Upload 13 files
Browse files- BHC_Test1/README.md +202 -0
- BHC_Test1/adapter_config.json +39 -0
- BHC_Test1/adapter_model.safetensors +3 -0
- BHC_Test1/latest +1 -0
- BHC_Test1/rng_state.pth +3 -0
- BHC_Test1/scheduler.pt +3 -0
- BHC_Test1/special_tokens_map.json +24 -0
- BHC_Test1/tokenizer.json +0 -0
- BHC_Test1/tokenizer.model +3 -0
- BHC_Test1/tokenizer_config.json +47 -0
- BHC_Test1/trainer_state.json +3234 -0
- BHC_Test1/training_args.bin +3 -0
- BHC_Test1/zero_to_fp32.py +760 -0
BHC_Test1/README.md
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: /home/bingxing2/ailab/gaoben/models/Mistral-7B-Instruct/Mistral-7B-Instruct-v0.2/AI-ModelScope/Mistral-7B-Instruct-v0___2
|
| 3 |
+
library_name: peft
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# Model Card for Model ID
|
| 7 |
+
|
| 8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
### Model Description
|
| 15 |
+
|
| 16 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [More Information Needed]
|
| 21 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 22 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 23 |
+
- **Model type:** [More Information Needed]
|
| 24 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 25 |
+
- **License:** [More Information Needed]
|
| 26 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 27 |
+
|
| 28 |
+
### Model Sources [optional]
|
| 29 |
+
|
| 30 |
+
<!-- Provide the basic links for the model. -->
|
| 31 |
+
|
| 32 |
+
- **Repository:** [More Information Needed]
|
| 33 |
+
- **Paper [optional]:** [More Information Needed]
|
| 34 |
+
- **Demo [optional]:** [More Information Needed]
|
| 35 |
+
|
| 36 |
+
## Uses
|
| 37 |
+
|
| 38 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 39 |
+
|
| 40 |
+
### Direct Use
|
| 41 |
+
|
| 42 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 43 |
+
|
| 44 |
+
[More Information Needed]
|
| 45 |
+
|
| 46 |
+
### Downstream Use [optional]
|
| 47 |
+
|
| 48 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 49 |
+
|
| 50 |
+
[More Information Needed]
|
| 51 |
+
|
| 52 |
+
### Out-of-Scope Use
|
| 53 |
+
|
| 54 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 55 |
+
|
| 56 |
+
[More Information Needed]
|
| 57 |
+
|
| 58 |
+
## Bias, Risks, and Limitations
|
| 59 |
+
|
| 60 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 61 |
+
|
| 62 |
+
[More Information Needed]
|
| 63 |
+
|
| 64 |
+
### Recommendations
|
| 65 |
+
|
| 66 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 67 |
+
|
| 68 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 69 |
+
|
| 70 |
+
## How to Get Started with the Model
|
| 71 |
+
|
| 72 |
+
Use the code below to get started with the model.
|
| 73 |
+
|
| 74 |
+
[More Information Needed]
|
| 75 |
+
|
| 76 |
+
## Training Details
|
| 77 |
+
|
| 78 |
+
### Training Data
|
| 79 |
+
|
| 80 |
+
<!-- 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. -->
|
| 81 |
+
|
| 82 |
+
[More Information Needed]
|
| 83 |
+
|
| 84 |
+
### Training Procedure
|
| 85 |
+
|
| 86 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 87 |
+
|
| 88 |
+
#### Preprocessing [optional]
|
| 89 |
+
|
| 90 |
+
[More Information Needed]
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
#### Training Hyperparameters
|
| 94 |
+
|
| 95 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 96 |
+
|
| 97 |
+
#### Speeds, Sizes, Times [optional]
|
| 98 |
+
|
| 99 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 100 |
+
|
| 101 |
+
[More Information Needed]
|
| 102 |
+
|
| 103 |
+
## Evaluation
|
| 104 |
+
|
| 105 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 106 |
+
|
| 107 |
+
### Testing Data, Factors & Metrics
|
| 108 |
+
|
| 109 |
+
#### Testing Data
|
| 110 |
+
|
| 111 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 112 |
+
|
| 113 |
+
[More Information Needed]
|
| 114 |
+
|
| 115 |
+
#### Factors
|
| 116 |
+
|
| 117 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 118 |
+
|
| 119 |
+
[More Information Needed]
|
| 120 |
+
|
| 121 |
+
#### Metrics
|
| 122 |
+
|
| 123 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 124 |
+
|
| 125 |
+
[More Information Needed]
|
| 126 |
+
|
| 127 |
+
### Results
|
| 128 |
+
|
| 129 |
+
[More Information Needed]
|
| 130 |
+
|
| 131 |
+
#### Summary
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
|
| 135 |
+
## Model Examination [optional]
|
| 136 |
+
|
| 137 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 138 |
+
|
| 139 |
+
[More Information Needed]
|
| 140 |
+
|
| 141 |
+
## Environmental Impact
|
| 142 |
+
|
| 143 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 144 |
+
|
| 145 |
+
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).
|
| 146 |
+
|
| 147 |
+
- **Hardware Type:** [More Information Needed]
|
| 148 |
+
- **Hours used:** [More Information Needed]
|
| 149 |
+
- **Cloud Provider:** [More Information Needed]
|
| 150 |
+
- **Compute Region:** [More Information Needed]
|
| 151 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 152 |
+
|
| 153 |
+
## Technical Specifications [optional]
|
| 154 |
+
|
| 155 |
+
### Model Architecture and Objective
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
### Compute Infrastructure
|
| 160 |
+
|
| 161 |
+
[More Information Needed]
|
| 162 |
+
|
| 163 |
+
#### Hardware
|
| 164 |
+
|
| 165 |
+
[More Information Needed]
|
| 166 |
+
|
| 167 |
+
#### Software
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
## Citation [optional]
|
| 172 |
+
|
| 173 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 174 |
+
|
| 175 |
+
**BibTeX:**
|
| 176 |
+
|
| 177 |
+
[More Information Needed]
|
| 178 |
+
|
| 179 |
+
**APA:**
|
| 180 |
+
|
| 181 |
+
[More Information Needed]
|
| 182 |
+
|
| 183 |
+
## Glossary [optional]
|
| 184 |
+
|
| 185 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 186 |
+
|
| 187 |
+
[More Information Needed]
|
| 188 |
+
|
| 189 |
+
## More Information [optional]
|
| 190 |
+
|
| 191 |
+
[More Information Needed]
|
| 192 |
+
|
| 193 |
+
## Model Card Authors [optional]
|
| 194 |
+
|
| 195 |
+
[More Information Needed]
|
| 196 |
+
|
| 197 |
+
## Model Card Contact
|
| 198 |
+
|
| 199 |
+
[More Information Needed]
|
| 200 |
+
### Framework versions
|
| 201 |
+
|
| 202 |
+
- PEFT 0.15.1
|
BHC_Test1/adapter_config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "/home/bingxing2/ailab/gaoben/models/Mistral-7B-Instruct/Mistral-7B-Instruct-v0.2/AI-ModelScope/Mistral-7B-Instruct-v0___2",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": null,
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 32,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.0,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"r": 16,
|
| 24 |
+
"rank_pattern": {},
|
| 25 |
+
"revision": null,
|
| 26 |
+
"target_modules": [
|
| 27 |
+
"o_proj",
|
| 28 |
+
"gate_proj",
|
| 29 |
+
"q_proj",
|
| 30 |
+
"v_proj",
|
| 31 |
+
"up_proj",
|
| 32 |
+
"down_proj",
|
| 33 |
+
"k_proj"
|
| 34 |
+
],
|
| 35 |
+
"task_type": "CAUSAL_LM",
|
| 36 |
+
"trainable_token_indices": null,
|
| 37 |
+
"use_dora": false,
|
| 38 |
+
"use_rslora": false
|
| 39 |
+
}
|
BHC_Test1/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c3c6253a9f2e2b59bb23380c6de138cadd49183138bcfc62923a4858f9540d9
|
| 3 |
+
size 83946192
|
BHC_Test1/latest
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
global_step3982
|
BHC_Test1/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0886b5e6b4eb6c54d008834760837138a75d96ac8156628b1654cc847af0e990
|
| 3 |
+
size 14244
|
BHC_Test1/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b7c8d9564d916465ba3eebeffbbcae87e1948f601895c602a2242e200667fcc
|
| 3 |
+
size 1064
|
BHC_Test1/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
BHC_Test1/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
BHC_Test1/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
|
| 3 |
+
size 493443
|
BHC_Test1/tokenizer_config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"additional_special_tokens": [],
|
| 32 |
+
"bos_token": "<s>",
|
| 33 |
+
"chat_template": "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ message['content'] + eos_token}}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}",
|
| 34 |
+
"clean_up_tokenization_spaces": false,
|
| 35 |
+
"eos_token": "</s>",
|
| 36 |
+
"extra_special_tokens": {},
|
| 37 |
+
"legacy": true,
|
| 38 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 39 |
+
"pad_token": "</s>",
|
| 40 |
+
"padding_side": "right",
|
| 41 |
+
"sp_model_kwargs": {},
|
| 42 |
+
"spaces_between_special_tokens": false,
|
| 43 |
+
"split_special_tokens": false,
|
| 44 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 45 |
+
"unk_token": "<unk>",
|
| 46 |
+
"use_default_system_prompt": false
|
| 47 |
+
}
|
BHC_Test1/trainer_state.json
ADDED
|
@@ -0,0 +1,3234 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 20.837148463047743,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.052321778940483975,
|
| 14 |
+
"grad_norm": 0.5172898769378662,
|
| 15 |
+
"learning_rate": 4.9999695642048685e-05,
|
| 16 |
+
"loss": 1.2268,
|
| 17 |
+
"num_input_tokens_seen": 55264,
|
| 18 |
+
"step": 10
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.10464355788096795,
|
| 22 |
+
"grad_norm": 0.2547127306461334,
|
| 23 |
+
"learning_rate": 4.9998643550002796e-05,
|
| 24 |
+
"loss": 0.163,
|
| 25 |
+
"num_input_tokens_seen": 111136,
|
| 26 |
+
"step": 20
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.15696533682145192,
|
| 30 |
+
"grad_norm": 0.19801495969295502,
|
| 31 |
+
"learning_rate": 4.999683999797514e-05,
|
| 32 |
+
"loss": 0.1584,
|
| 33 |
+
"num_input_tokens_seen": 167232,
|
| 34 |
+
"step": 30
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.2092871157619359,
|
| 38 |
+
"grad_norm": 0.17607145011425018,
|
| 39 |
+
"learning_rate": 4.999428504018057e-05,
|
| 40 |
+
"loss": 0.1571,
|
| 41 |
+
"num_input_tokens_seen": 222368,
|
| 42 |
+
"step": 40
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.2616088947024199,
|
| 46 |
+
"grad_norm": 0.123909592628479,
|
| 47 |
+
"learning_rate": 4.999097875342117e-05,
|
| 48 |
+
"loss": 0.1535,
|
| 49 |
+
"num_input_tokens_seen": 280080,
|
| 50 |
+
"step": 50
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.31393067364290383,
|
| 54 |
+
"grad_norm": 0.15530619025230408,
|
| 55 |
+
"learning_rate": 4.998692123708403e-05,
|
| 56 |
+
"loss": 0.1592,
|
| 57 |
+
"num_input_tokens_seen": 336144,
|
| 58 |
+
"step": 60
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.36625245258338784,
|
| 62 |
+
"grad_norm": 0.08514747023582458,
|
| 63 |
+
"learning_rate": 4.998211261313822e-05,
|
| 64 |
+
"loss": 0.1581,
|
| 65 |
+
"num_input_tokens_seen": 392928,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.4185742315238718,
|
| 70 |
+
"grad_norm": 0.08711712062358856,
|
| 71 |
+
"learning_rate": 4.997655302613111e-05,
|
| 72 |
+
"loss": 0.1479,
|
| 73 |
+
"num_input_tokens_seen": 448288,
|
| 74 |
+
"step": 80
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.4708960104643558,
|
| 78 |
+
"grad_norm": 0.14576469361782074,
|
| 79 |
+
"learning_rate": 4.997024264318406e-05,
|
| 80 |
+
"loss": 0.1574,
|
| 81 |
+
"num_input_tokens_seen": 504112,
|
| 82 |
+
"step": 90
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.5232177894048398,
|
| 86 |
+
"grad_norm": 0.14594055712223053,
|
| 87 |
+
"learning_rate": 4.9963181653987373e-05,
|
| 88 |
+
"loss": 0.1532,
|
| 89 |
+
"num_input_tokens_seen": 559824,
|
| 90 |
+
"step": 100
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.5755395683453237,
|
| 94 |
+
"grad_norm": 1.3401381969451904,
|
| 95 |
+
"learning_rate": 4.99553702707946e-05,
|
| 96 |
+
"loss": 0.1527,
|
| 97 |
+
"num_input_tokens_seen": 615856,
|
| 98 |
+
"step": 110
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.6278613472858077,
|
| 102 |
+
"grad_norm": 0.21902425587177277,
|
| 103 |
+
"learning_rate": 4.9946808728416143e-05,
|
| 104 |
+
"loss": 0.1464,
|
| 105 |
+
"num_input_tokens_seen": 672448,
|
| 106 |
+
"step": 120
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.6801831262262917,
|
| 110 |
+
"grad_norm": 0.24769121408462524,
|
| 111 |
+
"learning_rate": 4.993749728421224e-05,
|
| 112 |
+
"loss": 0.1545,
|
| 113 |
+
"num_input_tokens_seen": 727840,
|
| 114 |
+
"step": 130
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.7325049051667757,
|
| 118 |
+
"grad_norm": 0.2848469018936157,
|
| 119 |
+
"learning_rate": 4.992743621808518e-05,
|
| 120 |
+
"loss": 0.1448,
|
| 121 |
+
"num_input_tokens_seen": 784496,
|
| 122 |
+
"step": 140
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.7848266841072596,
|
| 126 |
+
"grad_norm": 0.31721895933151245,
|
| 127 |
+
"learning_rate": 4.991662583247092e-05,
|
| 128 |
+
"loss": 0.1458,
|
| 129 |
+
"num_input_tokens_seen": 840784,
|
| 130 |
+
"step": 150
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.8371484630477436,
|
| 134 |
+
"grad_norm": 0.21474598348140717,
|
| 135 |
+
"learning_rate": 4.9905066452329964e-05,
|
| 136 |
+
"loss": 0.1445,
|
| 137 |
+
"num_input_tokens_seen": 897920,
|
| 138 |
+
"step": 160
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.8894702419882276,
|
| 142 |
+
"grad_norm": 0.5110652446746826,
|
| 143 |
+
"learning_rate": 4.9892758425137643e-05,
|
| 144 |
+
"loss": 0.1471,
|
| 145 |
+
"num_input_tokens_seen": 953184,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.9417920209287116,
|
| 150 |
+
"grad_norm": 0.29877740144729614,
|
| 151 |
+
"learning_rate": 4.987970212087363e-05,
|
| 152 |
+
"loss": 0.1464,
|
| 153 |
+
"num_input_tokens_seen": 1009168,
|
| 154 |
+
"step": 180
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.9941137998691956,
|
| 158 |
+
"grad_norm": 0.16597020626068115,
|
| 159 |
+
"learning_rate": 4.986589793201081e-05,
|
| 160 |
+
"loss": 0.1407,
|
| 161 |
+
"num_input_tokens_seen": 1065200,
|
| 162 |
+
"step": 190
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.0418574231523872,
|
| 166 |
+
"grad_norm": 0.2527844309806824,
|
| 167 |
+
"learning_rate": 4.985134627350353e-05,
|
| 168 |
+
"loss": 0.1492,
|
| 169 |
+
"num_input_tokens_seen": 1115736,
|
| 170 |
+
"step": 200
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.0941792020928711,
|
| 174 |
+
"grad_norm": 0.19597752392292023,
|
| 175 |
+
"learning_rate": 4.9836047582775084e-05,
|
| 176 |
+
"loss": 0.1431,
|
| 177 |
+
"num_input_tokens_seen": 1171656,
|
| 178 |
+
"step": 210
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.146500981033355,
|
| 182 |
+
"grad_norm": 0.23703065514564514,
|
| 183 |
+
"learning_rate": 4.9820002319704576e-05,
|
| 184 |
+
"loss": 0.1434,
|
| 185 |
+
"num_input_tokens_seen": 1229224,
|
| 186 |
+
"step": 220
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.198822759973839,
|
| 190 |
+
"grad_norm": 0.19536039233207703,
|
| 191 |
+
"learning_rate": 4.98032109666131e-05,
|
| 192 |
+
"loss": 0.14,
|
| 193 |
+
"num_input_tokens_seen": 1285992,
|
| 194 |
+
"step": 230
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.251144538914323,
|
| 198 |
+
"grad_norm": 0.3111051321029663,
|
| 199 |
+
"learning_rate": 4.978567402824924e-05,
|
| 200 |
+
"loss": 0.1355,
|
| 201 |
+
"num_input_tokens_seen": 1341256,
|
| 202 |
+
"step": 240
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 1.3034663178548072,
|
| 206 |
+
"grad_norm": 0.19106437265872955,
|
| 207 |
+
"learning_rate": 4.97673920317739e-05,
|
| 208 |
+
"loss": 0.1425,
|
| 209 |
+
"num_input_tokens_seen": 1396392,
|
| 210 |
+
"step": 250
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.355788096795291,
|
| 214 |
+
"grad_norm": 0.30303388833999634,
|
| 215 |
+
"learning_rate": 4.9748365526744423e-05,
|
| 216 |
+
"loss": 0.1343,
|
| 217 |
+
"num_input_tokens_seen": 1454088,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.408109875735775,
|
| 222 |
+
"grad_norm": 0.3765646815299988,
|
| 223 |
+
"learning_rate": 4.972859508509816e-05,
|
| 224 |
+
"loss": 0.1343,
|
| 225 |
+
"num_input_tokens_seen": 1509656,
|
| 226 |
+
"step": 270
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.460431654676259,
|
| 230 |
+
"grad_norm": 0.3608955442905426,
|
| 231 |
+
"learning_rate": 4.9708081301135155e-05,
|
| 232 |
+
"loss": 0.1377,
|
| 233 |
+
"num_input_tokens_seen": 1565976,
|
| 234 |
+
"step": 280
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.512753433616743,
|
| 238 |
+
"grad_norm": 0.19421738386154175,
|
| 239 |
+
"learning_rate": 4.9686824791500396e-05,
|
| 240 |
+
"loss": 0.1381,
|
| 241 |
+
"num_input_tokens_seen": 1621544,
|
| 242 |
+
"step": 290
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.565075212557227,
|
| 246 |
+
"grad_norm": 0.6034450531005859,
|
| 247 |
+
"learning_rate": 4.96648261951652e-05,
|
| 248 |
+
"loss": 0.1339,
|
| 249 |
+
"num_input_tokens_seen": 1677624,
|
| 250 |
+
"step": 300
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.6173969914977109,
|
| 254 |
+
"grad_norm": 0.28776681423187256,
|
| 255 |
+
"learning_rate": 4.964208617340803e-05,
|
| 256 |
+
"loss": 0.1351,
|
| 257 |
+
"num_input_tokens_seen": 1733192,
|
| 258 |
+
"step": 310
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 1.669718770438195,
|
| 262 |
+
"grad_norm": 0.15169182419776917,
|
| 263 |
+
"learning_rate": 4.961860540979464e-05,
|
| 264 |
+
"loss": 0.1406,
|
| 265 |
+
"num_input_tokens_seen": 1788648,
|
| 266 |
+
"step": 320
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.7220405493786788,
|
| 270 |
+
"grad_norm": 0.24536730349063873,
|
| 271 |
+
"learning_rate": 4.9594384610157483e-05,
|
| 272 |
+
"loss": 0.1354,
|
| 273 |
+
"num_input_tokens_seen": 1844360,
|
| 274 |
+
"step": 330
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.774362328319163,
|
| 278 |
+
"grad_norm": 0.4901825487613678,
|
| 279 |
+
"learning_rate": 4.9569424502574544e-05,
|
| 280 |
+
"loss": 0.1318,
|
| 281 |
+
"num_input_tokens_seen": 1900888,
|
| 282 |
+
"step": 340
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.8266841072596467,
|
| 286 |
+
"grad_norm": 0.33420756459236145,
|
| 287 |
+
"learning_rate": 4.954372583734741e-05,
|
| 288 |
+
"loss": 0.1324,
|
| 289 |
+
"num_input_tokens_seen": 1956840,
|
| 290 |
+
"step": 350
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.8790058862001309,
|
| 294 |
+
"grad_norm": 0.24992604553699493,
|
| 295 |
+
"learning_rate": 4.951728938697872e-05,
|
| 296 |
+
"loss": 0.1352,
|
| 297 |
+
"num_input_tokens_seen": 2013624,
|
| 298 |
+
"step": 360
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.9313276651406148,
|
| 302 |
+
"grad_norm": 0.3095349669456482,
|
| 303 |
+
"learning_rate": 4.9490115946148985e-05,
|
| 304 |
+
"loss": 0.1332,
|
| 305 |
+
"num_input_tokens_seen": 2069768,
|
| 306 |
+
"step": 370
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 1.9836494440810988,
|
| 310 |
+
"grad_norm": 0.35465294122695923,
|
| 311 |
+
"learning_rate": 4.946220633169266e-05,
|
| 312 |
+
"loss": 0.128,
|
| 313 |
+
"num_input_tokens_seen": 2125768,
|
| 314 |
+
"step": 380
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.0313930673642906,
|
| 318 |
+
"grad_norm": 0.2656111419200897,
|
| 319 |
+
"learning_rate": 4.943356138257359e-05,
|
| 320 |
+
"loss": 0.1356,
|
| 321 |
+
"num_input_tokens_seen": 2176424,
|
| 322 |
+
"step": 390
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 2.0837148463047743,
|
| 326 |
+
"grad_norm": 0.16828453540802002,
|
| 327 |
+
"learning_rate": 4.940418195985983e-05,
|
| 328 |
+
"loss": 0.1332,
|
| 329 |
+
"num_input_tokens_seen": 2232200,
|
| 330 |
+
"step": 400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.1360366252452585,
|
| 334 |
+
"grad_norm": 0.22813495993614197,
|
| 335 |
+
"learning_rate": 4.9374068946697695e-05,
|
| 336 |
+
"loss": 0.1296,
|
| 337 |
+
"num_input_tokens_seen": 2289432,
|
| 338 |
+
"step": 410
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.1883584041857422,
|
| 342 |
+
"grad_norm": 0.30772924423217773,
|
| 343 |
+
"learning_rate": 4.934322324828529e-05,
|
| 344 |
+
"loss": 0.1288,
|
| 345 |
+
"num_input_tokens_seen": 2345576,
|
| 346 |
+
"step": 420
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.2406801831262264,
|
| 350 |
+
"grad_norm": 0.29362913966178894,
|
| 351 |
+
"learning_rate": 4.931164579184523e-05,
|
| 352 |
+
"loss": 0.1307,
|
| 353 |
+
"num_input_tokens_seen": 2400696,
|
| 354 |
+
"step": 430
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.29300196206671,
|
| 358 |
+
"grad_norm": 0.21813659369945526,
|
| 359 |
+
"learning_rate": 4.9279337526596814e-05,
|
| 360 |
+
"loss": 0.1292,
|
| 361 |
+
"num_input_tokens_seen": 2456552,
|
| 362 |
+
"step": 440
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.3453237410071943,
|
| 366 |
+
"grad_norm": 0.3129786550998688,
|
| 367 |
+
"learning_rate": 4.924629942372748e-05,
|
| 368 |
+
"loss": 0.1329,
|
| 369 |
+
"num_input_tokens_seen": 2512808,
|
| 370 |
+
"step": 450
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 2.397645519947678,
|
| 374 |
+
"grad_norm": 0.19762465357780457,
|
| 375 |
+
"learning_rate": 4.9212532476363596e-05,
|
| 376 |
+
"loss": 0.1261,
|
| 377 |
+
"num_input_tokens_seen": 2569016,
|
| 378 |
+
"step": 460
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 2.4499672988881622,
|
| 382 |
+
"grad_norm": 0.30003225803375244,
|
| 383 |
+
"learning_rate": 4.917803769954062e-05,
|
| 384 |
+
"loss": 0.124,
|
| 385 |
+
"num_input_tokens_seen": 2625208,
|
| 386 |
+
"step": 470
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.502289077828646,
|
| 390 |
+
"grad_norm": 0.2731872797012329,
|
| 391 |
+
"learning_rate": 4.9142816130172596e-05,
|
| 392 |
+
"loss": 0.1285,
|
| 393 |
+
"num_input_tokens_seen": 2680824,
|
| 394 |
+
"step": 480
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.55461085676913,
|
| 398 |
+
"grad_norm": 0.25849199295043945,
|
| 399 |
+
"learning_rate": 4.9106868827020955e-05,
|
| 400 |
+
"loss": 0.1323,
|
| 401 |
+
"num_input_tokens_seen": 2737304,
|
| 402 |
+
"step": 490
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 2.6069326357096143,
|
| 406 |
+
"grad_norm": 0.24799242615699768,
|
| 407 |
+
"learning_rate": 4.907019687066271e-05,
|
| 408 |
+
"loss": 0.1299,
|
| 409 |
+
"num_input_tokens_seen": 2793432,
|
| 410 |
+
"step": 500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.659254414650098,
|
| 414 |
+
"grad_norm": 0.1981252282857895,
|
| 415 |
+
"learning_rate": 4.9032801363458e-05,
|
| 416 |
+
"loss": 0.1281,
|
| 417 |
+
"num_input_tokens_seen": 2850008,
|
| 418 |
+
"step": 510
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 2.711576193590582,
|
| 422 |
+
"grad_norm": 0.40522608160972595,
|
| 423 |
+
"learning_rate": 4.8994683429516896e-05,
|
| 424 |
+
"loss": 0.1309,
|
| 425 |
+
"num_input_tokens_seen": 2905304,
|
| 426 |
+
"step": 520
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.763897972531066,
|
| 430 |
+
"grad_norm": 0.2624233067035675,
|
| 431 |
+
"learning_rate": 4.895584421466565e-05,
|
| 432 |
+
"loss": 0.1271,
|
| 433 |
+
"num_input_tokens_seen": 2961112,
|
| 434 |
+
"step": 530
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 2.81621975147155,
|
| 438 |
+
"grad_norm": 0.28190159797668457,
|
| 439 |
+
"learning_rate": 4.8916284886412214e-05,
|
| 440 |
+
"loss": 0.1222,
|
| 441 |
+
"num_input_tokens_seen": 3017208,
|
| 442 |
+
"step": 540
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 2.868541530412034,
|
| 446 |
+
"grad_norm": 0.21756145358085632,
|
| 447 |
+
"learning_rate": 4.887600663391122e-05,
|
| 448 |
+
"loss": 0.1288,
|
| 449 |
+
"num_input_tokens_seen": 3074216,
|
| 450 |
+
"step": 550
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.920863309352518,
|
| 454 |
+
"grad_norm": 0.33730047941207886,
|
| 455 |
+
"learning_rate": 4.883501066792814e-05,
|
| 456 |
+
"loss": 0.1267,
|
| 457 |
+
"num_input_tokens_seen": 3129784,
|
| 458 |
+
"step": 560
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.973185088293002,
|
| 462 |
+
"grad_norm": 0.24758093059062958,
|
| 463 |
+
"learning_rate": 4.8793298220802963e-05,
|
| 464 |
+
"loss": 0.1288,
|
| 465 |
+
"num_input_tokens_seen": 3187048,
|
| 466 |
+
"step": 570
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 3.0209287115761936,
|
| 470 |
+
"grad_norm": 0.15806905925273895,
|
| 471 |
+
"learning_rate": 4.87508705464131e-05,
|
| 472 |
+
"loss": 0.1316,
|
| 473 |
+
"num_input_tokens_seen": 3238608,
|
| 474 |
+
"step": 580
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 3.0732504905166778,
|
| 478 |
+
"grad_norm": 0.28546327352523804,
|
| 479 |
+
"learning_rate": 4.8707728920135744e-05,
|
| 480 |
+
"loss": 0.1256,
|
| 481 |
+
"num_input_tokens_seen": 3294400,
|
| 482 |
+
"step": 590
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 3.1255722694571615,
|
| 486 |
+
"grad_norm": 0.3583323359489441,
|
| 487 |
+
"learning_rate": 4.866387463880947e-05,
|
| 488 |
+
"loss": 0.1223,
|
| 489 |
+
"num_input_tokens_seen": 3350560,
|
| 490 |
+
"step": 600
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 3.1778940483976457,
|
| 494 |
+
"grad_norm": 0.21477927267551422,
|
| 495 |
+
"learning_rate": 4.861930902069531e-05,
|
| 496 |
+
"loss": 0.1192,
|
| 497 |
+
"num_input_tokens_seen": 3406256,
|
| 498 |
+
"step": 610
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.2302158273381294,
|
| 502 |
+
"grad_norm": 0.3112389147281647,
|
| 503 |
+
"learning_rate": 4.8574033405437094e-05,
|
| 504 |
+
"loss": 0.1209,
|
| 505 |
+
"num_input_tokens_seen": 3461680,
|
| 506 |
+
"step": 620
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.2825376062786136,
|
| 510 |
+
"grad_norm": 0.22634848952293396,
|
| 511 |
+
"learning_rate": 4.8528049154021186e-05,
|
| 512 |
+
"loss": 0.1318,
|
| 513 |
+
"num_input_tokens_seen": 3517984,
|
| 514 |
+
"step": 630
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 3.3348593852190973,
|
| 518 |
+
"grad_norm": 0.2349083423614502,
|
| 519 |
+
"learning_rate": 4.848135764873557e-05,
|
| 520 |
+
"loss": 0.1264,
|
| 521 |
+
"num_input_tokens_seen": 3573376,
|
| 522 |
+
"step": 640
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 3.3871811641595815,
|
| 526 |
+
"grad_norm": 0.29488080739974976,
|
| 527 |
+
"learning_rate": 4.843396029312832e-05,
|
| 528 |
+
"loss": 0.1238,
|
| 529 |
+
"num_input_tokens_seen": 3630544,
|
| 530 |
+
"step": 650
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.439502943100065,
|
| 534 |
+
"grad_norm": 0.28836989402770996,
|
| 535 |
+
"learning_rate": 4.838585851196537e-05,
|
| 536 |
+
"loss": 0.124,
|
| 537 |
+
"num_input_tokens_seen": 3686432,
|
| 538 |
+
"step": 660
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 3.4918247220405494,
|
| 542 |
+
"grad_norm": 0.2634967267513275,
|
| 543 |
+
"learning_rate": 4.833705375118772e-05,
|
| 544 |
+
"loss": 0.1212,
|
| 545 |
+
"num_input_tokens_seen": 3741552,
|
| 546 |
+
"step": 670
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 3.544146500981033,
|
| 550 |
+
"grad_norm": 0.30187705159187317,
|
| 551 |
+
"learning_rate": 4.828754747786796e-05,
|
| 552 |
+
"loss": 0.1225,
|
| 553 |
+
"num_input_tokens_seen": 3797760,
|
| 554 |
+
"step": 680
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 3.5964682799215173,
|
| 558 |
+
"grad_norm": 0.31589755415916443,
|
| 559 |
+
"learning_rate": 4.823734118016616e-05,
|
| 560 |
+
"loss": 0.1236,
|
| 561 |
+
"num_input_tokens_seen": 3854704,
|
| 562 |
+
"step": 690
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 3.6487900588620015,
|
| 566 |
+
"grad_norm": 0.24722936749458313,
|
| 567 |
+
"learning_rate": 4.818643636728515e-05,
|
| 568 |
+
"loss": 0.1154,
|
| 569 |
+
"num_input_tokens_seen": 3910544,
|
| 570 |
+
"step": 700
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.701111837802485,
|
| 574 |
+
"grad_norm": 0.20096950232982635,
|
| 575 |
+
"learning_rate": 4.813483456942515e-05,
|
| 576 |
+
"loss": 0.119,
|
| 577 |
+
"num_input_tokens_seen": 3966448,
|
| 578 |
+
"step": 710
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 3.7534336167429694,
|
| 582 |
+
"grad_norm": 0.22696824371814728,
|
| 583 |
+
"learning_rate": 4.808253733773775e-05,
|
| 584 |
+
"loss": 0.1277,
|
| 585 |
+
"num_input_tokens_seen": 4022880,
|
| 586 |
+
"step": 720
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 3.805755395683453,
|
| 590 |
+
"grad_norm": 0.2600644826889038,
|
| 591 |
+
"learning_rate": 4.8029546244279346e-05,
|
| 592 |
+
"loss": 0.1245,
|
| 593 |
+
"num_input_tokens_seen": 4079280,
|
| 594 |
+
"step": 730
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 3.8580771746239373,
|
| 598 |
+
"grad_norm": 0.3461725115776062,
|
| 599 |
+
"learning_rate": 4.797586288196378e-05,
|
| 600 |
+
"loss": 0.1245,
|
| 601 |
+
"num_input_tokens_seen": 4136096,
|
| 602 |
+
"step": 740
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.910398953564421,
|
| 606 |
+
"grad_norm": 0.20807071030139923,
|
| 607 |
+
"learning_rate": 4.792148886451456e-05,
|
| 608 |
+
"loss": 0.1238,
|
| 609 |
+
"num_input_tokens_seen": 4192832,
|
| 610 |
+
"step": 750
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 3.962720732504905,
|
| 614 |
+
"grad_norm": 0.25355812907218933,
|
| 615 |
+
"learning_rate": 4.7866425826416316e-05,
|
| 616 |
+
"loss": 0.1249,
|
| 617 |
+
"num_input_tokens_seen": 4248368,
|
| 618 |
+
"step": 760
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 4.010464355788097,
|
| 622 |
+
"grad_norm": 0.28272193670272827,
|
| 623 |
+
"learning_rate": 4.781067542286561e-05,
|
| 624 |
+
"loss": 0.1245,
|
| 625 |
+
"num_input_tokens_seen": 4299232,
|
| 626 |
+
"step": 770
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.062786134728581,
|
| 630 |
+
"grad_norm": 0.39174169301986694,
|
| 631 |
+
"learning_rate": 4.7754239329721274e-05,
|
| 632 |
+
"loss": 0.1216,
|
| 633 |
+
"num_input_tokens_seen": 4356192,
|
| 634 |
+
"step": 780
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 4.115107913669065,
|
| 638 |
+
"grad_norm": 0.2807229459285736,
|
| 639 |
+
"learning_rate": 4.769711924345397e-05,
|
| 640 |
+
"loss": 0.1195,
|
| 641 |
+
"num_input_tokens_seen": 4411648,
|
| 642 |
+
"step": 790
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 4.167429692609549,
|
| 646 |
+
"grad_norm": 0.24898973107337952,
|
| 647 |
+
"learning_rate": 4.763931688109524e-05,
|
| 648 |
+
"loss": 0.1174,
|
| 649 |
+
"num_input_tokens_seen": 4467568,
|
| 650 |
+
"step": 800
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 4.219751471550032,
|
| 654 |
+
"grad_norm": 0.3005022704601288,
|
| 655 |
+
"learning_rate": 4.7580833980185816e-05,
|
| 656 |
+
"loss": 0.1251,
|
| 657 |
+
"num_input_tokens_seen": 4522624,
|
| 658 |
+
"step": 810
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 4.272073250490517,
|
| 662 |
+
"grad_norm": 0.3487663269042969,
|
| 663 |
+
"learning_rate": 4.7521672298723495e-05,
|
| 664 |
+
"loss": 0.1182,
|
| 665 |
+
"num_input_tokens_seen": 4578640,
|
| 666 |
+
"step": 820
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 4.324395029431001,
|
| 670 |
+
"grad_norm": 0.19975335896015167,
|
| 671 |
+
"learning_rate": 4.7461833615110194e-05,
|
| 672 |
+
"loss": 0.1211,
|
| 673 |
+
"num_input_tokens_seen": 4633712,
|
| 674 |
+
"step": 830
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 4.3767168083714845,
|
| 678 |
+
"grad_norm": 0.2742094397544861,
|
| 679 |
+
"learning_rate": 4.740131972809856e-05,
|
| 680 |
+
"loss": 0.1208,
|
| 681 |
+
"num_input_tokens_seen": 4690160,
|
| 682 |
+
"step": 840
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 4.429038587311968,
|
| 686 |
+
"grad_norm": 0.49232161045074463,
|
| 687 |
+
"learning_rate": 4.734013245673788e-05,
|
| 688 |
+
"loss": 0.1213,
|
| 689 |
+
"num_input_tokens_seen": 4747104,
|
| 690 |
+
"step": 850
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 4.481360366252453,
|
| 694 |
+
"grad_norm": 0.3197735548019409,
|
| 695 |
+
"learning_rate": 4.727827364031936e-05,
|
| 696 |
+
"loss": 0.1137,
|
| 697 |
+
"num_input_tokens_seen": 4803376,
|
| 698 |
+
"step": 860
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.5336821451929366,
|
| 702 |
+
"grad_norm": 0.43819674849510193,
|
| 703 |
+
"learning_rate": 4.721574513832091e-05,
|
| 704 |
+
"loss": 0.1163,
|
| 705 |
+
"num_input_tokens_seen": 4859840,
|
| 706 |
+
"step": 870
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 4.58600392413342,
|
| 710 |
+
"grad_norm": 0.21390356123447418,
|
| 711 |
+
"learning_rate": 4.715254883035119e-05,
|
| 712 |
+
"loss": 0.121,
|
| 713 |
+
"num_input_tokens_seen": 4916272,
|
| 714 |
+
"step": 880
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 4.638325703073905,
|
| 718 |
+
"grad_norm": 0.30768075585365295,
|
| 719 |
+
"learning_rate": 4.708868661609314e-05,
|
| 720 |
+
"loss": 0.1194,
|
| 721 |
+
"num_input_tokens_seen": 4971728,
|
| 722 |
+
"step": 890
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 4.690647482014389,
|
| 726 |
+
"grad_norm": 0.34879791736602783,
|
| 727 |
+
"learning_rate": 4.702416041524683e-05,
|
| 728 |
+
"loss": 0.1223,
|
| 729 |
+
"num_input_tokens_seen": 5027680,
|
| 730 |
+
"step": 900
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 4.742969260954872,
|
| 734 |
+
"grad_norm": 0.23973870277404785,
|
| 735 |
+
"learning_rate": 4.695897216747183e-05,
|
| 736 |
+
"loss": 0.1225,
|
| 737 |
+
"num_input_tokens_seen": 5083392,
|
| 738 |
+
"step": 910
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 4.795291039895356,
|
| 742 |
+
"grad_norm": 0.33110499382019043,
|
| 743 |
+
"learning_rate": 4.689312383232883e-05,
|
| 744 |
+
"loss": 0.1248,
|
| 745 |
+
"num_input_tokens_seen": 5140112,
|
| 746 |
+
"step": 920
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 4.847612818835841,
|
| 750 |
+
"grad_norm": 0.2787615954875946,
|
| 751 |
+
"learning_rate": 4.682661738922078e-05,
|
| 752 |
+
"loss": 0.1204,
|
| 753 |
+
"num_input_tokens_seen": 5195072,
|
| 754 |
+
"step": 930
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 4.8999345977763245,
|
| 758 |
+
"grad_norm": 0.20854991674423218,
|
| 759 |
+
"learning_rate": 4.6759454837333376e-05,
|
| 760 |
+
"loss": 0.1181,
|
| 761 |
+
"num_input_tokens_seen": 5251408,
|
| 762 |
+
"step": 940
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 4.952256376716808,
|
| 766 |
+
"grad_norm": 0.3582795262336731,
|
| 767 |
+
"learning_rate": 4.6691638195574963e-05,
|
| 768 |
+
"loss": 0.118,
|
| 769 |
+
"num_input_tokens_seen": 5307776,
|
| 770 |
+
"step": 950
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 5.0,
|
| 774 |
+
"grad_norm": 0.18520788848400116,
|
| 775 |
+
"learning_rate": 4.662316950251584e-05,
|
| 776 |
+
"loss": 0.1192,
|
| 777 |
+
"num_input_tokens_seen": 5359280,
|
| 778 |
+
"step": 960
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 5.052321778940484,
|
| 782 |
+
"grad_norm": 0.2574616074562073,
|
| 783 |
+
"learning_rate": 4.655405081632699e-05,
|
| 784 |
+
"loss": 0.1179,
|
| 785 |
+
"num_input_tokens_seen": 5415968,
|
| 786 |
+
"step": 970
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 5.104643557880968,
|
| 790 |
+
"grad_norm": 0.3383731544017792,
|
| 791 |
+
"learning_rate": 4.648428421471822e-05,
|
| 792 |
+
"loss": 0.1137,
|
| 793 |
+
"num_input_tokens_seen": 5471840,
|
| 794 |
+
"step": 980
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 5.156965336821452,
|
| 798 |
+
"grad_norm": 0.3150075376033783,
|
| 799 |
+
"learning_rate": 4.641387179487569e-05,
|
| 800 |
+
"loss": 0.1179,
|
| 801 |
+
"num_input_tokens_seen": 5528224,
|
| 802 |
+
"step": 990
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 5.209287115761936,
|
| 806 |
+
"grad_norm": 0.2862718999385834,
|
| 807 |
+
"learning_rate": 4.634281567339885e-05,
|
| 808 |
+
"loss": 0.1117,
|
| 809 |
+
"num_input_tokens_seen": 5583680,
|
| 810 |
+
"step": 1000
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 5.2616088947024195,
|
| 814 |
+
"grad_norm": 0.32703572511672974,
|
| 815 |
+
"learning_rate": 4.627111798623688e-05,
|
| 816 |
+
"loss": 0.12,
|
| 817 |
+
"num_input_tokens_seen": 5638640,
|
| 818 |
+
"step": 1010
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 5.313930673642904,
|
| 822 |
+
"grad_norm": 0.3088448941707611,
|
| 823 |
+
"learning_rate": 4.619878088862443e-05,
|
| 824 |
+
"loss": 0.1134,
|
| 825 |
+
"num_input_tokens_seen": 5694208,
|
| 826 |
+
"step": 1020
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 5.366252452583388,
|
| 830 |
+
"grad_norm": 0.3116937279701233,
|
| 831 |
+
"learning_rate": 4.612580655501683e-05,
|
| 832 |
+
"loss": 0.1178,
|
| 833 |
+
"num_input_tokens_seen": 5749696,
|
| 834 |
+
"step": 1030
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 5.418574231523872,
|
| 838 |
+
"grad_norm": 0.3191389739513397,
|
| 839 |
+
"learning_rate": 4.605219717902476e-05,
|
| 840 |
+
"loss": 0.1136,
|
| 841 |
+
"num_input_tokens_seen": 5805264,
|
| 842 |
+
"step": 1040
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 5.470896010464356,
|
| 846 |
+
"grad_norm": 0.31546375155448914,
|
| 847 |
+
"learning_rate": 4.5977954973348294e-05,
|
| 848 |
+
"loss": 0.1167,
|
| 849 |
+
"num_input_tokens_seen": 5861616,
|
| 850 |
+
"step": 1050
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 5.52321778940484,
|
| 854 |
+
"grad_norm": 0.2683027386665344,
|
| 855 |
+
"learning_rate": 4.590308216971038e-05,
|
| 856 |
+
"loss": 0.1113,
|
| 857 |
+
"num_input_tokens_seen": 5918816,
|
| 858 |
+
"step": 1060
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 5.575539568345324,
|
| 862 |
+
"grad_norm": 0.26271599531173706,
|
| 863 |
+
"learning_rate": 4.582758101878977e-05,
|
| 864 |
+
"loss": 0.1113,
|
| 865 |
+
"num_input_tokens_seen": 5975184,
|
| 866 |
+
"step": 1070
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 5.6278613472858074,
|
| 870 |
+
"grad_norm": 0.2612985372543335,
|
| 871 |
+
"learning_rate": 4.5751453790153325e-05,
|
| 872 |
+
"loss": 0.1143,
|
| 873 |
+
"num_input_tokens_seen": 6030736,
|
| 874 |
+
"step": 1080
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 5.680183126226292,
|
| 878 |
+
"grad_norm": 0.29255977272987366,
|
| 879 |
+
"learning_rate": 4.567470277218786e-05,
|
| 880 |
+
"loss": 0.1159,
|
| 881 |
+
"num_input_tokens_seen": 6086848,
|
| 882 |
+
"step": 1090
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 5.732504905166776,
|
| 886 |
+
"grad_norm": 0.31331729888916016,
|
| 887 |
+
"learning_rate": 4.55973302720313e-05,
|
| 888 |
+
"loss": 0.1114,
|
| 889 |
+
"num_input_tokens_seen": 6142544,
|
| 890 |
+
"step": 1100
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 5.7848266841072595,
|
| 894 |
+
"grad_norm": 0.24986866116523743,
|
| 895 |
+
"learning_rate": 4.551933861550333e-05,
|
| 896 |
+
"loss": 0.1173,
|
| 897 |
+
"num_input_tokens_seen": 6199856,
|
| 898 |
+
"step": 1110
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 5.837148463047743,
|
| 902 |
+
"grad_norm": 0.26383090019226074,
|
| 903 |
+
"learning_rate": 4.5440730147035516e-05,
|
| 904 |
+
"loss": 0.1166,
|
| 905 |
+
"num_input_tokens_seen": 6255488,
|
| 906 |
+
"step": 1120
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 5.889470241988228,
|
| 910 |
+
"grad_norm": 0.28010377287864685,
|
| 911 |
+
"learning_rate": 4.5361507229600784e-05,
|
| 912 |
+
"loss": 0.1148,
|
| 913 |
+
"num_input_tokens_seen": 6311696,
|
| 914 |
+
"step": 1130
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 5.941792020928712,
|
| 918 |
+
"grad_norm": 0.2781098783016205,
|
| 919 |
+
"learning_rate": 4.528167224464245e-05,
|
| 920 |
+
"loss": 0.1152,
|
| 921 |
+
"num_input_tokens_seen": 6368064,
|
| 922 |
+
"step": 1140
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 5.994113799869195,
|
| 926 |
+
"grad_norm": 0.35192057490348816,
|
| 927 |
+
"learning_rate": 4.520122759200256e-05,
|
| 928 |
+
"loss": 0.1087,
|
| 929 |
+
"num_input_tokens_seen": 6424048,
|
| 930 |
+
"step": 1150
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 6.041857423152387,
|
| 934 |
+
"grad_norm": 0.408779114484787,
|
| 935 |
+
"learning_rate": 4.512017568984982e-05,
|
| 936 |
+
"loss": 0.1094,
|
| 937 |
+
"num_input_tokens_seen": 6475464,
|
| 938 |
+
"step": 1160
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 6.094179202092871,
|
| 942 |
+
"grad_norm": 0.5394303798675537,
|
| 943 |
+
"learning_rate": 4.503851897460686e-05,
|
| 944 |
+
"loss": 0.1005,
|
| 945 |
+
"num_input_tokens_seen": 6531112,
|
| 946 |
+
"step": 1170
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 6.1465009810333555,
|
| 950 |
+
"grad_norm": 0.4771885871887207,
|
| 951 |
+
"learning_rate": 4.4956259900877005e-05,
|
| 952 |
+
"loss": 0.107,
|
| 953 |
+
"num_input_tokens_seen": 6587352,
|
| 954 |
+
"step": 1180
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 6.198822759973839,
|
| 958 |
+
"grad_norm": 0.4851493835449219,
|
| 959 |
+
"learning_rate": 4.4873400941370506e-05,
|
| 960 |
+
"loss": 0.1093,
|
| 961 |
+
"num_input_tokens_seen": 6643608,
|
| 962 |
+
"step": 1190
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 6.251144538914323,
|
| 966 |
+
"grad_norm": 0.47404423356056213,
|
| 967 |
+
"learning_rate": 4.4789944586830196e-05,
|
| 968 |
+
"loss": 0.1082,
|
| 969 |
+
"num_input_tokens_seen": 6700616,
|
| 970 |
+
"step": 1200
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 6.303466317854807,
|
| 974 |
+
"grad_norm": 0.4265010356903076,
|
| 975 |
+
"learning_rate": 4.470589334595662e-05,
|
| 976 |
+
"loss": 0.1088,
|
| 977 |
+
"num_input_tokens_seen": 6756344,
|
| 978 |
+
"step": 1210
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 6.355788096795291,
|
| 982 |
+
"grad_norm": 0.2776981592178345,
|
| 983 |
+
"learning_rate": 4.462124974533261e-05,
|
| 984 |
+
"loss": 0.1124,
|
| 985 |
+
"num_input_tokens_seen": 6813144,
|
| 986 |
+
"step": 1220
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 6.408109875735775,
|
| 990 |
+
"grad_norm": 0.3706200122833252,
|
| 991 |
+
"learning_rate": 4.453601632934737e-05,
|
| 992 |
+
"loss": 0.1095,
|
| 993 |
+
"num_input_tokens_seen": 6868280,
|
| 994 |
+
"step": 1230
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 6.460431654676259,
|
| 998 |
+
"grad_norm": 0.4983665943145752,
|
| 999 |
+
"learning_rate": 4.4450195660119965e-05,
|
| 1000 |
+
"loss": 0.1114,
|
| 1001 |
+
"num_input_tokens_seen": 6924296,
|
| 1002 |
+
"step": 1240
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 6.5127534336167425,
|
| 1006 |
+
"grad_norm": 0.38131120800971985,
|
| 1007 |
+
"learning_rate": 4.4363790317422314e-05,
|
| 1008 |
+
"loss": 0.1141,
|
| 1009 |
+
"num_input_tokens_seen": 6980392,
|
| 1010 |
+
"step": 1250
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 6.565075212557227,
|
| 1014 |
+
"grad_norm": 0.24885649979114532,
|
| 1015 |
+
"learning_rate": 4.427680289860163e-05,
|
| 1016 |
+
"loss": 0.1128,
|
| 1017 |
+
"num_input_tokens_seen": 7036056,
|
| 1018 |
+
"step": 1260
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 6.617396991497711,
|
| 1022 |
+
"grad_norm": 0.4561365246772766,
|
| 1023 |
+
"learning_rate": 4.4189236018502356e-05,
|
| 1024 |
+
"loss": 0.1149,
|
| 1025 |
+
"num_input_tokens_seen": 7092152,
|
| 1026 |
+
"step": 1270
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 6.669718770438195,
|
| 1030 |
+
"grad_norm": 0.548459529876709,
|
| 1031 |
+
"learning_rate": 4.410109230938755e-05,
|
| 1032 |
+
"loss": 0.1079,
|
| 1033 |
+
"num_input_tokens_seen": 7147096,
|
| 1034 |
+
"step": 1280
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 6.722040549378679,
|
| 1038 |
+
"grad_norm": 0.46682003140449524,
|
| 1039 |
+
"learning_rate": 4.4012374420859786e-05,
|
| 1040 |
+
"loss": 0.1061,
|
| 1041 |
+
"num_input_tokens_seen": 7202584,
|
| 1042 |
+
"step": 1290
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 6.774362328319163,
|
| 1046 |
+
"grad_norm": 0.32285481691360474,
|
| 1047 |
+
"learning_rate": 4.392308501978148e-05,
|
| 1048 |
+
"loss": 0.1098,
|
| 1049 |
+
"num_input_tokens_seen": 7258552,
|
| 1050 |
+
"step": 1300
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 6.826684107259647,
|
| 1054 |
+
"grad_norm": 0.3166508078575134,
|
| 1055 |
+
"learning_rate": 4.383322679019472e-05,
|
| 1056 |
+
"loss": 0.1119,
|
| 1057 |
+
"num_input_tokens_seen": 7315768,
|
| 1058 |
+
"step": 1310
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 6.87900588620013,
|
| 1062 |
+
"grad_norm": 0.35454803705215454,
|
| 1063 |
+
"learning_rate": 4.3742802433240625e-05,
|
| 1064 |
+
"loss": 0.1107,
|
| 1065 |
+
"num_input_tokens_seen": 7371352,
|
| 1066 |
+
"step": 1320
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 6.931327665140615,
|
| 1070 |
+
"grad_norm": 0.2988300919532776,
|
| 1071 |
+
"learning_rate": 4.3651814667078086e-05,
|
| 1072 |
+
"loss": 0.1085,
|
| 1073 |
+
"num_input_tokens_seen": 7427800,
|
| 1074 |
+
"step": 1330
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 6.983649444081099,
|
| 1078 |
+
"grad_norm": 0.3938722014427185,
|
| 1079 |
+
"learning_rate": 4.35602662268021e-05,
|
| 1080 |
+
"loss": 0.1064,
|
| 1081 |
+
"num_input_tokens_seen": 7484616,
|
| 1082 |
+
"step": 1340
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 7.031393067364291,
|
| 1086 |
+
"grad_norm": 0.45549729466438293,
|
| 1087 |
+
"learning_rate": 4.346815986436158e-05,
|
| 1088 |
+
"loss": 0.1041,
|
| 1089 |
+
"num_input_tokens_seen": 7534896,
|
| 1090 |
+
"step": 1350
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 7.083714846304774,
|
| 1094 |
+
"grad_norm": 0.6130596995353699,
|
| 1095 |
+
"learning_rate": 4.337549834847655e-05,
|
| 1096 |
+
"loss": 0.0977,
|
| 1097 |
+
"num_input_tokens_seen": 7591328,
|
| 1098 |
+
"step": 1360
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 7.136036625245258,
|
| 1102 |
+
"grad_norm": 0.6437245607376099,
|
| 1103 |
+
"learning_rate": 4.328228446455498e-05,
|
| 1104 |
+
"loss": 0.0979,
|
| 1105 |
+
"num_input_tokens_seen": 7647760,
|
| 1106 |
+
"step": 1370
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 7.188358404185743,
|
| 1110 |
+
"grad_norm": 0.5056234002113342,
|
| 1111 |
+
"learning_rate": 4.3188521014609054e-05,
|
| 1112 |
+
"loss": 0.0994,
|
| 1113 |
+
"num_input_tokens_seen": 7704672,
|
| 1114 |
+
"step": 1380
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 7.240680183126226,
|
| 1118 |
+
"grad_norm": 0.5103983879089355,
|
| 1119 |
+
"learning_rate": 4.309421081717091e-05,
|
| 1120 |
+
"loss": 0.1001,
|
| 1121 |
+
"num_input_tokens_seen": 7761104,
|
| 1122 |
+
"step": 1390
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 7.29300196206671,
|
| 1126 |
+
"grad_norm": 0.47358736395835876,
|
| 1127 |
+
"learning_rate": 4.299935670720794e-05,
|
| 1128 |
+
"loss": 0.0926,
|
| 1129 |
+
"num_input_tokens_seen": 7818224,
|
| 1130 |
+
"step": 1400
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 7.345323741007194,
|
| 1134 |
+
"grad_norm": 0.6868699789047241,
|
| 1135 |
+
"learning_rate": 4.290396153603755e-05,
|
| 1136 |
+
"loss": 0.0975,
|
| 1137 |
+
"num_input_tokens_seen": 7874208,
|
| 1138 |
+
"step": 1410
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 7.3976455199476785,
|
| 1142 |
+
"grad_norm": 0.39971715211868286,
|
| 1143 |
+
"learning_rate": 4.280802817124149e-05,
|
| 1144 |
+
"loss": 0.1012,
|
| 1145 |
+
"num_input_tokens_seen": 7930496,
|
| 1146 |
+
"step": 1420
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 7.449967298888162,
|
| 1150 |
+
"grad_norm": 0.6476240754127502,
|
| 1151 |
+
"learning_rate": 4.271155949657959e-05,
|
| 1152 |
+
"loss": 0.1033,
|
| 1153 |
+
"num_input_tokens_seen": 7985552,
|
| 1154 |
+
"step": 1430
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 7.502289077828646,
|
| 1158 |
+
"grad_norm": 0.5134351253509521,
|
| 1159 |
+
"learning_rate": 4.261455841190314e-05,
|
| 1160 |
+
"loss": 0.0966,
|
| 1161 |
+
"num_input_tokens_seen": 8041568,
|
| 1162 |
+
"step": 1440
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 7.554610856769131,
|
| 1166 |
+
"grad_norm": 0.6232761144638062,
|
| 1167 |
+
"learning_rate": 4.2517027833067685e-05,
|
| 1168 |
+
"loss": 0.1001,
|
| 1169 |
+
"num_input_tokens_seen": 8098656,
|
| 1170 |
+
"step": 1450
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 7.606932635709614,
|
| 1174 |
+
"grad_norm": 0.4906207025051117,
|
| 1175 |
+
"learning_rate": 4.241897069184537e-05,
|
| 1176 |
+
"loss": 0.1072,
|
| 1177 |
+
"num_input_tokens_seen": 8154032,
|
| 1178 |
+
"step": 1460
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 7.659254414650098,
|
| 1182 |
+
"grad_norm": 0.6239527463912964,
|
| 1183 |
+
"learning_rate": 4.2320389935836836e-05,
|
| 1184 |
+
"loss": 0.1006,
|
| 1185 |
+
"num_input_tokens_seen": 8210032,
|
| 1186 |
+
"step": 1470
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 7.711576193590582,
|
| 1190 |
+
"grad_norm": 1.2038999795913696,
|
| 1191 |
+
"learning_rate": 4.2221288528382584e-05,
|
| 1192 |
+
"loss": 0.1015,
|
| 1193 |
+
"num_input_tokens_seen": 8265296,
|
| 1194 |
+
"step": 1480
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 7.763897972531066,
|
| 1198 |
+
"grad_norm": 0.365824431180954,
|
| 1199 |
+
"learning_rate": 4.212166944847392e-05,
|
| 1200 |
+
"loss": 0.0973,
|
| 1201 |
+
"num_input_tokens_seen": 8321840,
|
| 1202 |
+
"step": 1490
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 7.81621975147155,
|
| 1206 |
+
"grad_norm": 0.4954347610473633,
|
| 1207 |
+
"learning_rate": 4.2021535690663414e-05,
|
| 1208 |
+
"loss": 0.1015,
|
| 1209 |
+
"num_input_tokens_seen": 8378064,
|
| 1210 |
+
"step": 1500
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 7.868541530412034,
|
| 1214 |
+
"grad_norm": 0.5742694139480591,
|
| 1215 |
+
"learning_rate": 4.192089026497484e-05,
|
| 1216 |
+
"loss": 0.0997,
|
| 1217 |
+
"num_input_tokens_seen": 8434384,
|
| 1218 |
+
"step": 1510
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 7.920863309352518,
|
| 1222 |
+
"grad_norm": 0.48765209317207336,
|
| 1223 |
+
"learning_rate": 4.181973619681276e-05,
|
| 1224 |
+
"loss": 0.1003,
|
| 1225 |
+
"num_input_tokens_seen": 8490496,
|
| 1226 |
+
"step": 1520
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 7.973185088293002,
|
| 1230 |
+
"grad_norm": 0.4735221266746521,
|
| 1231 |
+
"learning_rate": 4.171807652687151e-05,
|
| 1232 |
+
"loss": 0.1051,
|
| 1233 |
+
"num_input_tokens_seen": 8545600,
|
| 1234 |
+
"step": 1530
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 8.020928711576193,
|
| 1238 |
+
"grad_norm": 0.35401248931884766,
|
| 1239 |
+
"learning_rate": 4.1615914311043855e-05,
|
| 1240 |
+
"loss": 0.0967,
|
| 1241 |
+
"num_input_tokens_seen": 8595848,
|
| 1242 |
+
"step": 1540
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 8.073250490516678,
|
| 1246 |
+
"grad_norm": 0.7497507929801941,
|
| 1247 |
+
"learning_rate": 4.151325262032908e-05,
|
| 1248 |
+
"loss": 0.0834,
|
| 1249 |
+
"num_input_tokens_seen": 8652536,
|
| 1250 |
+
"step": 1550
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 8.125572269457162,
|
| 1254 |
+
"grad_norm": 0.7235395908355713,
|
| 1255 |
+
"learning_rate": 4.1410094540740726e-05,
|
| 1256 |
+
"loss": 0.0804,
|
| 1257 |
+
"num_input_tokens_seen": 8708952,
|
| 1258 |
+
"step": 1560
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 8.177894048397645,
|
| 1262 |
+
"grad_norm": 0.9055391550064087,
|
| 1263 |
+
"learning_rate": 4.1306443173213785e-05,
|
| 1264 |
+
"loss": 0.085,
|
| 1265 |
+
"num_input_tokens_seen": 8765688,
|
| 1266 |
+
"step": 1570
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 8.23021582733813,
|
| 1270 |
+
"grad_norm": 0.8378714919090271,
|
| 1271 |
+
"learning_rate": 4.1202301633511506e-05,
|
| 1272 |
+
"loss": 0.0813,
|
| 1273 |
+
"num_input_tokens_seen": 8822376,
|
| 1274 |
+
"step": 1580
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 8.282537606278613,
|
| 1278 |
+
"grad_norm": 0.6385327577590942,
|
| 1279 |
+
"learning_rate": 4.109767305213173e-05,
|
| 1280 |
+
"loss": 0.0831,
|
| 1281 |
+
"num_input_tokens_seen": 8878456,
|
| 1282 |
+
"step": 1590
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 8.334859385219097,
|
| 1286 |
+
"grad_norm": 0.8356183767318726,
|
| 1287 |
+
"learning_rate": 4.0992560574212764e-05,
|
| 1288 |
+
"loss": 0.0893,
|
| 1289 |
+
"num_input_tokens_seen": 8934088,
|
| 1290 |
+
"step": 1600
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 8.387181164159582,
|
| 1294 |
+
"grad_norm": 0.8791753649711609,
|
| 1295 |
+
"learning_rate": 4.0886967359438885e-05,
|
| 1296 |
+
"loss": 0.087,
|
| 1297 |
+
"num_input_tokens_seen": 8990120,
|
| 1298 |
+
"step": 1610
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 8.439502943100065,
|
| 1302 |
+
"grad_norm": 0.6865050792694092,
|
| 1303 |
+
"learning_rate": 4.078089658194533e-05,
|
| 1304 |
+
"loss": 0.0873,
|
| 1305 |
+
"num_input_tokens_seen": 9045848,
|
| 1306 |
+
"step": 1620
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 8.49182472204055,
|
| 1310 |
+
"grad_norm": 0.7075687050819397,
|
| 1311 |
+
"learning_rate": 4.0674351430222864e-05,
|
| 1312 |
+
"loss": 0.0855,
|
| 1313 |
+
"num_input_tokens_seen": 9102456,
|
| 1314 |
+
"step": 1630
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 8.544146500981034,
|
| 1318 |
+
"grad_norm": 0.7803745269775391,
|
| 1319 |
+
"learning_rate": 4.0567335107021986e-05,
|
| 1320 |
+
"loss": 0.0887,
|
| 1321 |
+
"num_input_tokens_seen": 9158712,
|
| 1322 |
+
"step": 1640
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 8.596468279921517,
|
| 1326 |
+
"grad_norm": 0.6807383894920349,
|
| 1327 |
+
"learning_rate": 4.0459850829256604e-05,
|
| 1328 |
+
"loss": 0.0826,
|
| 1329 |
+
"num_input_tokens_seen": 9215496,
|
| 1330 |
+
"step": 1650
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 8.648790058862001,
|
| 1334 |
+
"grad_norm": 0.777777373790741,
|
| 1335 |
+
"learning_rate": 4.035190182790738e-05,
|
| 1336 |
+
"loss": 0.0922,
|
| 1337 |
+
"num_input_tokens_seen": 9269720,
|
| 1338 |
+
"step": 1660
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 8.701111837802486,
|
| 1342 |
+
"grad_norm": 0.6249901056289673,
|
| 1343 |
+
"learning_rate": 4.024349134792453e-05,
|
| 1344 |
+
"loss": 0.0862,
|
| 1345 |
+
"num_input_tokens_seen": 9325624,
|
| 1346 |
+
"step": 1670
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 8.753433616742969,
|
| 1350 |
+
"grad_norm": 0.7085604667663574,
|
| 1351 |
+
"learning_rate": 4.0134622648130394e-05,
|
| 1352 |
+
"loss": 0.0836,
|
| 1353 |
+
"num_input_tokens_seen": 9380984,
|
| 1354 |
+
"step": 1680
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 8.805755395683454,
|
| 1358 |
+
"grad_norm": 0.9964501261711121,
|
| 1359 |
+
"learning_rate": 4.0025299001121365e-05,
|
| 1360 |
+
"loss": 0.0862,
|
| 1361 |
+
"num_input_tokens_seen": 9437320,
|
| 1362 |
+
"step": 1690
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 8.858077174623936,
|
| 1366 |
+
"grad_norm": 1.0215933322906494,
|
| 1367 |
+
"learning_rate": 3.991552369316958e-05,
|
| 1368 |
+
"loss": 0.0926,
|
| 1369 |
+
"num_input_tokens_seen": 9492888,
|
| 1370 |
+
"step": 1700
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 8.910398953564421,
|
| 1374 |
+
"grad_norm": 0.7546895146369934,
|
| 1375 |
+
"learning_rate": 3.9805300024124125e-05,
|
| 1376 |
+
"loss": 0.0893,
|
| 1377 |
+
"num_input_tokens_seen": 9550504,
|
| 1378 |
+
"step": 1710
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 8.962720732504906,
|
| 1382 |
+
"grad_norm": 0.8502705693244934,
|
| 1383 |
+
"learning_rate": 3.969463130731183e-05,
|
| 1384 |
+
"loss": 0.0948,
|
| 1385 |
+
"num_input_tokens_seen": 9606120,
|
| 1386 |
+
"step": 1720
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 9.010464355788097,
|
| 1390 |
+
"grad_norm": 0.5291112065315247,
|
| 1391 |
+
"learning_rate": 3.9583520869437666e-05,
|
| 1392 |
+
"loss": 0.0893,
|
| 1393 |
+
"num_input_tokens_seen": 9656472,
|
| 1394 |
+
"step": 1730
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 9.062786134728581,
|
| 1398 |
+
"grad_norm": 0.6509256958961487,
|
| 1399 |
+
"learning_rate": 3.9471972050484764e-05,
|
| 1400 |
+
"loss": 0.063,
|
| 1401 |
+
"num_input_tokens_seen": 9712712,
|
| 1402 |
+
"step": 1740
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 9.115107913669064,
|
| 1406 |
+
"grad_norm": 1.1916331052780151,
|
| 1407 |
+
"learning_rate": 3.9359988203614e-05,
|
| 1408 |
+
"loss": 0.0602,
|
| 1409 |
+
"num_input_tokens_seen": 9768584,
|
| 1410 |
+
"step": 1750
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 9.167429692609549,
|
| 1414 |
+
"grad_norm": 0.8077856302261353,
|
| 1415 |
+
"learning_rate": 3.924757269506319e-05,
|
| 1416 |
+
"loss": 0.0643,
|
| 1417 |
+
"num_input_tokens_seen": 9824360,
|
| 1418 |
+
"step": 1760
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 9.219751471550033,
|
| 1422 |
+
"grad_norm": 1.0554853677749634,
|
| 1423 |
+
"learning_rate": 3.913472890404593e-05,
|
| 1424 |
+
"loss": 0.0668,
|
| 1425 |
+
"num_input_tokens_seen": 9880328,
|
| 1426 |
+
"step": 1770
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 9.272073250490516,
|
| 1430 |
+
"grad_norm": 0.8842382431030273,
|
| 1431 |
+
"learning_rate": 3.9021460222649986e-05,
|
| 1432 |
+
"loss": 0.0633,
|
| 1433 |
+
"num_input_tokens_seen": 9936248,
|
| 1434 |
+
"step": 1780
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 9.324395029431,
|
| 1438 |
+
"grad_norm": 0.9562489986419678,
|
| 1439 |
+
"learning_rate": 3.890777005573537e-05,
|
| 1440 |
+
"loss": 0.0625,
|
| 1441 |
+
"num_input_tokens_seen": 9992088,
|
| 1442 |
+
"step": 1790
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 9.376716808371485,
|
| 1446 |
+
"grad_norm": 0.8844836950302124,
|
| 1447 |
+
"learning_rate": 3.8793661820831915e-05,
|
| 1448 |
+
"loss": 0.0718,
|
| 1449 |
+
"num_input_tokens_seen": 10048520,
|
| 1450 |
+
"step": 1800
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 9.429038587311968,
|
| 1454 |
+
"grad_norm": 1.2325676679611206,
|
| 1455 |
+
"learning_rate": 3.867913894803663e-05,
|
| 1456 |
+
"loss": 0.0646,
|
| 1457 |
+
"num_input_tokens_seen": 10104776,
|
| 1458 |
+
"step": 1810
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 9.481360366252453,
|
| 1462 |
+
"grad_norm": 0.7227656245231628,
|
| 1463 |
+
"learning_rate": 3.8564204879910535e-05,
|
| 1464 |
+
"loss": 0.0657,
|
| 1465 |
+
"num_input_tokens_seen": 10160888,
|
| 1466 |
+
"step": 1820
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 9.533682145192937,
|
| 1470 |
+
"grad_norm": 0.9736649990081787,
|
| 1471 |
+
"learning_rate": 3.844886307137519e-05,
|
| 1472 |
+
"loss": 0.0691,
|
| 1473 |
+
"num_input_tokens_seen": 10216920,
|
| 1474 |
+
"step": 1830
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 9.58600392413342,
|
| 1478 |
+
"grad_norm": 1.0427885055541992,
|
| 1479 |
+
"learning_rate": 3.833311698960888e-05,
|
| 1480 |
+
"loss": 0.0667,
|
| 1481 |
+
"num_input_tokens_seen": 10273880,
|
| 1482 |
+
"step": 1840
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 9.638325703073905,
|
| 1486 |
+
"grad_norm": 0.9749732613563538,
|
| 1487 |
+
"learning_rate": 3.8216970113942284e-05,
|
| 1488 |
+
"loss": 0.0699,
|
| 1489 |
+
"num_input_tokens_seen": 10329752,
|
| 1490 |
+
"step": 1850
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 9.690647482014388,
|
| 1494 |
+
"grad_norm": 0.6216735243797302,
|
| 1495 |
+
"learning_rate": 3.8100425935754025e-05,
|
| 1496 |
+
"loss": 0.072,
|
| 1497 |
+
"num_input_tokens_seen": 10387128,
|
| 1498 |
+
"step": 1860
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 9.742969260954872,
|
| 1502 |
+
"grad_norm": 0.8047035932540894,
|
| 1503 |
+
"learning_rate": 3.798348795836562e-05,
|
| 1504 |
+
"loss": 0.0697,
|
| 1505 |
+
"num_input_tokens_seen": 10443144,
|
| 1506 |
+
"step": 1870
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 9.795291039895357,
|
| 1510 |
+
"grad_norm": 0.793222188949585,
|
| 1511 |
+
"learning_rate": 3.786615969693621e-05,
|
| 1512 |
+
"loss": 0.0709,
|
| 1513 |
+
"num_input_tokens_seen": 10497784,
|
| 1514 |
+
"step": 1880
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 9.84761281883584,
|
| 1518 |
+
"grad_norm": 0.7294532656669617,
|
| 1519 |
+
"learning_rate": 3.7748444678356886e-05,
|
| 1520 |
+
"loss": 0.0716,
|
| 1521 |
+
"num_input_tokens_seen": 10553416,
|
| 1522 |
+
"step": 1890
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 9.899934597776324,
|
| 1526 |
+
"grad_norm": 1.014732003211975,
|
| 1527 |
+
"learning_rate": 3.7630346441144656e-05,
|
| 1528 |
+
"loss": 0.0695,
|
| 1529 |
+
"num_input_tokens_seen": 10610168,
|
| 1530 |
+
"step": 1900
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 9.952256376716809,
|
| 1534 |
+
"grad_norm": 1.22804856300354,
|
| 1535 |
+
"learning_rate": 3.7511868535336134e-05,
|
| 1536 |
+
"loss": 0.0731,
|
| 1537 |
+
"num_input_tokens_seen": 10665912,
|
| 1538 |
+
"step": 1910
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 10.0,
|
| 1542 |
+
"grad_norm": 0.7104299068450928,
|
| 1543 |
+
"learning_rate": 3.7393014522380734e-05,
|
| 1544 |
+
"loss": 0.0652,
|
| 1545 |
+
"num_input_tokens_seen": 10717424,
|
| 1546 |
+
"step": 1920
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 10.052321778940485,
|
| 1550 |
+
"grad_norm": 0.9654238224029541,
|
| 1551 |
+
"learning_rate": 3.7273787975033686e-05,
|
| 1552 |
+
"loss": 0.0435,
|
| 1553 |
+
"num_input_tokens_seen": 10772656,
|
| 1554 |
+
"step": 1930
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 10.104643557880967,
|
| 1558 |
+
"grad_norm": 1.2412711381912231,
|
| 1559 |
+
"learning_rate": 3.7154192477248614e-05,
|
| 1560 |
+
"loss": 0.0425,
|
| 1561 |
+
"num_input_tokens_seen": 10829200,
|
| 1562 |
+
"step": 1940
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 10.156965336821452,
|
| 1566 |
+
"grad_norm": 0.6960983276367188,
|
| 1567 |
+
"learning_rate": 3.7034231624069796e-05,
|
| 1568 |
+
"loss": 0.0376,
|
| 1569 |
+
"num_input_tokens_seen": 10884064,
|
| 1570 |
+
"step": 1950
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 10.209287115761937,
|
| 1574 |
+
"grad_norm": 1.214837908744812,
|
| 1575 |
+
"learning_rate": 3.691390902152412e-05,
|
| 1576 |
+
"loss": 0.0487,
|
| 1577 |
+
"num_input_tokens_seen": 10939824,
|
| 1578 |
+
"step": 1960
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 10.26160889470242,
|
| 1582 |
+
"grad_norm": 0.7800888419151306,
|
| 1583 |
+
"learning_rate": 3.679322828651265e-05,
|
| 1584 |
+
"loss": 0.043,
|
| 1585 |
+
"num_input_tokens_seen": 10996128,
|
| 1586 |
+
"step": 1970
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 10.313930673642904,
|
| 1590 |
+
"grad_norm": 1.8462598323822021,
|
| 1591 |
+
"learning_rate": 3.667219304670193e-05,
|
| 1592 |
+
"loss": 0.0454,
|
| 1593 |
+
"num_input_tokens_seen": 11053280,
|
| 1594 |
+
"step": 1980
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 10.366252452583387,
|
| 1598 |
+
"grad_norm": 1.075537085533142,
|
| 1599 |
+
"learning_rate": 3.655080694041495e-05,
|
| 1600 |
+
"loss": 0.0479,
|
| 1601 |
+
"num_input_tokens_seen": 11109664,
|
| 1602 |
+
"step": 1990
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 10.418574231523872,
|
| 1606 |
+
"grad_norm": 0.7474583983421326,
|
| 1607 |
+
"learning_rate": 3.642907361652172e-05,
|
| 1608 |
+
"loss": 0.0458,
|
| 1609 |
+
"num_input_tokens_seen": 11166016,
|
| 1610 |
+
"step": 2000
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 10.470896010464356,
|
| 1614 |
+
"grad_norm": 1.0032005310058594,
|
| 1615 |
+
"learning_rate": 3.6306996734329656e-05,
|
| 1616 |
+
"loss": 0.0518,
|
| 1617 |
+
"num_input_tokens_seen": 11221648,
|
| 1618 |
+
"step": 2010
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 10.523217789404839,
|
| 1622 |
+
"grad_norm": 1.0080397129058838,
|
| 1623 |
+
"learning_rate": 3.618457996347352e-05,
|
| 1624 |
+
"loss": 0.0501,
|
| 1625 |
+
"num_input_tokens_seen": 11277952,
|
| 1626 |
+
"step": 2020
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 10.575539568345324,
|
| 1630 |
+
"grad_norm": 1.130409598350525,
|
| 1631 |
+
"learning_rate": 3.606182698380515e-05,
|
| 1632 |
+
"loss": 0.0476,
|
| 1633 |
+
"num_input_tokens_seen": 11334272,
|
| 1634 |
+
"step": 2030
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 10.627861347285808,
|
| 1638 |
+
"grad_norm": 0.9204115867614746,
|
| 1639 |
+
"learning_rate": 3.593874148528284e-05,
|
| 1640 |
+
"loss": 0.0515,
|
| 1641 |
+
"num_input_tokens_seen": 11389760,
|
| 1642 |
+
"step": 2040
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 10.680183126226291,
|
| 1646 |
+
"grad_norm": 0.8044779896736145,
|
| 1647 |
+
"learning_rate": 3.58153271678604e-05,
|
| 1648 |
+
"loss": 0.0477,
|
| 1649 |
+
"num_input_tokens_seen": 11446144,
|
| 1650 |
+
"step": 2050
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 10.732504905166776,
|
| 1654 |
+
"grad_norm": 1.4428260326385498,
|
| 1655 |
+
"learning_rate": 3.5691587741375934e-05,
|
| 1656 |
+
"loss": 0.0488,
|
| 1657 |
+
"num_input_tokens_seen": 11502320,
|
| 1658 |
+
"step": 2060
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 10.78482668410726,
|
| 1662 |
+
"grad_norm": 0.9769074320793152,
|
| 1663 |
+
"learning_rate": 3.5567526925440353e-05,
|
| 1664 |
+
"loss": 0.0565,
|
| 1665 |
+
"num_input_tokens_seen": 11559392,
|
| 1666 |
+
"step": 2070
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 10.837148463047743,
|
| 1670 |
+
"grad_norm": 0.6837288737297058,
|
| 1671 |
+
"learning_rate": 3.5443148449325545e-05,
|
| 1672 |
+
"loss": 0.051,
|
| 1673 |
+
"num_input_tokens_seen": 11615824,
|
| 1674 |
+
"step": 2080
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 10.889470241988228,
|
| 1678 |
+
"grad_norm": 1.1239960193634033,
|
| 1679 |
+
"learning_rate": 3.5318456051852264e-05,
|
| 1680 |
+
"loss": 0.0528,
|
| 1681 |
+
"num_input_tokens_seen": 11671968,
|
| 1682 |
+
"step": 2090
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 10.941792020928713,
|
| 1686 |
+
"grad_norm": 1.2493700981140137,
|
| 1687 |
+
"learning_rate": 3.519345348127775e-05,
|
| 1688 |
+
"loss": 0.0479,
|
| 1689 |
+
"num_input_tokens_seen": 11727632,
|
| 1690 |
+
"step": 2100
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 10.994113799869195,
|
| 1694 |
+
"grad_norm": 1.0371160507202148,
|
| 1695 |
+
"learning_rate": 3.506814449518306e-05,
|
| 1696 |
+
"loss": 0.0446,
|
| 1697 |
+
"num_input_tokens_seen": 11784032,
|
| 1698 |
+
"step": 2110
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 11.041857423152388,
|
| 1702 |
+
"grad_norm": 0.6289274096488953,
|
| 1703 |
+
"learning_rate": 3.494253286036011e-05,
|
| 1704 |
+
"loss": 0.0392,
|
| 1705 |
+
"num_input_tokens_seen": 11835760,
|
| 1706 |
+
"step": 2120
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 11.094179202092871,
|
| 1710 |
+
"grad_norm": 1.0801132917404175,
|
| 1711 |
+
"learning_rate": 3.481662235269844e-05,
|
| 1712 |
+
"loss": 0.0311,
|
| 1713 |
+
"num_input_tokens_seen": 11891584,
|
| 1714 |
+
"step": 2130
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 11.146500981033356,
|
| 1718 |
+
"grad_norm": 0.7098966836929321,
|
| 1719 |
+
"learning_rate": 3.469041675707173e-05,
|
| 1720 |
+
"loss": 0.0253,
|
| 1721 |
+
"num_input_tokens_seen": 11947824,
|
| 1722 |
+
"step": 2140
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 11.198822759973838,
|
| 1726 |
+
"grad_norm": 0.9716883897781372,
|
| 1727 |
+
"learning_rate": 3.4563919867224e-05,
|
| 1728 |
+
"loss": 0.0313,
|
| 1729 |
+
"num_input_tokens_seen": 12003328,
|
| 1730 |
+
"step": 2150
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 11.251144538914323,
|
| 1734 |
+
"grad_norm": 1.6546238660812378,
|
| 1735 |
+
"learning_rate": 3.4437135485655575e-05,
|
| 1736 |
+
"loss": 0.0339,
|
| 1737 |
+
"num_input_tokens_seen": 12060512,
|
| 1738 |
+
"step": 2160
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 11.303466317854808,
|
| 1742 |
+
"grad_norm": 0.8873046040534973,
|
| 1743 |
+
"learning_rate": 3.4310067423508815e-05,
|
| 1744 |
+
"loss": 0.0344,
|
| 1745 |
+
"num_input_tokens_seen": 12117584,
|
| 1746 |
+
"step": 2170
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 11.35578809679529,
|
| 1750 |
+
"grad_norm": 1.2070609331130981,
|
| 1751 |
+
"learning_rate": 3.418271950045352e-05,
|
| 1752 |
+
"loss": 0.0274,
|
| 1753 |
+
"num_input_tokens_seen": 12172512,
|
| 1754 |
+
"step": 2180
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 11.408109875735775,
|
| 1758 |
+
"grad_norm": 1.0395190715789795,
|
| 1759 |
+
"learning_rate": 3.405509554457211e-05,
|
| 1760 |
+
"loss": 0.0268,
|
| 1761 |
+
"num_input_tokens_seen": 12227744,
|
| 1762 |
+
"step": 2190
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 11.46043165467626,
|
| 1766 |
+
"grad_norm": 1.2013176679611206,
|
| 1767 |
+
"learning_rate": 3.392719939224453e-05,
|
| 1768 |
+
"loss": 0.0363,
|
| 1769 |
+
"num_input_tokens_seen": 12283776,
|
| 1770 |
+
"step": 2200
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 11.512753433616743,
|
| 1774 |
+
"grad_norm": 1.1089166402816772,
|
| 1775 |
+
"learning_rate": 3.379903488803304e-05,
|
| 1776 |
+
"loss": 0.0339,
|
| 1777 |
+
"num_input_tokens_seen": 12340288,
|
| 1778 |
+
"step": 2210
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 11.565075212557227,
|
| 1782 |
+
"grad_norm": 1.4544755220413208,
|
| 1783 |
+
"learning_rate": 3.3670605884566484e-05,
|
| 1784 |
+
"loss": 0.0325,
|
| 1785 |
+
"num_input_tokens_seen": 12396368,
|
| 1786 |
+
"step": 2220
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 11.61739699149771,
|
| 1790 |
+
"grad_norm": 1.0953842401504517,
|
| 1791 |
+
"learning_rate": 3.3541916242424606e-05,
|
| 1792 |
+
"loss": 0.0333,
|
| 1793 |
+
"num_input_tokens_seen": 12451872,
|
| 1794 |
+
"step": 2230
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 11.669718770438195,
|
| 1798 |
+
"grad_norm": 0.7939156889915466,
|
| 1799 |
+
"learning_rate": 3.341296983002193e-05,
|
| 1800 |
+
"loss": 0.0336,
|
| 1801 |
+
"num_input_tokens_seen": 12507776,
|
| 1802 |
+
"step": 2240
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 11.72204054937868,
|
| 1806 |
+
"grad_norm": 1.114825963973999,
|
| 1807 |
+
"learning_rate": 3.3283770523491535e-05,
|
| 1808 |
+
"loss": 0.0357,
|
| 1809 |
+
"num_input_tokens_seen": 12564320,
|
| 1810 |
+
"step": 2250
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 11.774362328319162,
|
| 1814 |
+
"grad_norm": 1.0780022144317627,
|
| 1815 |
+
"learning_rate": 3.3154322206568475e-05,
|
| 1816 |
+
"loss": 0.0356,
|
| 1817 |
+
"num_input_tokens_seen": 12620912,
|
| 1818 |
+
"step": 2260
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 11.826684107259647,
|
| 1822 |
+
"grad_norm": 0.9889864325523376,
|
| 1823 |
+
"learning_rate": 3.302462877047307e-05,
|
| 1824 |
+
"loss": 0.0318,
|
| 1825 |
+
"num_input_tokens_seen": 12676464,
|
| 1826 |
+
"step": 2270
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 11.879005886200131,
|
| 1830 |
+
"grad_norm": 1.8913953304290771,
|
| 1831 |
+
"learning_rate": 3.2894694113793935e-05,
|
| 1832 |
+
"loss": 0.039,
|
| 1833 |
+
"num_input_tokens_seen": 12731408,
|
| 1834 |
+
"step": 2280
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 11.931327665140614,
|
| 1838 |
+
"grad_norm": 0.854158341884613,
|
| 1839 |
+
"learning_rate": 3.27645221423708e-05,
|
| 1840 |
+
"loss": 0.0408,
|
| 1841 |
+
"num_input_tokens_seen": 12787552,
|
| 1842 |
+
"step": 2290
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 11.983649444081099,
|
| 1846 |
+
"grad_norm": 0.7944401502609253,
|
| 1847 |
+
"learning_rate": 3.263411676917704e-05,
|
| 1848 |
+
"loss": 0.034,
|
| 1849 |
+
"num_input_tokens_seen": 12843808,
|
| 1850 |
+
"step": 2300
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 12.03139306736429,
|
| 1854 |
+
"grad_norm": 1.3224999904632568,
|
| 1855 |
+
"learning_rate": 3.250348191420214e-05,
|
| 1856 |
+
"loss": 0.03,
|
| 1857 |
+
"num_input_tokens_seen": 12895184,
|
| 1858 |
+
"step": 2310
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 12.083714846304774,
|
| 1862 |
+
"grad_norm": 0.8005937933921814,
|
| 1863 |
+
"learning_rate": 3.237262150433379e-05,
|
| 1864 |
+
"loss": 0.0219,
|
| 1865 |
+
"num_input_tokens_seen": 12951408,
|
| 1866 |
+
"step": 2320
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 12.136036625245259,
|
| 1870 |
+
"grad_norm": 1.3628501892089844,
|
| 1871 |
+
"learning_rate": 3.224153947323987e-05,
|
| 1872 |
+
"loss": 0.0181,
|
| 1873 |
+
"num_input_tokens_seen": 13007776,
|
| 1874 |
+
"step": 2330
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 12.188358404185742,
|
| 1878 |
+
"grad_norm": 0.7954509258270264,
|
| 1879 |
+
"learning_rate": 3.21102397612502e-05,
|
| 1880 |
+
"loss": 0.0183,
|
| 1881 |
+
"num_input_tokens_seen": 13064144,
|
| 1882 |
+
"step": 2340
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 12.240680183126226,
|
| 1886 |
+
"grad_norm": 0.8565235733985901,
|
| 1887 |
+
"learning_rate": 3.1978726315238094e-05,
|
| 1888 |
+
"loss": 0.0183,
|
| 1889 |
+
"num_input_tokens_seen": 13120320,
|
| 1890 |
+
"step": 2350
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 12.293001962066711,
|
| 1894 |
+
"grad_norm": 0.7555674910545349,
|
| 1895 |
+
"learning_rate": 3.1847003088501726e-05,
|
| 1896 |
+
"loss": 0.017,
|
| 1897 |
+
"num_input_tokens_seen": 13177168,
|
| 1898 |
+
"step": 2360
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 12.345323741007194,
|
| 1902 |
+
"grad_norm": 0.7122445106506348,
|
| 1903 |
+
"learning_rate": 3.1715074040645275e-05,
|
| 1904 |
+
"loss": 0.0206,
|
| 1905 |
+
"num_input_tokens_seen": 13232784,
|
| 1906 |
+
"step": 2370
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 12.397645519947678,
|
| 1910 |
+
"grad_norm": 0.9428816437721252,
|
| 1911 |
+
"learning_rate": 3.158294313745992e-05,
|
| 1912 |
+
"loss": 0.0194,
|
| 1913 |
+
"num_input_tokens_seen": 13287312,
|
| 1914 |
+
"step": 2380
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 12.449967298888161,
|
| 1918 |
+
"grad_norm": 1.027761459350586,
|
| 1919 |
+
"learning_rate": 3.145061435080461e-05,
|
| 1920 |
+
"loss": 0.0165,
|
| 1921 |
+
"num_input_tokens_seen": 13343616,
|
| 1922 |
+
"step": 2390
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 12.502289077828646,
|
| 1926 |
+
"grad_norm": 0.9591146111488342,
|
| 1927 |
+
"learning_rate": 3.1318091658486655e-05,
|
| 1928 |
+
"loss": 0.0192,
|
| 1929 |
+
"num_input_tokens_seen": 13398656,
|
| 1930 |
+
"step": 2400
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 12.55461085676913,
|
| 1934 |
+
"grad_norm": 2.0098116397857666,
|
| 1935 |
+
"learning_rate": 3.1185379044142225e-05,
|
| 1936 |
+
"loss": 0.0202,
|
| 1937 |
+
"num_input_tokens_seen": 13453888,
|
| 1938 |
+
"step": 2410
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 12.606932635709613,
|
| 1942 |
+
"grad_norm": 1.243646502494812,
|
| 1943 |
+
"learning_rate": 3.105248049711651e-05,
|
| 1944 |
+
"loss": 0.0184,
|
| 1945 |
+
"num_input_tokens_seen": 13511168,
|
| 1946 |
+
"step": 2420
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 12.659254414650098,
|
| 1950 |
+
"grad_norm": 0.7831906676292419,
|
| 1951 |
+
"learning_rate": 3.091940001234386e-05,
|
| 1952 |
+
"loss": 0.0215,
|
| 1953 |
+
"num_input_tokens_seen": 13567168,
|
| 1954 |
+
"step": 2430
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 12.711576193590583,
|
| 1958 |
+
"grad_norm": 0.6232236623764038,
|
| 1959 |
+
"learning_rate": 3.078614159022767e-05,
|
| 1960 |
+
"loss": 0.0192,
|
| 1961 |
+
"num_input_tokens_seen": 13623200,
|
| 1962 |
+
"step": 2440
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 12.763897972531066,
|
| 1966 |
+
"grad_norm": 1.3829624652862549,
|
| 1967 |
+
"learning_rate": 3.065270923652015e-05,
|
| 1968 |
+
"loss": 0.0222,
|
| 1969 |
+
"num_input_tokens_seen": 13678880,
|
| 1970 |
+
"step": 2450
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 12.81621975147155,
|
| 1974 |
+
"grad_norm": 0.9216393232345581,
|
| 1975 |
+
"learning_rate": 3.051910696220188e-05,
|
| 1976 |
+
"loss": 0.0159,
|
| 1977 |
+
"num_input_tokens_seen": 13734624,
|
| 1978 |
+
"step": 2460
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 12.868541530412035,
|
| 1982 |
+
"grad_norm": 1.1284723281860352,
|
| 1983 |
+
"learning_rate": 3.0385338783361283e-05,
|
| 1984 |
+
"loss": 0.0248,
|
| 1985 |
+
"num_input_tokens_seen": 13790576,
|
| 1986 |
+
"step": 2470
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 12.920863309352518,
|
| 1990 |
+
"grad_norm": 1.4107064008712769,
|
| 1991 |
+
"learning_rate": 3.025140872107386e-05,
|
| 1992 |
+
"loss": 0.0227,
|
| 1993 |
+
"num_input_tokens_seen": 13845984,
|
| 1994 |
+
"step": 2480
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 12.973185088293002,
|
| 1998 |
+
"grad_norm": 0.7538278102874756,
|
| 1999 |
+
"learning_rate": 3.0117320801281335e-05,
|
| 2000 |
+
"loss": 0.0265,
|
| 2001 |
+
"num_input_tokens_seen": 13902400,
|
| 2002 |
+
"step": 2490
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 13.020928711576193,
|
| 2006 |
+
"grad_norm": 0.9580565690994263,
|
| 2007 |
+
"learning_rate": 2.9983079054670627e-05,
|
| 2008 |
+
"loss": 0.0195,
|
| 2009 |
+
"num_input_tokens_seen": 13953344,
|
| 2010 |
+
"step": 2500
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 13.073250490516678,
|
| 2014 |
+
"grad_norm": 0.4210267961025238,
|
| 2015 |
+
"learning_rate": 2.9848687516552725e-05,
|
| 2016 |
+
"loss": 0.0107,
|
| 2017 |
+
"num_input_tokens_seen": 14009424,
|
| 2018 |
+
"step": 2510
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 13.125572269457162,
|
| 2022 |
+
"grad_norm": 0.534055233001709,
|
| 2023 |
+
"learning_rate": 2.9714150226741312e-05,
|
| 2024 |
+
"loss": 0.013,
|
| 2025 |
+
"num_input_tokens_seen": 14064880,
|
| 2026 |
+
"step": 2520
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 13.177894048397645,
|
| 2030 |
+
"grad_norm": 0.8242612481117249,
|
| 2031 |
+
"learning_rate": 2.9579471229431394e-05,
|
| 2032 |
+
"loss": 0.0095,
|
| 2033 |
+
"num_input_tokens_seen": 14120896,
|
| 2034 |
+
"step": 2530
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 13.23021582733813,
|
| 2038 |
+
"grad_norm": 1.1644961833953857,
|
| 2039 |
+
"learning_rate": 2.944465457307771e-05,
|
| 2040 |
+
"loss": 0.0125,
|
| 2041 |
+
"num_input_tokens_seen": 14176512,
|
| 2042 |
+
"step": 2540
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 13.282537606278613,
|
| 2046 |
+
"grad_norm": 0.682830274105072,
|
| 2047 |
+
"learning_rate": 2.930970431027304e-05,
|
| 2048 |
+
"loss": 0.0106,
|
| 2049 |
+
"num_input_tokens_seen": 14232608,
|
| 2050 |
+
"step": 2550
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 13.334859385219097,
|
| 2054 |
+
"grad_norm": 0.7904958724975586,
|
| 2055 |
+
"learning_rate": 2.9174624497626353e-05,
|
| 2056 |
+
"loss": 0.012,
|
| 2057 |
+
"num_input_tokens_seen": 14289360,
|
| 2058 |
+
"step": 2560
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 13.387181164159582,
|
| 2062 |
+
"grad_norm": 0.8092711567878723,
|
| 2063 |
+
"learning_rate": 2.903941919564091e-05,
|
| 2064 |
+
"loss": 0.0124,
|
| 2065 |
+
"num_input_tokens_seen": 14346096,
|
| 2066 |
+
"step": 2570
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 13.439502943100065,
|
| 2070 |
+
"grad_norm": 0.4784017503261566,
|
| 2071 |
+
"learning_rate": 2.8904092468592187e-05,
|
| 2072 |
+
"loss": 0.0132,
|
| 2073 |
+
"num_input_tokens_seen": 14401872,
|
| 2074 |
+
"step": 2580
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 13.49182472204055,
|
| 2078 |
+
"grad_norm": 0.5194114446640015,
|
| 2079 |
+
"learning_rate": 2.8768648384405695e-05,
|
| 2080 |
+
"loss": 0.0101,
|
| 2081 |
+
"num_input_tokens_seen": 14458864,
|
| 2082 |
+
"step": 2590
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 13.544146500981034,
|
| 2086 |
+
"grad_norm": 0.6601864099502563,
|
| 2087 |
+
"learning_rate": 2.863309101453469e-05,
|
| 2088 |
+
"loss": 0.0135,
|
| 2089 |
+
"num_input_tokens_seen": 14515664,
|
| 2090 |
+
"step": 2600
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 13.596468279921517,
|
| 2094 |
+
"grad_norm": 0.9567685723304749,
|
| 2095 |
+
"learning_rate": 2.8497424433837833e-05,
|
| 2096 |
+
"loss": 0.0138,
|
| 2097 |
+
"num_input_tokens_seen": 14572256,
|
| 2098 |
+
"step": 2610
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 13.648790058862001,
|
| 2102 |
+
"grad_norm": 0.5563291311264038,
|
| 2103 |
+
"learning_rate": 2.836165272045663e-05,
|
| 2104 |
+
"loss": 0.0132,
|
| 2105 |
+
"num_input_tokens_seen": 14627248,
|
| 2106 |
+
"step": 2620
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 13.701111837802486,
|
| 2110 |
+
"grad_norm": 0.9716143608093262,
|
| 2111 |
+
"learning_rate": 2.8225779955692905e-05,
|
| 2112 |
+
"loss": 0.0134,
|
| 2113 |
+
"num_input_tokens_seen": 14683728,
|
| 2114 |
+
"step": 2630
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 13.753433616742969,
|
| 2118 |
+
"grad_norm": 0.9606854915618896,
|
| 2119 |
+
"learning_rate": 2.8089810223886076e-05,
|
| 2120 |
+
"loss": 0.0154,
|
| 2121 |
+
"num_input_tokens_seen": 14740864,
|
| 2122 |
+
"step": 2640
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 13.805755395683454,
|
| 2126 |
+
"grad_norm": 1.01091730594635,
|
| 2127 |
+
"learning_rate": 2.79537476122904e-05,
|
| 2128 |
+
"loss": 0.0121,
|
| 2129 |
+
"num_input_tokens_seen": 14796176,
|
| 2130 |
+
"step": 2650
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 13.858077174623936,
|
| 2134 |
+
"grad_norm": 0.6134788990020752,
|
| 2135 |
+
"learning_rate": 2.781759621095209e-05,
|
| 2136 |
+
"loss": 0.0119,
|
| 2137 |
+
"num_input_tokens_seen": 14852304,
|
| 2138 |
+
"step": 2660
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 13.910398953564421,
|
| 2142 |
+
"grad_norm": 1.1731514930725098,
|
| 2143 |
+
"learning_rate": 2.7681360112586403e-05,
|
| 2144 |
+
"loss": 0.0188,
|
| 2145 |
+
"num_input_tokens_seen": 14908624,
|
| 2146 |
+
"step": 2670
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 13.962720732504906,
|
| 2150 |
+
"grad_norm": 0.7572017908096313,
|
| 2151 |
+
"learning_rate": 2.7545043412454568e-05,
|
| 2152 |
+
"loss": 0.0153,
|
| 2153 |
+
"num_input_tokens_seen": 14964784,
|
| 2154 |
+
"step": 2680
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 14.010464355788097,
|
| 2158 |
+
"grad_norm": 0.2877230942249298,
|
| 2159 |
+
"learning_rate": 2.7408650208240733e-05,
|
| 2160 |
+
"loss": 0.0093,
|
| 2161 |
+
"num_input_tokens_seen": 15016112,
|
| 2162 |
+
"step": 2690
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 14.062786134728581,
|
| 2166 |
+
"grad_norm": 1.2057104110717773,
|
| 2167 |
+
"learning_rate": 2.7272184599928723e-05,
|
| 2168 |
+
"loss": 0.006,
|
| 2169 |
+
"num_input_tokens_seen": 15072240,
|
| 2170 |
+
"step": 2700
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 14.115107913669064,
|
| 2174 |
+
"grad_norm": 0.2993405759334564,
|
| 2175 |
+
"learning_rate": 2.7135650689678873e-05,
|
| 2176 |
+
"loss": 0.0082,
|
| 2177 |
+
"num_input_tokens_seen": 15128432,
|
| 2178 |
+
"step": 2710
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 14.167429692609549,
|
| 2182 |
+
"grad_norm": 0.4322413504123688,
|
| 2183 |
+
"learning_rate": 2.6999052581704643e-05,
|
| 2184 |
+
"loss": 0.0052,
|
| 2185 |
+
"num_input_tokens_seen": 15185232,
|
| 2186 |
+
"step": 2720
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 14.219751471550033,
|
| 2190 |
+
"grad_norm": 0.4944108724594116,
|
| 2191 |
+
"learning_rate": 2.6862394382149308e-05,
|
| 2192 |
+
"loss": 0.0066,
|
| 2193 |
+
"num_input_tokens_seen": 15241040,
|
| 2194 |
+
"step": 2730
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 14.272073250490516,
|
| 2198 |
+
"grad_norm": 0.6533095836639404,
|
| 2199 |
+
"learning_rate": 2.672568019896248e-05,
|
| 2200 |
+
"loss": 0.0088,
|
| 2201 |
+
"num_input_tokens_seen": 15297904,
|
| 2202 |
+
"step": 2740
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 14.324395029431,
|
| 2206 |
+
"grad_norm": 0.316057026386261,
|
| 2207 |
+
"learning_rate": 2.6588914141776626e-05,
|
| 2208 |
+
"loss": 0.0056,
|
| 2209 |
+
"num_input_tokens_seen": 15355584,
|
| 2210 |
+
"step": 2750
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 14.376716808371485,
|
| 2214 |
+
"grad_norm": 0.502487063407898,
|
| 2215 |
+
"learning_rate": 2.6452100321783585e-05,
|
| 2216 |
+
"loss": 0.0029,
|
| 2217 |
+
"num_input_tokens_seen": 15410592,
|
| 2218 |
+
"step": 2760
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 14.429038587311968,
|
| 2222 |
+
"grad_norm": 0.5012995004653931,
|
| 2223 |
+
"learning_rate": 2.6315242851610923e-05,
|
| 2224 |
+
"loss": 0.0109,
|
| 2225 |
+
"num_input_tokens_seen": 15466448,
|
| 2226 |
+
"step": 2770
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 14.481360366252453,
|
| 2230 |
+
"grad_norm": 0.8451622128486633,
|
| 2231 |
+
"learning_rate": 2.6178345845198328e-05,
|
| 2232 |
+
"loss": 0.0057,
|
| 2233 |
+
"num_input_tokens_seen": 15522816,
|
| 2234 |
+
"step": 2780
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 14.533682145192937,
|
| 2238 |
+
"grad_norm": 0.33364975452423096,
|
| 2239 |
+
"learning_rate": 2.6041413417673966e-05,
|
| 2240 |
+
"loss": 0.009,
|
| 2241 |
+
"num_input_tokens_seen": 15578672,
|
| 2242 |
+
"step": 2790
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 14.58600392413342,
|
| 2246 |
+
"grad_norm": 0.3638412058353424,
|
| 2247 |
+
"learning_rate": 2.590444968523074e-05,
|
| 2248 |
+
"loss": 0.0089,
|
| 2249 |
+
"num_input_tokens_seen": 15635408,
|
| 2250 |
+
"step": 2800
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 14.638325703073905,
|
| 2254 |
+
"grad_norm": 0.5961637496948242,
|
| 2255 |
+
"learning_rate": 2.5767458765002606e-05,
|
| 2256 |
+
"loss": 0.008,
|
| 2257 |
+
"num_input_tokens_seen": 15691648,
|
| 2258 |
+
"step": 2810
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 14.690647482014388,
|
| 2262 |
+
"grad_norm": 0.7401494979858398,
|
| 2263 |
+
"learning_rate": 2.5630444774940765e-05,
|
| 2264 |
+
"loss": 0.0081,
|
| 2265 |
+
"num_input_tokens_seen": 15748032,
|
| 2266 |
+
"step": 2820
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 14.742969260954872,
|
| 2270 |
+
"grad_norm": 0.18349328637123108,
|
| 2271 |
+
"learning_rate": 2.5493411833689907e-05,
|
| 2272 |
+
"loss": 0.0071,
|
| 2273 |
+
"num_input_tokens_seen": 15803232,
|
| 2274 |
+
"step": 2830
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 14.795291039895357,
|
| 2278 |
+
"grad_norm": 1.5877436399459839,
|
| 2279 |
+
"learning_rate": 2.5356364060464398e-05,
|
| 2280 |
+
"loss": 0.0078,
|
| 2281 |
+
"num_input_tokens_seen": 15859120,
|
| 2282 |
+
"step": 2840
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 14.84761281883584,
|
| 2286 |
+
"grad_norm": 1.041685938835144,
|
| 2287 |
+
"learning_rate": 2.521930557492444e-05,
|
| 2288 |
+
"loss": 0.0089,
|
| 2289 |
+
"num_input_tokens_seen": 15915872,
|
| 2290 |
+
"step": 2850
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 14.899934597776324,
|
| 2294 |
+
"grad_norm": 0.6710309982299805,
|
| 2295 |
+
"learning_rate": 2.5082240497052267e-05,
|
| 2296 |
+
"loss": 0.0088,
|
| 2297 |
+
"num_input_tokens_seen": 15973472,
|
| 2298 |
+
"step": 2860
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 14.952256376716809,
|
| 2302 |
+
"grad_norm": 0.9839669466018677,
|
| 2303 |
+
"learning_rate": 2.494517294702826e-05,
|
| 2304 |
+
"loss": 0.0069,
|
| 2305 |
+
"num_input_tokens_seen": 16029920,
|
| 2306 |
+
"step": 2870
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 15.0,
|
| 2310 |
+
"grad_norm": 1.3188862800598145,
|
| 2311 |
+
"learning_rate": 2.4808107045107123e-05,
|
| 2312 |
+
"loss": 0.0098,
|
| 2313 |
+
"num_input_tokens_seen": 16080272,
|
| 2314 |
+
"step": 2880
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 15.052321778940485,
|
| 2318 |
+
"grad_norm": 0.41307705640792847,
|
| 2319 |
+
"learning_rate": 2.4671046911494025e-05,
|
| 2320 |
+
"loss": 0.0037,
|
| 2321 |
+
"num_input_tokens_seen": 16136752,
|
| 2322 |
+
"step": 2890
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 15.104643557880967,
|
| 2326 |
+
"grad_norm": 0.8025826811790466,
|
| 2327 |
+
"learning_rate": 2.453399666622072e-05,
|
| 2328 |
+
"loss": 0.0032,
|
| 2329 |
+
"num_input_tokens_seen": 16191920,
|
| 2330 |
+
"step": 2900
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 15.156965336821452,
|
| 2334 |
+
"grad_norm": 0.2138717621564865,
|
| 2335 |
+
"learning_rate": 2.4396960429021738e-05,
|
| 2336 |
+
"loss": 0.0028,
|
| 2337 |
+
"num_input_tokens_seen": 16246912,
|
| 2338 |
+
"step": 2910
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 15.209287115761937,
|
| 2342 |
+
"grad_norm": 0.16525974869728088,
|
| 2343 |
+
"learning_rate": 2.4259942319210498e-05,
|
| 2344 |
+
"loss": 0.0058,
|
| 2345 |
+
"num_input_tokens_seen": 16303520,
|
| 2346 |
+
"step": 2920
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 15.26160889470242,
|
| 2350 |
+
"grad_norm": 0.22977286577224731,
|
| 2351 |
+
"learning_rate": 2.412294645555555e-05,
|
| 2352 |
+
"loss": 0.005,
|
| 2353 |
+
"num_input_tokens_seen": 16359888,
|
| 2354 |
+
"step": 2930
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 15.313930673642904,
|
| 2358 |
+
"grad_norm": 1.8685427904129028,
|
| 2359 |
+
"learning_rate": 2.39859769561567e-05,
|
| 2360 |
+
"loss": 0.0048,
|
| 2361 |
+
"num_input_tokens_seen": 16416512,
|
| 2362 |
+
"step": 2940
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 15.366252452583387,
|
| 2366 |
+
"grad_norm": 0.7653654217720032,
|
| 2367 |
+
"learning_rate": 2.3849037938321235e-05,
|
| 2368 |
+
"loss": 0.0056,
|
| 2369 |
+
"num_input_tokens_seen": 16473664,
|
| 2370 |
+
"step": 2950
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 15.418574231523872,
|
| 2374 |
+
"grad_norm": 0.21836823225021362,
|
| 2375 |
+
"learning_rate": 2.3712133518440176e-05,
|
| 2376 |
+
"loss": 0.0072,
|
| 2377 |
+
"num_input_tokens_seen": 16529312,
|
| 2378 |
+
"step": 2960
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 15.470896010464356,
|
| 2382 |
+
"grad_norm": 0.6065989136695862,
|
| 2383 |
+
"learning_rate": 2.3575267811864543e-05,
|
| 2384 |
+
"loss": 0.0074,
|
| 2385 |
+
"num_input_tokens_seen": 16586048,
|
| 2386 |
+
"step": 2970
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 15.523217789404839,
|
| 2390 |
+
"grad_norm": 0.21767759323120117,
|
| 2391 |
+
"learning_rate": 2.34384449327816e-05,
|
| 2392 |
+
"loss": 0.0037,
|
| 2393 |
+
"num_input_tokens_seen": 16642560,
|
| 2394 |
+
"step": 2980
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 15.575539568345324,
|
| 2398 |
+
"grad_norm": 0.35699793696403503,
|
| 2399 |
+
"learning_rate": 2.330166899409124e-05,
|
| 2400 |
+
"loss": 0.0039,
|
| 2401 |
+
"num_input_tokens_seen": 16699248,
|
| 2402 |
+
"step": 2990
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 15.627861347285808,
|
| 2406 |
+
"grad_norm": 0.46648091077804565,
|
| 2407 |
+
"learning_rate": 2.3164944107282333e-05,
|
| 2408 |
+
"loss": 0.0067,
|
| 2409 |
+
"num_input_tokens_seen": 16755952,
|
| 2410 |
+
"step": 3000
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 15.680183126226291,
|
| 2414 |
+
"grad_norm": 0.8464080691337585,
|
| 2415 |
+
"learning_rate": 2.3028274382309097e-05,
|
| 2416 |
+
"loss": 0.0061,
|
| 2417 |
+
"num_input_tokens_seen": 16811536,
|
| 2418 |
+
"step": 3010
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 15.732504905166776,
|
| 2422 |
+
"grad_norm": 0.30818283557891846,
|
| 2423 |
+
"learning_rate": 2.2891663927467604e-05,
|
| 2424 |
+
"loss": 0.0046,
|
| 2425 |
+
"num_input_tokens_seen": 16867824,
|
| 2426 |
+
"step": 3020
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 15.78482668410726,
|
| 2430 |
+
"grad_norm": 0.5573921799659729,
|
| 2431 |
+
"learning_rate": 2.2755116849272274e-05,
|
| 2432 |
+
"loss": 0.0041,
|
| 2433 |
+
"num_input_tokens_seen": 16924080,
|
| 2434 |
+
"step": 3030
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 15.837148463047743,
|
| 2438 |
+
"grad_norm": 0.5058010816574097,
|
| 2439 |
+
"learning_rate": 2.2618637252332398e-05,
|
| 2440 |
+
"loss": 0.0065,
|
| 2441 |
+
"num_input_tokens_seen": 16979728,
|
| 2442 |
+
"step": 3040
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 15.889470241988228,
|
| 2446 |
+
"grad_norm": 0.4849563241004944,
|
| 2447 |
+
"learning_rate": 2.2482229239228785e-05,
|
| 2448 |
+
"loss": 0.0047,
|
| 2449 |
+
"num_input_tokens_seen": 17035488,
|
| 2450 |
+
"step": 3050
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 15.941792020928713,
|
| 2454 |
+
"grad_norm": 0.10410932451486588,
|
| 2455 |
+
"learning_rate": 2.234589691039046e-05,
|
| 2456 |
+
"loss": 0.0054,
|
| 2457 |
+
"num_input_tokens_seen": 17091072,
|
| 2458 |
+
"step": 3060
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 15.994113799869195,
|
| 2462 |
+
"grad_norm": 0.31193605065345764,
|
| 2463 |
+
"learning_rate": 2.2209644363971337e-05,
|
| 2464 |
+
"loss": 0.0043,
|
| 2465 |
+
"num_input_tokens_seen": 17147328,
|
| 2466 |
+
"step": 3070
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 16.041857423152386,
|
| 2470 |
+
"grad_norm": 0.15345972776412964,
|
| 2471 |
+
"learning_rate": 2.2073475695727096e-05,
|
| 2472 |
+
"loss": 0.0045,
|
| 2473 |
+
"num_input_tokens_seen": 17198200,
|
| 2474 |
+
"step": 3080
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 16.09417920209287,
|
| 2478 |
+
"grad_norm": 0.8098782896995544,
|
| 2479 |
+
"learning_rate": 2.193739499889201e-05,
|
| 2480 |
+
"loss": 0.0042,
|
| 2481 |
+
"num_input_tokens_seen": 17254408,
|
| 2482 |
+
"step": 3090
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 16.146500981033356,
|
| 2486 |
+
"grad_norm": 0.6010912656784058,
|
| 2487 |
+
"learning_rate": 2.1801406364055958e-05,
|
| 2488 |
+
"loss": 0.0049,
|
| 2489 |
+
"num_input_tokens_seen": 17311304,
|
| 2490 |
+
"step": 3100
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 16.19882275997384,
|
| 2494 |
+
"grad_norm": 0.0812903568148613,
|
| 2495 |
+
"learning_rate": 2.1665513879041418e-05,
|
| 2496 |
+
"loss": 0.001,
|
| 2497 |
+
"num_input_tokens_seen": 17368152,
|
| 2498 |
+
"step": 3110
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 16.251144538914325,
|
| 2502 |
+
"grad_norm": 0.08344978094100952,
|
| 2503 |
+
"learning_rate": 2.1529721628780593e-05,
|
| 2504 |
+
"loss": 0.0037,
|
| 2505 |
+
"num_input_tokens_seen": 17423480,
|
| 2506 |
+
"step": 3120
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 16.303466317854806,
|
| 2510 |
+
"grad_norm": 0.3543494641780853,
|
| 2511 |
+
"learning_rate": 2.1394033695192645e-05,
|
| 2512 |
+
"loss": 0.0016,
|
| 2513 |
+
"num_input_tokens_seen": 17478984,
|
| 2514 |
+
"step": 3130
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 16.35578809679529,
|
| 2518 |
+
"grad_norm": 0.683672308921814,
|
| 2519 |
+
"learning_rate": 2.125845415706097e-05,
|
| 2520 |
+
"loss": 0.0019,
|
| 2521 |
+
"num_input_tokens_seen": 17535592,
|
| 2522 |
+
"step": 3140
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 16.408109875735775,
|
| 2526 |
+
"grad_norm": 0.07661443203687668,
|
| 2527 |
+
"learning_rate": 2.1122987089910577e-05,
|
| 2528 |
+
"loss": 0.0012,
|
| 2529 |
+
"num_input_tokens_seen": 17591960,
|
| 2530 |
+
"step": 3150
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 16.46043165467626,
|
| 2534 |
+
"grad_norm": 0.21092914044857025,
|
| 2535 |
+
"learning_rate": 2.0987636565885606e-05,
|
| 2536 |
+
"loss": 0.004,
|
| 2537 |
+
"num_input_tokens_seen": 17648504,
|
| 2538 |
+
"step": 3160
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 16.512753433616744,
|
| 2542 |
+
"grad_norm": 0.8990269303321838,
|
| 2543 |
+
"learning_rate": 2.0852406653626916e-05,
|
| 2544 |
+
"loss": 0.003,
|
| 2545 |
+
"num_input_tokens_seen": 17705240,
|
| 2546 |
+
"step": 3170
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 16.565075212557225,
|
| 2550 |
+
"grad_norm": 0.22914479672908783,
|
| 2551 |
+
"learning_rate": 2.0717301418149742e-05,
|
| 2552 |
+
"loss": 0.0028,
|
| 2553 |
+
"num_input_tokens_seen": 17760392,
|
| 2554 |
+
"step": 3180
|
| 2555 |
+
},
|
| 2556 |
+
{
|
| 2557 |
+
"epoch": 16.61739699149771,
|
| 2558 |
+
"grad_norm": 1.7169655561447144,
|
| 2559 |
+
"learning_rate": 2.058232492072157e-05,
|
| 2560 |
+
"loss": 0.0033,
|
| 2561 |
+
"num_input_tokens_seen": 17816744,
|
| 2562 |
+
"step": 3190
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 16.669718770438195,
|
| 2566 |
+
"grad_norm": 0.07557539641857147,
|
| 2567 |
+
"learning_rate": 2.044748121874e-05,
|
| 2568 |
+
"loss": 0.0021,
|
| 2569 |
+
"num_input_tokens_seen": 17872760,
|
| 2570 |
+
"step": 3200
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 16.72204054937868,
|
| 2574 |
+
"grad_norm": 0.1961260586977005,
|
| 2575 |
+
"learning_rate": 2.0312774365610783e-05,
|
| 2576 |
+
"loss": 0.0017,
|
| 2577 |
+
"num_input_tokens_seen": 17928696,
|
| 2578 |
+
"step": 3210
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 16.774362328319164,
|
| 2582 |
+
"grad_norm": 0.6838825941085815,
|
| 2583 |
+
"learning_rate": 2.0178208410626006e-05,
|
| 2584 |
+
"loss": 0.0011,
|
| 2585 |
+
"num_input_tokens_seen": 17984232,
|
| 2586 |
+
"step": 3220
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 16.82668410725965,
|
| 2590 |
+
"grad_norm": 0.12144844979047775,
|
| 2591 |
+
"learning_rate": 2.0043787398842347e-05,
|
| 2592 |
+
"loss": 0.0022,
|
| 2593 |
+
"num_input_tokens_seen": 18040712,
|
| 2594 |
+
"step": 3230
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 16.87900588620013,
|
| 2598 |
+
"grad_norm": 0.3987827003002167,
|
| 2599 |
+
"learning_rate": 1.9909515370959493e-05,
|
| 2600 |
+
"loss": 0.0029,
|
| 2601 |
+
"num_input_tokens_seen": 18097016,
|
| 2602 |
+
"step": 3240
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 16.931327665140614,
|
| 2606 |
+
"grad_norm": 0.10082973539829254,
|
| 2607 |
+
"learning_rate": 1.9775396363198654e-05,
|
| 2608 |
+
"loss": 0.0015,
|
| 2609 |
+
"num_input_tokens_seen": 18152776,
|
| 2610 |
+
"step": 3250
|
| 2611 |
+
},
|
| 2612 |
+
{
|
| 2613 |
+
"epoch": 16.9836494440811,
|
| 2614 |
+
"grad_norm": 0.020283468067646027,
|
| 2615 |
+
"learning_rate": 1.9641434407181285e-05,
|
| 2616 |
+
"loss": 0.002,
|
| 2617 |
+
"num_input_tokens_seen": 18208456,
|
| 2618 |
+
"step": 3260
|
| 2619 |
+
},
|
| 2620 |
+
{
|
| 2621 |
+
"epoch": 17.03139306736429,
|
| 2622 |
+
"grad_norm": 0.23080122470855713,
|
| 2623 |
+
"learning_rate": 1.950763352980782e-05,
|
| 2624 |
+
"loss": 0.0006,
|
| 2625 |
+
"num_input_tokens_seen": 18259784,
|
| 2626 |
+
"step": 3270
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 17.083714846304776,
|
| 2630 |
+
"grad_norm": 0.00946098379790783,
|
| 2631 |
+
"learning_rate": 1.9373997753136695e-05,
|
| 2632 |
+
"loss": 0.0003,
|
| 2633 |
+
"num_input_tokens_seen": 18316008,
|
| 2634 |
+
"step": 3280
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 17.136036625245257,
|
| 2638 |
+
"grad_norm": 0.033656761050224304,
|
| 2639 |
+
"learning_rate": 1.9240531094263388e-05,
|
| 2640 |
+
"loss": 0.0008,
|
| 2641 |
+
"num_input_tokens_seen": 18372696,
|
| 2642 |
+
"step": 3290
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 17.188358404185742,
|
| 2646 |
+
"grad_norm": 0.03647719696164131,
|
| 2647 |
+
"learning_rate": 1.9107237565199716e-05,
|
| 2648 |
+
"loss": 0.0009,
|
| 2649 |
+
"num_input_tokens_seen": 18428488,
|
| 2650 |
+
"step": 3300
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 17.240680183126226,
|
| 2654 |
+
"grad_norm": 0.030296266078948975,
|
| 2655 |
+
"learning_rate": 1.8974121172753192e-05,
|
| 2656 |
+
"loss": 0.0004,
|
| 2657 |
+
"num_input_tokens_seen": 18484120,
|
| 2658 |
+
"step": 3310
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 17.29300196206671,
|
| 2662 |
+
"grad_norm": 0.043923936784267426,
|
| 2663 |
+
"learning_rate": 1.8841185918406594e-05,
|
| 2664 |
+
"loss": 0.0004,
|
| 2665 |
+
"num_input_tokens_seen": 18539976,
|
| 2666 |
+
"step": 3320
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 17.345323741007196,
|
| 2670 |
+
"grad_norm": 0.038960348814725876,
|
| 2671 |
+
"learning_rate": 1.870843579819771e-05,
|
| 2672 |
+
"loss": 0.0008,
|
| 2673 |
+
"num_input_tokens_seen": 18596792,
|
| 2674 |
+
"step": 3330
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"epoch": 17.397645519947677,
|
| 2678 |
+
"grad_norm": 0.010129265487194061,
|
| 2679 |
+
"learning_rate": 1.8575874802599162e-05,
|
| 2680 |
+
"loss": 0.001,
|
| 2681 |
+
"num_input_tokens_seen": 18652776,
|
| 2682 |
+
"step": 3340
|
| 2683 |
+
},
|
| 2684 |
+
{
|
| 2685 |
+
"epoch": 17.44996729888816,
|
| 2686 |
+
"grad_norm": 0.04958868771791458,
|
| 2687 |
+
"learning_rate": 1.8443506916398485e-05,
|
| 2688 |
+
"loss": 0.0004,
|
| 2689 |
+
"num_input_tokens_seen": 18709320,
|
| 2690 |
+
"step": 3350
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 17.502289077828646,
|
| 2694 |
+
"grad_norm": 2.036484956741333,
|
| 2695 |
+
"learning_rate": 1.8311336118578355e-05,
|
| 2696 |
+
"loss": 0.0018,
|
| 2697 |
+
"num_input_tokens_seen": 18766376,
|
| 2698 |
+
"step": 3360
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 17.55461085676913,
|
| 2702 |
+
"grad_norm": 0.10162217170000076,
|
| 2703 |
+
"learning_rate": 1.8179366382196944e-05,
|
| 2704 |
+
"loss": 0.0015,
|
| 2705 |
+
"num_input_tokens_seen": 18822440,
|
| 2706 |
+
"step": 3370
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 17.606932635709615,
|
| 2710 |
+
"grad_norm": 0.41068536043167114,
|
| 2711 |
+
"learning_rate": 1.8047601674268522e-05,
|
| 2712 |
+
"loss": 0.0011,
|
| 2713 |
+
"num_input_tokens_seen": 18877976,
|
| 2714 |
+
"step": 3380
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 17.6592544146501,
|
| 2718 |
+
"grad_norm": 0.35680681467056274,
|
| 2719 |
+
"learning_rate": 1.7916045955644207e-05,
|
| 2720 |
+
"loss": 0.0015,
|
| 2721 |
+
"num_input_tokens_seen": 18934728,
|
| 2722 |
+
"step": 3390
|
| 2723 |
+
},
|
| 2724 |
+
{
|
| 2725 |
+
"epoch": 17.71157619359058,
|
| 2726 |
+
"grad_norm": 0.16506262123584747,
|
| 2727 |
+
"learning_rate": 1.7784703180892882e-05,
|
| 2728 |
+
"loss": 0.0004,
|
| 2729 |
+
"num_input_tokens_seen": 18990088,
|
| 2730 |
+
"step": 3400
|
| 2731 |
+
},
|
| 2732 |
+
{
|
| 2733 |
+
"epoch": 17.763897972531066,
|
| 2734 |
+
"grad_norm": 0.05834071710705757,
|
| 2735 |
+
"learning_rate": 1.7653577298182327e-05,
|
| 2736 |
+
"loss": 0.0004,
|
| 2737 |
+
"num_input_tokens_seen": 19046728,
|
| 2738 |
+
"step": 3410
|
| 2739 |
+
},
|
| 2740 |
+
{
|
| 2741 |
+
"epoch": 17.81621975147155,
|
| 2742 |
+
"grad_norm": 0.14426672458648682,
|
| 2743 |
+
"learning_rate": 1.752267224916055e-05,
|
| 2744 |
+
"loss": 0.0022,
|
| 2745 |
+
"num_input_tokens_seen": 19101672,
|
| 2746 |
+
"step": 3420
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 17.868541530412035,
|
| 2750 |
+
"grad_norm": 0.2213674634695053,
|
| 2751 |
+
"learning_rate": 1.7391991968837272e-05,
|
| 2752 |
+
"loss": 0.0007,
|
| 2753 |
+
"num_input_tokens_seen": 19159128,
|
| 2754 |
+
"step": 3430
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 17.92086330935252,
|
| 2758 |
+
"grad_norm": 0.6867311000823975,
|
| 2759 |
+
"learning_rate": 1.726154038546569e-05,
|
| 2760 |
+
"loss": 0.0005,
|
| 2761 |
+
"num_input_tokens_seen": 19215448,
|
| 2762 |
+
"step": 3440
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 17.973185088293,
|
| 2766 |
+
"grad_norm": 0.022834990173578262,
|
| 2767 |
+
"learning_rate": 1.713132142042434e-05,
|
| 2768 |
+
"loss": 0.0013,
|
| 2769 |
+
"num_input_tokens_seen": 19270328,
|
| 2770 |
+
"step": 3450
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 18.020928711576193,
|
| 2774 |
+
"grad_norm": 0.017958860844373703,
|
| 2775 |
+
"learning_rate": 1.7001338988099264e-05,
|
| 2776 |
+
"loss": 0.0004,
|
| 2777 |
+
"num_input_tokens_seen": 19321096,
|
| 2778 |
+
"step": 3460
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 18.073250490516678,
|
| 2782 |
+
"grad_norm": 0.013346249237656593,
|
| 2783 |
+
"learning_rate": 1.68715969957663e-05,
|
| 2784 |
+
"loss": 0.0001,
|
| 2785 |
+
"num_input_tokens_seen": 19377144,
|
| 2786 |
+
"step": 3470
|
| 2787 |
+
},
|
| 2788 |
+
{
|
| 2789 |
+
"epoch": 18.125572269457162,
|
| 2790 |
+
"grad_norm": 0.016560234129428864,
|
| 2791 |
+
"learning_rate": 1.6742099343473674e-05,
|
| 2792 |
+
"loss": 0.0009,
|
| 2793 |
+
"num_input_tokens_seen": 19433080,
|
| 2794 |
+
"step": 3480
|
| 2795 |
+
},
|
| 2796 |
+
{
|
| 2797 |
+
"epoch": 18.177894048397647,
|
| 2798 |
+
"grad_norm": 0.3799145519733429,
|
| 2799 |
+
"learning_rate": 1.6612849923924723e-05,
|
| 2800 |
+
"loss": 0.0003,
|
| 2801 |
+
"num_input_tokens_seen": 19489176,
|
| 2802 |
+
"step": 3490
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 18.230215827338128,
|
| 2806 |
+
"grad_norm": 0.9833922982215881,
|
| 2807 |
+
"learning_rate": 1.6483852622360923e-05,
|
| 2808 |
+
"loss": 0.0003,
|
| 2809 |
+
"num_input_tokens_seen": 19544920,
|
| 2810 |
+
"step": 3500
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 18.282537606278613,
|
| 2814 |
+
"grad_norm": 0.017415842041373253,
|
| 2815 |
+
"learning_rate": 1.635511131644505e-05,
|
| 2816 |
+
"loss": 0.0008,
|
| 2817 |
+
"num_input_tokens_seen": 19600888,
|
| 2818 |
+
"step": 3510
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 18.334859385219097,
|
| 2822 |
+
"grad_norm": 0.05391722172498703,
|
| 2823 |
+
"learning_rate": 1.6226629876144657e-05,
|
| 2824 |
+
"loss": 0.0002,
|
| 2825 |
+
"num_input_tokens_seen": 19656168,
|
| 2826 |
+
"step": 3520
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"epoch": 18.387181164159582,
|
| 2830 |
+
"grad_norm": 0.32104507088661194,
|
| 2831 |
+
"learning_rate": 1.609841216361574e-05,
|
| 2832 |
+
"loss": 0.0005,
|
| 2833 |
+
"num_input_tokens_seen": 19711224,
|
| 2834 |
+
"step": 3530
|
| 2835 |
+
},
|
| 2836 |
+
{
|
| 2837 |
+
"epoch": 18.439502943100067,
|
| 2838 |
+
"grad_norm": 0.019494058564305305,
|
| 2839 |
+
"learning_rate": 1.597046203308662e-05,
|
| 2840 |
+
"loss": 0.0004,
|
| 2841 |
+
"num_input_tokens_seen": 19767768,
|
| 2842 |
+
"step": 3540
|
| 2843 |
+
},
|
| 2844 |
+
{
|
| 2845 |
+
"epoch": 18.491824722040548,
|
| 2846 |
+
"grad_norm": 0.03266040235757828,
|
| 2847 |
+
"learning_rate": 1.584278333074208e-05,
|
| 2848 |
+
"loss": 0.0001,
|
| 2849 |
+
"num_input_tokens_seen": 19824616,
|
| 2850 |
+
"step": 3550
|
| 2851 |
+
},
|
| 2852 |
+
{
|
| 2853 |
+
"epoch": 18.544146500981032,
|
| 2854 |
+
"grad_norm": 0.024637416005134583,
|
| 2855 |
+
"learning_rate": 1.571537989460779e-05,
|
| 2856 |
+
"loss": 0.0007,
|
| 2857 |
+
"num_input_tokens_seen": 19880024,
|
| 2858 |
+
"step": 3560
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 18.596468279921517,
|
| 2862 |
+
"grad_norm": 0.03031347133219242,
|
| 2863 |
+
"learning_rate": 1.5588255554434883e-05,
|
| 2864 |
+
"loss": 0.0003,
|
| 2865 |
+
"num_input_tokens_seen": 19936504,
|
| 2866 |
+
"step": 3570
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 18.648790058862,
|
| 2870 |
+
"grad_norm": 0.005317925941199064,
|
| 2871 |
+
"learning_rate": 1.5461414131584873e-05,
|
| 2872 |
+
"loss": 0.0006,
|
| 2873 |
+
"num_input_tokens_seen": 19992136,
|
| 2874 |
+
"step": 3580
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 18.701111837802486,
|
| 2878 |
+
"grad_norm": 0.024280209094285965,
|
| 2879 |
+
"learning_rate": 1.533485943891478e-05,
|
| 2880 |
+
"loss": 0.0035,
|
| 2881 |
+
"num_input_tokens_seen": 20049128,
|
| 2882 |
+
"step": 3590
|
| 2883 |
+
},
|
| 2884 |
+
{
|
| 2885 |
+
"epoch": 18.75343361674297,
|
| 2886 |
+
"grad_norm": 0.9302756190299988,
|
| 2887 |
+
"learning_rate": 1.5208595280662497e-05,
|
| 2888 |
+
"loss": 0.0002,
|
| 2889 |
+
"num_input_tokens_seen": 20106488,
|
| 2890 |
+
"step": 3600
|
| 2891 |
+
},
|
| 2892 |
+
{
|
| 2893 |
+
"epoch": 18.805755395683452,
|
| 2894 |
+
"grad_norm": 0.013046924024820328,
|
| 2895 |
+
"learning_rate": 1.5082625452332433e-05,
|
| 2896 |
+
"loss": 0.0003,
|
| 2897 |
+
"num_input_tokens_seen": 20162536,
|
| 2898 |
+
"step": 3610
|
| 2899 |
+
},
|
| 2900 |
+
{
|
| 2901 |
+
"epoch": 18.858077174623936,
|
| 2902 |
+
"grad_norm": 0.01642036624252796,
|
| 2903 |
+
"learning_rate": 1.4956953740581454e-05,
|
| 2904 |
+
"loss": 0.0003,
|
| 2905 |
+
"num_input_tokens_seen": 20219032,
|
| 2906 |
+
"step": 3620
|
| 2907 |
+
},
|
| 2908 |
+
{
|
| 2909 |
+
"epoch": 18.91039895356442,
|
| 2910 |
+
"grad_norm": 0.01480098720639944,
|
| 2911 |
+
"learning_rate": 1.4831583923104999e-05,
|
| 2912 |
+
"loss": 0.0003,
|
| 2913 |
+
"num_input_tokens_seen": 20275880,
|
| 2914 |
+
"step": 3630
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 18.962720732504906,
|
| 2918 |
+
"grad_norm": 0.05232972651720047,
|
| 2919 |
+
"learning_rate": 1.4706519768523597e-05,
|
| 2920 |
+
"loss": 0.0006,
|
| 2921 |
+
"num_input_tokens_seen": 20332264,
|
| 2922 |
+
"step": 3640
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 19.0104643557881,
|
| 2926 |
+
"grad_norm": 0.008723029866814613,
|
| 2927 |
+
"learning_rate": 1.458176503626949e-05,
|
| 2928 |
+
"loss": 0.0001,
|
| 2929 |
+
"num_input_tokens_seen": 20382464,
|
| 2930 |
+
"step": 3650
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 19.06278613472858,
|
| 2934 |
+
"grad_norm": 0.35439035296440125,
|
| 2935 |
+
"learning_rate": 1.4457323476473738e-05,
|
| 2936 |
+
"loss": 0.0005,
|
| 2937 |
+
"num_input_tokens_seen": 20438720,
|
| 2938 |
+
"step": 3660
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 19.115107913669064,
|
| 2942 |
+
"grad_norm": 0.02343558706343174,
|
| 2943 |
+
"learning_rate": 1.4333198829853394e-05,
|
| 2944 |
+
"loss": 0.0001,
|
| 2945 |
+
"num_input_tokens_seen": 20493616,
|
| 2946 |
+
"step": 3670
|
| 2947 |
+
},
|
| 2948 |
+
{
|
| 2949 |
+
"epoch": 19.16742969260955,
|
| 2950 |
+
"grad_norm": 0.014349430799484253,
|
| 2951 |
+
"learning_rate": 1.420939482759907e-05,
|
| 2952 |
+
"loss": 0.0003,
|
| 2953 |
+
"num_input_tokens_seen": 20550000,
|
| 2954 |
+
"step": 3680
|
| 2955 |
+
},
|
| 2956 |
+
{
|
| 2957 |
+
"epoch": 19.219751471550033,
|
| 2958 |
+
"grad_norm": 0.044981323182582855,
|
| 2959 |
+
"learning_rate": 1.4085915191262832e-05,
|
| 2960 |
+
"loss": 0.0001,
|
| 2961 |
+
"num_input_tokens_seen": 20606144,
|
| 2962 |
+
"step": 3690
|
| 2963 |
+
},
|
| 2964 |
+
{
|
| 2965 |
+
"epoch": 19.272073250490518,
|
| 2966 |
+
"grad_norm": 0.023957155644893646,
|
| 2967 |
+
"learning_rate": 1.396276363264629e-05,
|
| 2968 |
+
"loss": 0.0001,
|
| 2969 |
+
"num_input_tokens_seen": 20662720,
|
| 2970 |
+
"step": 3700
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 19.324395029431,
|
| 2974 |
+
"grad_norm": 0.015040691941976547,
|
| 2975 |
+
"learning_rate": 1.3839943853689024e-05,
|
| 2976 |
+
"loss": 0.0002,
|
| 2977 |
+
"num_input_tokens_seen": 20718992,
|
| 2978 |
+
"step": 3710
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 19.376716808371484,
|
| 2982 |
+
"grad_norm": 0.011137389577925205,
|
| 2983 |
+
"learning_rate": 1.3717459546357284e-05,
|
| 2984 |
+
"loss": 0.0001,
|
| 2985 |
+
"num_input_tokens_seen": 20776096,
|
| 2986 |
+
"step": 3720
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 19.429038587311968,
|
| 2990 |
+
"grad_norm": 0.059972431510686874,
|
| 2991 |
+
"learning_rate": 1.3595314392533083e-05,
|
| 2992 |
+
"loss": 0.0003,
|
| 2993 |
+
"num_input_tokens_seen": 20831584,
|
| 2994 |
+
"step": 3730
|
| 2995 |
+
},
|
| 2996 |
+
{
|
| 2997 |
+
"epoch": 19.481360366252453,
|
| 2998 |
+
"grad_norm": 0.003549647517502308,
|
| 2999 |
+
"learning_rate": 1.3473512063903432e-05,
|
| 3000 |
+
"loss": 0.0006,
|
| 3001 |
+
"num_input_tokens_seen": 20887408,
|
| 3002 |
+
"step": 3740
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 19.533682145192937,
|
| 3006 |
+
"grad_norm": 0.01864522323012352,
|
| 3007 |
+
"learning_rate": 1.335205622185003e-05,
|
| 3008 |
+
"loss": 0.0003,
|
| 3009 |
+
"num_input_tokens_seen": 20944080,
|
| 3010 |
+
"step": 3750
|
| 3011 |
+
},
|
| 3012 |
+
{
|
| 3013 |
+
"epoch": 19.586003924133422,
|
| 3014 |
+
"grad_norm": 0.013788777403533459,
|
| 3015 |
+
"learning_rate": 1.3230950517339141e-05,
|
| 3016 |
+
"loss": 0.0001,
|
| 3017 |
+
"num_input_tokens_seen": 21000576,
|
| 3018 |
+
"step": 3760
|
| 3019 |
+
},
|
| 3020 |
+
{
|
| 3021 |
+
"epoch": 19.638325703073903,
|
| 3022 |
+
"grad_norm": 0.007708332501351833,
|
| 3023 |
+
"learning_rate": 1.3110198590811918e-05,
|
| 3024 |
+
"loss": 0.0001,
|
| 3025 |
+
"num_input_tokens_seen": 21056608,
|
| 3026 |
+
"step": 3770
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 19.690647482014388,
|
| 3030 |
+
"grad_norm": 0.008117050863802433,
|
| 3031 |
+
"learning_rate": 1.2989804072074918e-05,
|
| 3032 |
+
"loss": 0.0001,
|
| 3033 |
+
"num_input_tokens_seen": 21112528,
|
| 3034 |
+
"step": 3780
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 19.742969260954872,
|
| 3038 |
+
"grad_norm": 0.005804878659546375,
|
| 3039 |
+
"learning_rate": 1.2869770580191051e-05,
|
| 3040 |
+
"loss": 0.0001,
|
| 3041 |
+
"num_input_tokens_seen": 21169104,
|
| 3042 |
+
"step": 3790
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 19.795291039895357,
|
| 3046 |
+
"grad_norm": 0.021145416423678398,
|
| 3047 |
+
"learning_rate": 1.2750101723370683e-05,
|
| 3048 |
+
"loss": 0.0021,
|
| 3049 |
+
"num_input_tokens_seen": 21225440,
|
| 3050 |
+
"step": 3800
|
| 3051 |
+
},
|
| 3052 |
+
{
|
| 3053 |
+
"epoch": 19.84761281883584,
|
| 3054 |
+
"grad_norm": 0.031260546296834946,
|
| 3055 |
+
"learning_rate": 1.2630801098863284e-05,
|
| 3056 |
+
"loss": 0.0014,
|
| 3057 |
+
"num_input_tokens_seen": 21281952,
|
| 3058 |
+
"step": 3810
|
| 3059 |
+
},
|
| 3060 |
+
{
|
| 3061 |
+
"epoch": 19.899934597776323,
|
| 3062 |
+
"grad_norm": 0.024749331176280975,
|
| 3063 |
+
"learning_rate": 1.2511872292849236e-05,
|
| 3064 |
+
"loss": 0.0001,
|
| 3065 |
+
"num_input_tokens_seen": 21338448,
|
| 3066 |
+
"step": 3820
|
| 3067 |
+
},
|
| 3068 |
+
{
|
| 3069 |
+
"epoch": 19.952256376716807,
|
| 3070 |
+
"grad_norm": 0.11673085391521454,
|
| 3071 |
+
"learning_rate": 1.2393318880332062e-05,
|
| 3072 |
+
"loss": 0.0001,
|
| 3073 |
+
"num_input_tokens_seen": 21394640,
|
| 3074 |
+
"step": 3830
|
| 3075 |
+
},
|
| 3076 |
+
{
|
| 3077 |
+
"epoch": 20.0,
|
| 3078 |
+
"grad_norm": 0.07176396250724792,
|
| 3079 |
+
"learning_rate": 1.2275144425030902e-05,
|
| 3080 |
+
"loss": 0.0003,
|
| 3081 |
+
"num_input_tokens_seen": 21445504,
|
| 3082 |
+
"step": 3840
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 20.052321778940485,
|
| 3086 |
+
"grad_norm": 0.024757077917456627,
|
| 3087 |
+
"learning_rate": 1.2157352479273465e-05,
|
| 3088 |
+
"loss": 0.0004,
|
| 3089 |
+
"num_input_tokens_seen": 21503072,
|
| 3090 |
+
"step": 3850
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 20.10464355788097,
|
| 3094 |
+
"grad_norm": 0.016589034348726273,
|
| 3095 |
+
"learning_rate": 1.2039946583889225e-05,
|
| 3096 |
+
"loss": 0.0001,
|
| 3097 |
+
"num_input_tokens_seen": 21559312,
|
| 3098 |
+
"step": 3860
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 20.15696533682145,
|
| 3102 |
+
"grad_norm": 0.004230449441820383,
|
| 3103 |
+
"learning_rate": 1.1922930268102949e-05,
|
| 3104 |
+
"loss": 0.0001,
|
| 3105 |
+
"num_input_tokens_seen": 21616032,
|
| 3106 |
+
"step": 3870
|
| 3107 |
+
},
|
| 3108 |
+
{
|
| 3109 |
+
"epoch": 20.209287115761935,
|
| 3110 |
+
"grad_norm": 0.007515770383179188,
|
| 3111 |
+
"learning_rate": 1.1806307049428616e-05,
|
| 3112 |
+
"loss": 0.0003,
|
| 3113 |
+
"num_input_tokens_seen": 21671872,
|
| 3114 |
+
"step": 3880
|
| 3115 |
+
},
|
| 3116 |
+
{
|
| 3117 |
+
"epoch": 20.26160889470242,
|
| 3118 |
+
"grad_norm": 0.0041547054424881935,
|
| 3119 |
+
"learning_rate": 1.1690080433563716e-05,
|
| 3120 |
+
"loss": 0.0001,
|
| 3121 |
+
"num_input_tokens_seen": 21727616,
|
| 3122 |
+
"step": 3890
|
| 3123 |
+
},
|
| 3124 |
+
{
|
| 3125 |
+
"epoch": 20.313930673642904,
|
| 3126 |
+
"grad_norm": 0.0042037139646708965,
|
| 3127 |
+
"learning_rate": 1.157425391428384e-05,
|
| 3128 |
+
"loss": 0.0,
|
| 3129 |
+
"num_input_tokens_seen": 21784400,
|
| 3130 |
+
"step": 3900
|
| 3131 |
+
},
|
| 3132 |
+
{
|
| 3133 |
+
"epoch": 20.36625245258339,
|
| 3134 |
+
"grad_norm": 0.024101046845316887,
|
| 3135 |
+
"learning_rate": 1.145883097333767e-05,
|
| 3136 |
+
"loss": 0.0001,
|
| 3137 |
+
"num_input_tokens_seen": 21841584,
|
| 3138 |
+
"step": 3910
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 20.418574231523873,
|
| 3142 |
+
"grad_norm": 0.01927500218153,
|
| 3143 |
+
"learning_rate": 1.1343815080342279e-05,
|
| 3144 |
+
"loss": 0.0,
|
| 3145 |
+
"num_input_tokens_seen": 21897120,
|
| 3146 |
+
"step": 3920
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 20.470896010464354,
|
| 3150 |
+
"grad_norm": 0.005595839582383633,
|
| 3151 |
+
"learning_rate": 1.1229209692678921e-05,
|
| 3152 |
+
"loss": 0.0001,
|
| 3153 |
+
"num_input_tokens_seen": 21952320,
|
| 3154 |
+
"step": 3930
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 20.52321778940484,
|
| 3158 |
+
"grad_norm": 0.00780284171923995,
|
| 3159 |
+
"learning_rate": 1.1115018255389006e-05,
|
| 3160 |
+
"loss": 0.0001,
|
| 3161 |
+
"num_input_tokens_seen": 22008432,
|
| 3162 |
+
"step": 3940
|
| 3163 |
+
},
|
| 3164 |
+
{
|
| 3165 |
+
"epoch": 20.575539568345324,
|
| 3166 |
+
"grad_norm": 0.0051256874576210976,
|
| 3167 |
+
"learning_rate": 1.1001244201070606e-05,
|
| 3168 |
+
"loss": 0.0001,
|
| 3169 |
+
"num_input_tokens_seen": 22063664,
|
| 3170 |
+
"step": 3950
|
| 3171 |
+
},
|
| 3172 |
+
{
|
| 3173 |
+
"epoch": 20.62786134728581,
|
| 3174 |
+
"grad_norm": 0.0060654510743916035,
|
| 3175 |
+
"learning_rate": 1.088789094977522e-05,
|
| 3176 |
+
"loss": 0.0,
|
| 3177 |
+
"num_input_tokens_seen": 22119488,
|
| 3178 |
+
"step": 3960
|
| 3179 |
+
},
|
| 3180 |
+
{
|
| 3181 |
+
"epoch": 20.680183126226293,
|
| 3182 |
+
"grad_norm": 0.019679056480526924,
|
| 3183 |
+
"learning_rate": 1.077496190890502e-05,
|
| 3184 |
+
"loss": 0.0,
|
| 3185 |
+
"num_input_tokens_seen": 22175568,
|
| 3186 |
+
"step": 3970
|
| 3187 |
+
},
|
| 3188 |
+
{
|
| 3189 |
+
"epoch": 20.732504905166774,
|
| 3190 |
+
"grad_norm": 0.02322162687778473,
|
| 3191 |
+
"learning_rate": 1.0662460473110384e-05,
|
| 3192 |
+
"loss": 0.0,
|
| 3193 |
+
"num_input_tokens_seen": 22231472,
|
| 3194 |
+
"step": 3980
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 20.78482668410726,
|
| 3198 |
+
"grad_norm": 0.001980294706299901,
|
| 3199 |
+
"learning_rate": 1.0550390024187906e-05,
|
| 3200 |
+
"loss": 0.0007,
|
| 3201 |
+
"num_input_tokens_seen": 22287120,
|
| 3202 |
+
"step": 3990
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 20.837148463047743,
|
| 3206 |
+
"grad_norm": 0.3102468252182007,
|
| 3207 |
+
"learning_rate": 1.0438753930978643e-05,
|
| 3208 |
+
"loss": 0.0001,
|
| 3209 |
+
"num_input_tokens_seen": 22342736,
|
| 3210 |
+
"step": 4000
|
| 3211 |
+
}
|
| 3212 |
+
],
|
| 3213 |
+
"logging_steps": 10,
|
| 3214 |
+
"max_steps": 5730,
|
| 3215 |
+
"num_input_tokens_seen": 22342736,
|
| 3216 |
+
"num_train_epochs": 30,
|
| 3217 |
+
"save_steps": 100,
|
| 3218 |
+
"stateful_callbacks": {
|
| 3219 |
+
"TrainerControl": {
|
| 3220 |
+
"args": {
|
| 3221 |
+
"should_epoch_stop": false,
|
| 3222 |
+
"should_evaluate": false,
|
| 3223 |
+
"should_log": false,
|
| 3224 |
+
"should_save": true,
|
| 3225 |
+
"should_training_stop": false
|
| 3226 |
+
},
|
| 3227 |
+
"attributes": {}
|
| 3228 |
+
}
|
| 3229 |
+
},
|
| 3230 |
+
"total_flos": 9.588523420509798e+17,
|
| 3231 |
+
"train_batch_size": 2,
|
| 3232 |
+
"trial_name": null,
|
| 3233 |
+
"trial_params": null
|
| 3234 |
+
}
|
BHC_Test1/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70a1a9572fede9132d2f690b06d50636ee37349818605d61d0d34856cabc787c
|
| 3 |
+
size 7544
|
BHC_Test1/zero_to_fp32.py
ADDED
|
@@ -0,0 +1,760 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#!/usr/bin/env python
|
| 2 |
+
|
| 3 |
+
# Copyright (c) Microsoft Corporation.
|
| 4 |
+
# SPDX-License-Identifier: Apache-2.0
|
| 5 |
+
|
| 6 |
+
# DeepSpeed Team
|
| 7 |
+
|
| 8 |
+
# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
|
| 9 |
+
# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
|
| 10 |
+
# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
|
| 11 |
+
# application.
|
| 12 |
+
#
|
| 13 |
+
# example:
|
| 14 |
+
# python zero_to_fp32.py . output_dir/
|
| 15 |
+
# or
|
| 16 |
+
# python zero_to_fp32.py . output_dir/ --safe_serialization
|
| 17 |
+
|
| 18 |
+
import argparse
|
| 19 |
+
import torch
|
| 20 |
+
import glob
|
| 21 |
+
import math
|
| 22 |
+
import os
|
| 23 |
+
import re
|
| 24 |
+
import gc
|
| 25 |
+
import json
|
| 26 |
+
import numpy as np
|
| 27 |
+
from tqdm import tqdm
|
| 28 |
+
from collections import OrderedDict
|
| 29 |
+
from dataclasses import dataclass
|
| 30 |
+
|
| 31 |
+
# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
|
| 32 |
+
# DeepSpeed data structures it has to be available in the current python environment.
|
| 33 |
+
from deepspeed.utils import logger
|
| 34 |
+
from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
|
| 35 |
+
FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
|
| 36 |
+
FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
|
| 37 |
+
|
| 38 |
+
|
| 39 |
+
@dataclass
|
| 40 |
+
class zero_model_state:
|
| 41 |
+
buffers: dict()
|
| 42 |
+
param_shapes: dict()
|
| 43 |
+
shared_params: list
|
| 44 |
+
ds_version: int
|
| 45 |
+
frozen_param_shapes: dict()
|
| 46 |
+
frozen_param_fragments: dict()
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
debug = 0
|
| 50 |
+
|
| 51 |
+
# load to cpu
|
| 52 |
+
device = torch.device('cpu')
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
def atoi(text):
|
| 56 |
+
return int(text) if text.isdigit() else text
|
| 57 |
+
|
| 58 |
+
|
| 59 |
+
def natural_keys(text):
|
| 60 |
+
'''
|
| 61 |
+
alist.sort(key=natural_keys) sorts in human order
|
| 62 |
+
http://nedbatchelder.com/blog/200712/human_sorting.html
|
| 63 |
+
(See Toothy's implementation in the comments)
|
| 64 |
+
'''
|
| 65 |
+
return [atoi(c) for c in re.split(r'(\d+)', text)]
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
def get_model_state_file(checkpoint_dir, zero_stage):
|
| 69 |
+
if not os.path.isdir(checkpoint_dir):
|
| 70 |
+
raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
|
| 71 |
+
|
| 72 |
+
# there should be only one file
|
| 73 |
+
if zero_stage <= 2:
|
| 74 |
+
file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
|
| 75 |
+
elif zero_stage == 3:
|
| 76 |
+
file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
|
| 77 |
+
|
| 78 |
+
if not os.path.exists(file):
|
| 79 |
+
raise FileNotFoundError(f"can't find model states file at '{file}'")
|
| 80 |
+
|
| 81 |
+
return file
|
| 82 |
+
|
| 83 |
+
|
| 84 |
+
def get_checkpoint_files(checkpoint_dir, glob_pattern):
|
| 85 |
+
# XXX: need to test that this simple glob rule works for multi-node setup too
|
| 86 |
+
ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
|
| 87 |
+
|
| 88 |
+
if len(ckpt_files) == 0:
|
| 89 |
+
raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
|
| 90 |
+
|
| 91 |
+
return ckpt_files
|
| 92 |
+
|
| 93 |
+
|
| 94 |
+
def get_optim_files(checkpoint_dir):
|
| 95 |
+
return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
|
| 96 |
+
|
| 97 |
+
|
| 98 |
+
def get_model_state_files(checkpoint_dir):
|
| 99 |
+
return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
|
| 100 |
+
|
| 101 |
+
|
| 102 |
+
def parse_model_states(files):
|
| 103 |
+
zero_model_states = []
|
| 104 |
+
for file in files:
|
| 105 |
+
state_dict = torch.load(file, map_location=device, weights_only=False)
|
| 106 |
+
|
| 107 |
+
if BUFFER_NAMES not in state_dict:
|
| 108 |
+
raise ValueError(f"{file} is not a model state checkpoint")
|
| 109 |
+
buffer_names = state_dict[BUFFER_NAMES]
|
| 110 |
+
if debug:
|
| 111 |
+
print("Found buffers:", buffer_names)
|
| 112 |
+
|
| 113 |
+
# recover just the buffers while restoring them to fp32 if they were saved in fp16
|
| 114 |
+
buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
|
| 115 |
+
param_shapes = state_dict[PARAM_SHAPES]
|
| 116 |
+
|
| 117 |
+
# collect parameters that are included in param_shapes
|
| 118 |
+
param_names = []
|
| 119 |
+
for s in param_shapes:
|
| 120 |
+
for name in s.keys():
|
| 121 |
+
param_names.append(name)
|
| 122 |
+
|
| 123 |
+
# update with frozen parameters
|
| 124 |
+
frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
|
| 125 |
+
if frozen_param_shapes is not None:
|
| 126 |
+
if debug:
|
| 127 |
+
print(f"Found frozen_param_shapes: {frozen_param_shapes}")
|
| 128 |
+
param_names += list(frozen_param_shapes.keys())
|
| 129 |
+
|
| 130 |
+
# handle shared params
|
| 131 |
+
shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
|
| 132 |
+
|
| 133 |
+
ds_version = state_dict.get(DS_VERSION, None)
|
| 134 |
+
|
| 135 |
+
frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
|
| 136 |
+
|
| 137 |
+
z_model_state = zero_model_state(buffers=buffers,
|
| 138 |
+
param_shapes=param_shapes,
|
| 139 |
+
shared_params=shared_params,
|
| 140 |
+
ds_version=ds_version,
|
| 141 |
+
frozen_param_shapes=frozen_param_shapes,
|
| 142 |
+
frozen_param_fragments=frozen_param_fragments)
|
| 143 |
+
zero_model_states.append(z_model_state)
|
| 144 |
+
|
| 145 |
+
return zero_model_states
|
| 146 |
+
|
| 147 |
+
|
| 148 |
+
def parse_optim_states(files, ds_checkpoint_dir):
|
| 149 |
+
total_files = len(files)
|
| 150 |
+
state_dicts = []
|
| 151 |
+
for f in tqdm(files, desc='Loading checkpoint shards'):
|
| 152 |
+
state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
|
| 153 |
+
# immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
|
| 154 |
+
# and also handle the case where it was already removed by another helper script
|
| 155 |
+
state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
|
| 156 |
+
state_dicts.append(state_dict)
|
| 157 |
+
|
| 158 |
+
if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
|
| 159 |
+
raise ValueError(f"{files[0]} is not a zero checkpoint")
|
| 160 |
+
zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
|
| 161 |
+
world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
|
| 162 |
+
|
| 163 |
+
# For ZeRO-2 each param group can have different partition_count as data parallelism for expert
|
| 164 |
+
# parameters can be different from data parallelism for non-expert parameters. So we can just
|
| 165 |
+
# use the max of the partition_count to get the dp world_size.
|
| 166 |
+
|
| 167 |
+
if type(world_size) is list:
|
| 168 |
+
world_size = max(world_size)
|
| 169 |
+
|
| 170 |
+
if world_size != total_files:
|
| 171 |
+
raise ValueError(
|
| 172 |
+
f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
|
| 173 |
+
"Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
|
| 174 |
+
)
|
| 175 |
+
|
| 176 |
+
# the groups are named differently in each stage
|
| 177 |
+
if zero_stage <= 2:
|
| 178 |
+
fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
|
| 179 |
+
elif zero_stage == 3:
|
| 180 |
+
fp32_groups_key = FP32_FLAT_GROUPS
|
| 181 |
+
else:
|
| 182 |
+
raise ValueError(f"unknown zero stage {zero_stage}")
|
| 183 |
+
|
| 184 |
+
fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
|
| 185 |
+
return zero_stage, world_size, fp32_flat_groups
|
| 186 |
+
|
| 187 |
+
|
| 188 |
+
def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
|
| 189 |
+
"""
|
| 190 |
+
Returns fp32 state_dict reconstructed from ds checkpoint
|
| 191 |
+
|
| 192 |
+
Args:
|
| 193 |
+
- ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
|
| 194 |
+
|
| 195 |
+
"""
|
| 196 |
+
print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
|
| 197 |
+
|
| 198 |
+
optim_files = get_optim_files(ds_checkpoint_dir)
|
| 199 |
+
zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
|
| 200 |
+
print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
|
| 201 |
+
|
| 202 |
+
model_files = get_model_state_files(ds_checkpoint_dir)
|
| 203 |
+
|
| 204 |
+
zero_model_states = parse_model_states(model_files)
|
| 205 |
+
print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
|
| 206 |
+
|
| 207 |
+
if zero_stage <= 2:
|
| 208 |
+
return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 209 |
+
exclude_frozen_parameters)
|
| 210 |
+
elif zero_stage == 3:
|
| 211 |
+
return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 212 |
+
exclude_frozen_parameters)
|
| 213 |
+
|
| 214 |
+
|
| 215 |
+
def _zero2_merge_frozen_params(state_dict, zero_model_states):
|
| 216 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 217 |
+
return
|
| 218 |
+
|
| 219 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 220 |
+
frozen_param_fragments = zero_model_states[0].frozen_param_fragments
|
| 221 |
+
|
| 222 |
+
if debug:
|
| 223 |
+
num_elem = sum(s.numel() for s in frozen_param_shapes.values())
|
| 224 |
+
print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 225 |
+
|
| 226 |
+
wanted_params = len(frozen_param_shapes)
|
| 227 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 228 |
+
avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
|
| 229 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 230 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 231 |
+
|
| 232 |
+
total_params = 0
|
| 233 |
+
total_numel = 0
|
| 234 |
+
for name, shape in frozen_param_shapes.items():
|
| 235 |
+
total_params += 1
|
| 236 |
+
unpartitioned_numel = shape.numel()
|
| 237 |
+
total_numel += unpartitioned_numel
|
| 238 |
+
|
| 239 |
+
state_dict[name] = frozen_param_fragments[name]
|
| 240 |
+
|
| 241 |
+
if debug:
|
| 242 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 243 |
+
|
| 244 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 245 |
+
|
| 246 |
+
|
| 247 |
+
def _has_callable(obj, fn):
|
| 248 |
+
attr = getattr(obj, fn, None)
|
| 249 |
+
return callable(attr)
|
| 250 |
+
|
| 251 |
+
|
| 252 |
+
def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 253 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 254 |
+
|
| 255 |
+
# Reconstruction protocol:
|
| 256 |
+
#
|
| 257 |
+
# XXX: document this
|
| 258 |
+
|
| 259 |
+
if debug:
|
| 260 |
+
for i in range(world_size):
|
| 261 |
+
for j in range(len(fp32_flat_groups[0])):
|
| 262 |
+
print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
|
| 263 |
+
|
| 264 |
+
# XXX: memory usage doubles here (zero2)
|
| 265 |
+
num_param_groups = len(fp32_flat_groups[0])
|
| 266 |
+
merged_single_partition_of_fp32_groups = []
|
| 267 |
+
for i in range(num_param_groups):
|
| 268 |
+
merged_partitions = [sd[i] for sd in fp32_flat_groups]
|
| 269 |
+
full_single_fp32_vector = torch.cat(merged_partitions, 0)
|
| 270 |
+
merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
|
| 271 |
+
avail_numel = sum(
|
| 272 |
+
[full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
|
| 273 |
+
|
| 274 |
+
if debug:
|
| 275 |
+
wanted_params = sum([len(shapes) for shapes in param_shapes])
|
| 276 |
+
wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
|
| 277 |
+
# not asserting if there is a mismatch due to possible padding
|
| 278 |
+
print(f"Have {avail_numel} numels to process.")
|
| 279 |
+
print(f"Need {wanted_numel} numels in {wanted_params} params.")
|
| 280 |
+
|
| 281 |
+
# params
|
| 282 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 283 |
+
# out-of-core computing solution
|
| 284 |
+
total_numel = 0
|
| 285 |
+
total_params = 0
|
| 286 |
+
for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
|
| 287 |
+
offset = 0
|
| 288 |
+
avail_numel = full_single_fp32_vector.numel()
|
| 289 |
+
for name, shape in shapes.items():
|
| 290 |
+
|
| 291 |
+
unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
|
| 292 |
+
total_numel += unpartitioned_numel
|
| 293 |
+
total_params += 1
|
| 294 |
+
|
| 295 |
+
if debug:
|
| 296 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 297 |
+
state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
|
| 298 |
+
offset += unpartitioned_numel
|
| 299 |
+
|
| 300 |
+
# Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
|
| 301 |
+
# avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
|
| 302 |
+
# paddings performed in the code it's almost impossible to predict the exact numbers w/o the
|
| 303 |
+
# live optimizer object, so we are checking that the numbers are within the right range
|
| 304 |
+
align_to = 2 * world_size
|
| 305 |
+
|
| 306 |
+
def zero2_align(x):
|
| 307 |
+
return align_to * math.ceil(x / align_to)
|
| 308 |
+
|
| 309 |
+
if debug:
|
| 310 |
+
print(f"original offset={offset}, avail_numel={avail_numel}")
|
| 311 |
+
|
| 312 |
+
offset = zero2_align(offset)
|
| 313 |
+
avail_numel = zero2_align(avail_numel)
|
| 314 |
+
|
| 315 |
+
if debug:
|
| 316 |
+
print(f"aligned offset={offset}, avail_numel={avail_numel}")
|
| 317 |
+
|
| 318 |
+
# Sanity check
|
| 319 |
+
if offset != avail_numel:
|
| 320 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 321 |
+
|
| 322 |
+
print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
|
| 323 |
+
|
| 324 |
+
|
| 325 |
+
def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 326 |
+
exclude_frozen_parameters):
|
| 327 |
+
state_dict = OrderedDict()
|
| 328 |
+
|
| 329 |
+
# buffers
|
| 330 |
+
buffers = zero_model_states[0].buffers
|
| 331 |
+
state_dict.update(buffers)
|
| 332 |
+
if debug:
|
| 333 |
+
print(f"added {len(buffers)} buffers")
|
| 334 |
+
|
| 335 |
+
if not exclude_frozen_parameters:
|
| 336 |
+
_zero2_merge_frozen_params(state_dict, zero_model_states)
|
| 337 |
+
|
| 338 |
+
_zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 339 |
+
|
| 340 |
+
# recover shared parameters
|
| 341 |
+
for pair in zero_model_states[0].shared_params:
|
| 342 |
+
if pair[1] in state_dict:
|
| 343 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 344 |
+
|
| 345 |
+
return state_dict
|
| 346 |
+
|
| 347 |
+
|
| 348 |
+
def zero3_partitioned_param_info(unpartitioned_numel, world_size):
|
| 349 |
+
remainder = unpartitioned_numel % world_size
|
| 350 |
+
padding_numel = (world_size - remainder) if remainder else 0
|
| 351 |
+
partitioned_numel = math.ceil(unpartitioned_numel / world_size)
|
| 352 |
+
return partitioned_numel, padding_numel
|
| 353 |
+
|
| 354 |
+
|
| 355 |
+
def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
|
| 356 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 357 |
+
return
|
| 358 |
+
|
| 359 |
+
if debug:
|
| 360 |
+
for i in range(world_size):
|
| 361 |
+
num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
|
| 362 |
+
print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 363 |
+
|
| 364 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 365 |
+
wanted_params = len(frozen_param_shapes)
|
| 366 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 367 |
+
avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
|
| 368 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 369 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 370 |
+
|
| 371 |
+
total_params = 0
|
| 372 |
+
total_numel = 0
|
| 373 |
+
for name, shape in zero_model_states[0].frozen_param_shapes.items():
|
| 374 |
+
total_params += 1
|
| 375 |
+
unpartitioned_numel = shape.numel()
|
| 376 |
+
total_numel += unpartitioned_numel
|
| 377 |
+
|
| 378 |
+
param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
|
| 379 |
+
state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
|
| 380 |
+
|
| 381 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 382 |
+
|
| 383 |
+
if debug:
|
| 384 |
+
print(
|
| 385 |
+
f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 386 |
+
)
|
| 387 |
+
|
| 388 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 389 |
+
|
| 390 |
+
|
| 391 |
+
class GatheredTensor:
|
| 392 |
+
"""
|
| 393 |
+
A pseudo tensor that collects partitioned weights.
|
| 394 |
+
It is more memory efficient when there are multiple groups.
|
| 395 |
+
"""
|
| 396 |
+
|
| 397 |
+
def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
|
| 398 |
+
self.flat_groups = flat_groups
|
| 399 |
+
self.flat_groups_offset = flat_groups_offset
|
| 400 |
+
self.offset = offset
|
| 401 |
+
self.partitioned_numel = partitioned_numel
|
| 402 |
+
self.shape = shape
|
| 403 |
+
self.dtype = self.flat_groups[0][0].dtype
|
| 404 |
+
|
| 405 |
+
def contiguous(self):
|
| 406 |
+
"""
|
| 407 |
+
Merge partitioned weights from flat_groups into a single tensor.
|
| 408 |
+
"""
|
| 409 |
+
end_idx = self.offset + self.partitioned_numel
|
| 410 |
+
world_size = len(self.flat_groups)
|
| 411 |
+
pad_flat_param_chunks = []
|
| 412 |
+
|
| 413 |
+
for rank_i in range(world_size):
|
| 414 |
+
# for each rank, we need to collect weights from related group/groups
|
| 415 |
+
flat_groups_at_rank_i = self.flat_groups[rank_i]
|
| 416 |
+
start_group_id = None
|
| 417 |
+
end_group_id = None
|
| 418 |
+
for group_id in range(len(self.flat_groups_offset)):
|
| 419 |
+
if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
|
| 420 |
+
start_group_id = group_id
|
| 421 |
+
if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
|
| 422 |
+
end_group_id = group_id
|
| 423 |
+
break
|
| 424 |
+
# collect weights from related group/groups
|
| 425 |
+
for group_id in range(start_group_id, end_group_id + 1):
|
| 426 |
+
flat_tensor = flat_groups_at_rank_i[group_id]
|
| 427 |
+
start_offset = self.offset - self.flat_groups_offset[group_id]
|
| 428 |
+
end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
|
| 429 |
+
pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
|
| 430 |
+
|
| 431 |
+
# collect weights from all ranks
|
| 432 |
+
pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
|
| 433 |
+
param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
|
| 434 |
+
return param
|
| 435 |
+
|
| 436 |
+
|
| 437 |
+
def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 438 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 439 |
+
avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
|
| 440 |
+
|
| 441 |
+
# Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
|
| 442 |
+
# param, re-consolidating each param, while dealing with padding if any
|
| 443 |
+
|
| 444 |
+
# merge list of dicts, preserving order
|
| 445 |
+
param_shapes = {k: v for d in param_shapes for k, v in d.items()}
|
| 446 |
+
|
| 447 |
+
if debug:
|
| 448 |
+
for i in range(world_size):
|
| 449 |
+
print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
|
| 450 |
+
|
| 451 |
+
wanted_params = len(param_shapes)
|
| 452 |
+
wanted_numel = sum(shape.numel() for shape in param_shapes.values())
|
| 453 |
+
# not asserting if there is a mismatch due to possible padding
|
| 454 |
+
avail_numel = fp32_flat_groups[0].numel() * world_size
|
| 455 |
+
print(f"Trainable params: Have {avail_numel} numels to process.")
|
| 456 |
+
print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
|
| 457 |
+
|
| 458 |
+
# params
|
| 459 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 460 |
+
# out-of-core computing solution
|
| 461 |
+
offset = 0
|
| 462 |
+
total_numel = 0
|
| 463 |
+
total_params = 0
|
| 464 |
+
flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
|
| 465 |
+
for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
|
| 466 |
+
unpartitioned_numel = shape.numel()
|
| 467 |
+
total_numel += unpartitioned_numel
|
| 468 |
+
total_params += 1
|
| 469 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 470 |
+
|
| 471 |
+
if debug:
|
| 472 |
+
print(
|
| 473 |
+
f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 474 |
+
)
|
| 475 |
+
|
| 476 |
+
# memory efficient tensor
|
| 477 |
+
tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
|
| 478 |
+
state_dict[name] = tensor
|
| 479 |
+
offset += partitioned_numel
|
| 480 |
+
|
| 481 |
+
offset *= world_size
|
| 482 |
+
|
| 483 |
+
# Sanity check
|
| 484 |
+
if offset != avail_numel:
|
| 485 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 486 |
+
|
| 487 |
+
print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
|
| 488 |
+
|
| 489 |
+
|
| 490 |
+
def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 491 |
+
exclude_frozen_parameters):
|
| 492 |
+
state_dict = OrderedDict()
|
| 493 |
+
|
| 494 |
+
# buffers
|
| 495 |
+
buffers = zero_model_states[0].buffers
|
| 496 |
+
state_dict.update(buffers)
|
| 497 |
+
if debug:
|
| 498 |
+
print(f"added {len(buffers)} buffers")
|
| 499 |
+
|
| 500 |
+
if not exclude_frozen_parameters:
|
| 501 |
+
_zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
|
| 502 |
+
|
| 503 |
+
_zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 504 |
+
|
| 505 |
+
# recover shared parameters
|
| 506 |
+
for pair in zero_model_states[0].shared_params:
|
| 507 |
+
if pair[1] in state_dict:
|
| 508 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 509 |
+
|
| 510 |
+
return state_dict
|
| 511 |
+
|
| 512 |
+
|
| 513 |
+
def to_torch_tensor(state_dict, return_empty_tensor=False):
|
| 514 |
+
"""
|
| 515 |
+
Convert state_dict of GatheredTensor to torch tensor
|
| 516 |
+
"""
|
| 517 |
+
torch_state_dict = {}
|
| 518 |
+
converted_tensors = {}
|
| 519 |
+
for name, tensor in state_dict.items():
|
| 520 |
+
tensor_id = id(tensor)
|
| 521 |
+
if tensor_id in converted_tensors: # shared tensors
|
| 522 |
+
shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
|
| 523 |
+
torch_state_dict[name] = shared_tensor
|
| 524 |
+
else:
|
| 525 |
+
converted_tensors[tensor_id] = name
|
| 526 |
+
if return_empty_tensor:
|
| 527 |
+
torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
|
| 528 |
+
else:
|
| 529 |
+
torch_state_dict[name] = tensor.contiguous()
|
| 530 |
+
return torch_state_dict
|
| 531 |
+
|
| 532 |
+
|
| 533 |
+
def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
|
| 534 |
+
tag=None,
|
| 535 |
+
exclude_frozen_parameters=False,
|
| 536 |
+
lazy_mode=False):
|
| 537 |
+
"""
|
| 538 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
|
| 539 |
+
``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
|
| 540 |
+
via a model hub.
|
| 541 |
+
|
| 542 |
+
Args:
|
| 543 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder
|
| 544 |
+
- ``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``
|
| 545 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 546 |
+
- ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
|
| 547 |
+
Convert the pesduo tensor to torch tensor by ``.contiguous()``
|
| 548 |
+
|
| 549 |
+
Returns:
|
| 550 |
+
- pytorch ``state_dict``
|
| 551 |
+
|
| 552 |
+
A typical usage might be ::
|
| 553 |
+
|
| 554 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 555 |
+
# do the training and checkpoint saving
|
| 556 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
|
| 557 |
+
model = model.cpu() # move to cpu
|
| 558 |
+
model.load_state_dict(state_dict)
|
| 559 |
+
# submit to model hub or save the model to share with others
|
| 560 |
+
|
| 561 |
+
In this example the ``model`` will no longer be usable in the deepspeed context of the same
|
| 562 |
+
application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 563 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 564 |
+
|
| 565 |
+
If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
|
| 566 |
+
|
| 567 |
+
Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
|
| 568 |
+
You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
|
| 569 |
+
the checkpoint. Or you can load state_dict in lazy mode ::
|
| 570 |
+
|
| 571 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 572 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
|
| 573 |
+
for name, lazy_tensor in state_dict.item():
|
| 574 |
+
tensor = lazy_tensor.contiguous() # to cpu
|
| 575 |
+
print(name, tensor)
|
| 576 |
+
# del tensor to release memory if it no longer in use
|
| 577 |
+
"""
|
| 578 |
+
if tag is None:
|
| 579 |
+
latest_path = os.path.join(checkpoint_dir, 'latest')
|
| 580 |
+
if os.path.isfile(latest_path):
|
| 581 |
+
with open(latest_path, 'r') as fd:
|
| 582 |
+
tag = fd.read().strip()
|
| 583 |
+
else:
|
| 584 |
+
raise ValueError(f"Unable to find 'latest' file at {latest_path}")
|
| 585 |
+
|
| 586 |
+
ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
|
| 587 |
+
|
| 588 |
+
if not os.path.isdir(ds_checkpoint_dir):
|
| 589 |
+
raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
|
| 590 |
+
|
| 591 |
+
state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
|
| 592 |
+
if lazy_mode:
|
| 593 |
+
return state_dict
|
| 594 |
+
else:
|
| 595 |
+
return to_torch_tensor(state_dict)
|
| 596 |
+
|
| 597 |
+
|
| 598 |
+
def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
|
| 599 |
+
output_dir,
|
| 600 |
+
max_shard_size="5GB",
|
| 601 |
+
safe_serialization=False,
|
| 602 |
+
tag=None,
|
| 603 |
+
exclude_frozen_parameters=False):
|
| 604 |
+
"""
|
| 605 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
|
| 606 |
+
loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
|
| 607 |
+
|
| 608 |
+
Args:
|
| 609 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 610 |
+
- ``output_dir``: directory to the pytorch fp32 state_dict output files
|
| 611 |
+
- ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
|
| 612 |
+
- ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
|
| 613 |
+
- ``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``
|
| 614 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 615 |
+
"""
|
| 616 |
+
|
| 617 |
+
# Dependency pre-check
|
| 618 |
+
if safe_serialization:
|
| 619 |
+
try:
|
| 620 |
+
from safetensors.torch import save_file
|
| 621 |
+
except ImportError:
|
| 622 |
+
print('If you want to use `safe_serialization`, please `pip install safetensors`')
|
| 623 |
+
raise
|
| 624 |
+
if max_shard_size is not None:
|
| 625 |
+
try:
|
| 626 |
+
from huggingface_hub import split_torch_state_dict_into_shards
|
| 627 |
+
except ImportError:
|
| 628 |
+
print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
|
| 629 |
+
raise
|
| 630 |
+
|
| 631 |
+
# Convert zero checkpoint to state_dict
|
| 632 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
|
| 633 |
+
tag,
|
| 634 |
+
exclude_frozen_parameters,
|
| 635 |
+
lazy_mode=True)
|
| 636 |
+
|
| 637 |
+
# Shard the model if it is too big.
|
| 638 |
+
weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
|
| 639 |
+
if max_shard_size is not None:
|
| 640 |
+
filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
|
| 641 |
+
# an memory-efficient approach for sharding
|
| 642 |
+
empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
|
| 643 |
+
state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
|
| 644 |
+
filename_pattern=filename_pattern,
|
| 645 |
+
max_shard_size=max_shard_size)
|
| 646 |
+
else:
|
| 647 |
+
from collections import namedtuple
|
| 648 |
+
StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
|
| 649 |
+
state_dict_split = StateDictSplit(is_sharded=False,
|
| 650 |
+
filename_to_tensors={weights_name: list(state_dict.keys())})
|
| 651 |
+
|
| 652 |
+
# Save the model by shard
|
| 653 |
+
os.makedirs(output_dir, exist_ok=True)
|
| 654 |
+
filename_to_tensors = state_dict_split.filename_to_tensors.items()
|
| 655 |
+
for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
|
| 656 |
+
shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
|
| 657 |
+
shard_state_dict = to_torch_tensor(shard_state_dict)
|
| 658 |
+
output_path = os.path.join(output_dir, shard_file)
|
| 659 |
+
if safe_serialization:
|
| 660 |
+
save_file(shard_state_dict, output_path, metadata={"format": "pt"})
|
| 661 |
+
else:
|
| 662 |
+
torch.save(shard_state_dict, output_path)
|
| 663 |
+
# release the memory of current shard
|
| 664 |
+
for tensor_name in list(shard_state_dict.keys()):
|
| 665 |
+
del state_dict[tensor_name]
|
| 666 |
+
del shard_state_dict[tensor_name]
|
| 667 |
+
del shard_state_dict
|
| 668 |
+
gc.collect()
|
| 669 |
+
|
| 670 |
+
# Save index if sharded
|
| 671 |
+
if state_dict_split.is_sharded:
|
| 672 |
+
index = {
|
| 673 |
+
"metadata": state_dict_split.metadata,
|
| 674 |
+
"weight_map": state_dict_split.tensor_to_filename,
|
| 675 |
+
}
|
| 676 |
+
save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
|
| 677 |
+
save_index_file = os.path.join(output_dir, save_index_file)
|
| 678 |
+
with open(save_index_file, "w", encoding="utf-8") as f:
|
| 679 |
+
content = json.dumps(index, indent=2, sort_keys=True) + "\n"
|
| 680 |
+
f.write(content)
|
| 681 |
+
|
| 682 |
+
|
| 683 |
+
def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
|
| 684 |
+
"""
|
| 685 |
+
1. Put the provided model to cpu
|
| 686 |
+
2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
|
| 687 |
+
3. Load it into the provided model
|
| 688 |
+
|
| 689 |
+
Args:
|
| 690 |
+
- ``model``: the model object to update
|
| 691 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 692 |
+
- ``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``
|
| 693 |
+
|
| 694 |
+
Returns:
|
| 695 |
+
- ``model`: modified model
|
| 696 |
+
|
| 697 |
+
Make sure you have plenty of CPU memory available before you call this function. If you don't
|
| 698 |
+
have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
|
| 699 |
+
conveniently placed for you in the checkpoint folder.
|
| 700 |
+
|
| 701 |
+
A typical usage might be ::
|
| 702 |
+
|
| 703 |
+
from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
|
| 704 |
+
model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
|
| 705 |
+
# submit to model hub or save the model to share with others
|
| 706 |
+
|
| 707 |
+
Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
|
| 708 |
+
of the same application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 709 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 710 |
+
|
| 711 |
+
"""
|
| 712 |
+
logger.info(f"Extracting fp32 weights")
|
| 713 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
|
| 714 |
+
|
| 715 |
+
logger.info(f"Overwriting model with fp32 weights")
|
| 716 |
+
model = model.cpu()
|
| 717 |
+
model.load_state_dict(state_dict, strict=False)
|
| 718 |
+
|
| 719 |
+
return model
|
| 720 |
+
|
| 721 |
+
|
| 722 |
+
if __name__ == "__main__":
|
| 723 |
+
parser = argparse.ArgumentParser()
|
| 724 |
+
parser.add_argument("checkpoint_dir",
|
| 725 |
+
type=str,
|
| 726 |
+
help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
|
| 727 |
+
parser.add_argument("output_dir",
|
| 728 |
+
type=str,
|
| 729 |
+
help="directory to the pytorch fp32 state_dict output files"
|
| 730 |
+
"(e.g. path/checkpoint-12-output/)")
|
| 731 |
+
parser.add_argument(
|
| 732 |
+
"--max_shard_size",
|
| 733 |
+
type=str,
|
| 734 |
+
default="5GB",
|
| 735 |
+
help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
|
| 736 |
+
"lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
|
| 737 |
+
"We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
|
| 738 |
+
"without CPU OOM issues.")
|
| 739 |
+
parser.add_argument(
|
| 740 |
+
"--safe_serialization",
|
| 741 |
+
default=False,
|
| 742 |
+
action='store_true',
|
| 743 |
+
help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
|
| 744 |
+
parser.add_argument("-t",
|
| 745 |
+
"--tag",
|
| 746 |
+
type=str,
|
| 747 |
+
default=None,
|
| 748 |
+
help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
|
| 749 |
+
parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
|
| 750 |
+
parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
|
| 751 |
+
args = parser.parse_args()
|
| 752 |
+
|
| 753 |
+
debug = args.debug
|
| 754 |
+
|
| 755 |
+
convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
|
| 756 |
+
args.output_dir,
|
| 757 |
+
max_shard_size=args.max_shard_size,
|
| 758 |
+
safe_serialization=args.safe_serialization,
|
| 759 |
+
tag=args.tag,
|
| 760 |
+
exclude_frozen_parameters=args.exclude_frozen_parameters)
|