efd722a181ca5b0d63578ce8f7cd71cb25200956f56df
+size 7736

1_128_e5_3e-5/checkpoint-1445/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/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, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in 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,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, 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_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``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    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("--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_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)

1_128_e5_3e-5/checkpoint-289/README.md

@@ -0,0 +1,202 @@
+---
+base_model: ibm-granite/granite-3.3-8b-base
+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 0.15.2

1_128_e5_3e-5/checkpoint-289/adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
+  "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": 256,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "k_proj",
+    "v_proj",
+    "o_proj",
+    "up_proj",
+    "q_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

1_128_e5_3e-5/checkpoint-289/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3066ebce96d6e0fdadf27a547b3387d49eb57d461f1f26e069986172b60786e
+size 791751704

1_128_e5_3e-5/checkpoint-289/latest

@@ -0,0 +1 @@
+global_step289

1_128_e5_3e-5/checkpoint-289/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f3d6c903ec24dc245f39b3daff6d05b81a5f27b64bc4ad7e6f12f760729dbb2
+size 15920

1_128_e5_3e-5/checkpoint-289/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0961be97a7dd99ba1dee72cc3c0b208ff54eab836c7837aaab0e55d5ccb16439
+size 15920

1_128_e5_3e-5/checkpoint-289/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1598be54c33a6bc3c6cc27a80335a5f0dc5af6598ebc18d2b8f0ae828a6855f1
+size 15920