Model save
Browse files- README.md +58 -0
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +655 -0
README.md
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Qwen/Qwen3-0.6B
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: Qwen3-0.6B-math-orca-qlora-10k-ep1
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- sft
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for Qwen3-0.6B-math-orca-qlora-10k-ep1
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [Qwen/Qwen3-0.6B](https://huggingface.co/Qwen/Qwen3-0.6B).
|
| 15 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
+
|
| 17 |
+
## Quick start
|
| 18 |
+
|
| 19 |
+
```python
|
| 20 |
+
from transformers import pipeline
|
| 21 |
+
|
| 22 |
+
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
|
| 23 |
+
generator = pipeline("text-generation", model="tyfeng1997/Qwen3-0.6B-math-orca-qlora-10k-ep1", device="cuda")
|
| 24 |
+
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
|
| 25 |
+
print(output["generated_text"])
|
| 26 |
+
```
|
| 27 |
+
|
| 28 |
+
## Training procedure
|
| 29 |
+
|
| 30 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/bofeng1997-ty/qwen3-finetune/runs/pd4yxl0p)
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
This model was trained with SFT.
|
| 34 |
+
|
| 35 |
+
### Framework versions
|
| 36 |
+
|
| 37 |
+
- TRL: 0.18.0.dev0
|
| 38 |
+
- Transformers: 4.52.0.dev0
|
| 39 |
+
- Pytorch: 2.6.0
|
| 40 |
+
- Datasets: 3.5.1
|
| 41 |
+
- Tokenizers: 0.21.1
|
| 42 |
+
|
| 43 |
+
## Citations
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
Cite TRL as:
|
| 48 |
+
|
| 49 |
+
```bibtex
|
| 50 |
+
@misc{vonwerra2022trl,
|
| 51 |
+
title = {{TRL: Transformer Reinforcement Learning}},
|
| 52 |
+
author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
|
| 53 |
+
year = 2020,
|
| 54 |
+
journal = {GitHub repository},
|
| 55 |
+
publisher = {GitHub},
|
| 56 |
+
howpublished = {\url{https://github.com/huggingface/trl}}
|
| 57 |
+
}
|
| 58 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 2.024814536766259e+16,
|
| 3 |
+
"train_loss": 0.3264346291037167,
|
| 4 |
+
"train_runtime": 814.8752,
|
| 5 |
+
"train_samples": 10000,
|
| 6 |
+
"train_samples_per_second": 6.672,
|
| 7 |
+
"train_steps_per_second": 0.417
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 2.024814536766259e+16,
|
| 3 |
+
"train_loss": 0.3264346291037167,
|
| 4 |
+
"train_runtime": 814.8752,
|
| 5 |
+
"train_samples": 10000,
|
| 6 |
+
"train_samples_per_second": 6.672,
|
| 7 |
+
"train_steps_per_second": 0.417
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,655 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 340,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.014705882352941176,
|
| 14 |
+
"grad_norm": 1.2318819761276245,
|
| 15 |
+
"learning_rate": 0.0002,
|
| 16 |
+
"loss": 1.1427,
|
| 17 |
+
"mean_token_accuracy": 0.7514051914215087,
|
| 18 |
+
"num_tokens": 81920.0,
|
| 19 |
+
"step": 5
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"epoch": 0.029411764705882353,
|
| 23 |
+
"grad_norm": 0.612349808216095,
|
| 24 |
+
"learning_rate": 0.0002,
|
| 25 |
+
"loss": 0.5799,
|
| 26 |
+
"mean_token_accuracy": 0.8597629547119141,
|
| 27 |
+
"num_tokens": 163840.0,
|
| 28 |
+
"step": 10
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0.04411764705882353,
|
| 32 |
+
"grad_norm": 0.4277164936065674,
|
| 33 |
+
"learning_rate": 0.0002,
|
| 34 |
+
"loss": 0.4103,
|
| 35 |
+
"mean_token_accuracy": 0.8957844376564026,
|
| 36 |
+
"num_tokens": 245760.0,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.058823529411764705,
|
| 41 |
+
"grad_norm": 0.3906048536300659,
|
| 42 |
+
"learning_rate": 0.0002,
|
| 43 |
+
"loss": 0.4176,
|
| 44 |
+
"mean_token_accuracy": 0.8923509180545807,
|
| 45 |
+
"num_tokens": 327680.0,
|
| 46 |
+
"step": 20
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.07352941176470588,
|
| 50 |
+
"grad_norm": 0.3483637571334839,
|
| 51 |
+
"learning_rate": 0.0002,
|
| 52 |
+
"loss": 0.3835,
|
| 53 |
+
"mean_token_accuracy": 0.8978738963603974,
|
| 54 |
+
"num_tokens": 409600.0,
|
| 55 |
+
"step": 25
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.08823529411764706,
|
| 59 |
+
"grad_norm": 0.35842204093933105,
|
| 60 |
+
"learning_rate": 0.0002,
|
| 61 |
+
"loss": 0.3728,
|
| 62 |
+
"mean_token_accuracy": 0.9008919656276703,
|
| 63 |
+
"num_tokens": 491520.0,
|
| 64 |
+
"step": 30
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.10294117647058823,
|
| 68 |
+
"grad_norm": 0.29427987337112427,
|
| 69 |
+
"learning_rate": 0.0002,
|
| 70 |
+
"loss": 0.339,
|
| 71 |
+
"mean_token_accuracy": 0.9093719601631165,
|
| 72 |
+
"num_tokens": 573440.0,
|
| 73 |
+
"step": 35
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.11764705882352941,
|
| 77 |
+
"grad_norm": 0.3110213875770569,
|
| 78 |
+
"learning_rate": 0.0002,
|
| 79 |
+
"loss": 0.3479,
|
| 80 |
+
"mean_token_accuracy": 0.9065738081932068,
|
| 81 |
+
"num_tokens": 655360.0,
|
| 82 |
+
"step": 40
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.1323529411764706,
|
| 86 |
+
"grad_norm": 0.30392590165138245,
|
| 87 |
+
"learning_rate": 0.0002,
|
| 88 |
+
"loss": 0.3632,
|
| 89 |
+
"mean_token_accuracy": 0.90327467918396,
|
| 90 |
+
"num_tokens": 737280.0,
|
| 91 |
+
"step": 45
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.14705882352941177,
|
| 95 |
+
"grad_norm": 0.3164522647857666,
|
| 96 |
+
"learning_rate": 0.0002,
|
| 97 |
+
"loss": 0.3519,
|
| 98 |
+
"mean_token_accuracy": 0.9053433358669281,
|
| 99 |
+
"num_tokens": 818998.0,
|
| 100 |
+
"step": 50
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.16176470588235295,
|
| 104 |
+
"grad_norm": 0.29904985427856445,
|
| 105 |
+
"learning_rate": 0.0002,
|
| 106 |
+
"loss": 0.359,
|
| 107 |
+
"mean_token_accuracy": 0.9019794702529907,
|
| 108 |
+
"num_tokens": 900918.0,
|
| 109 |
+
"step": 55
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.17647058823529413,
|
| 113 |
+
"grad_norm": 0.3399337828159332,
|
| 114 |
+
"learning_rate": 0.0002,
|
| 115 |
+
"loss": 0.3472,
|
| 116 |
+
"mean_token_accuracy": 0.9052419304847718,
|
| 117 |
+
"num_tokens": 982838.0,
|
| 118 |
+
"step": 60
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.19117647058823528,
|
| 122 |
+
"grad_norm": 0.3235512673854828,
|
| 123 |
+
"learning_rate": 0.0002,
|
| 124 |
+
"loss": 0.3527,
|
| 125 |
+
"mean_token_accuracy": 0.9031891584396362,
|
| 126 |
+
"num_tokens": 1064758.0,
|
| 127 |
+
"step": 65
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 0.20588235294117646,
|
| 131 |
+
"grad_norm": 0.32987555861473083,
|
| 132 |
+
"learning_rate": 0.0002,
|
| 133 |
+
"loss": 0.3596,
|
| 134 |
+
"mean_token_accuracy": 0.9010885059833527,
|
| 135 |
+
"num_tokens": 1145665.0,
|
| 136 |
+
"step": 70
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.22058823529411764,
|
| 140 |
+
"grad_norm": 0.28296959400177,
|
| 141 |
+
"learning_rate": 0.0002,
|
| 142 |
+
"loss": 0.3186,
|
| 143 |
+
"mean_token_accuracy": 0.9128421485424042,
|
| 144 |
+
"num_tokens": 1227585.0,
|
| 145 |
+
"step": 75
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.23529411764705882,
|
| 149 |
+
"grad_norm": 0.2940562665462494,
|
| 150 |
+
"learning_rate": 0.0002,
|
| 151 |
+
"loss": 0.3189,
|
| 152 |
+
"mean_token_accuracy": 0.9119745969772339,
|
| 153 |
+
"num_tokens": 1309505.0,
|
| 154 |
+
"step": 80
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.25,
|
| 158 |
+
"grad_norm": 0.31312814354896545,
|
| 159 |
+
"learning_rate": 0.0002,
|
| 160 |
+
"loss": 0.3366,
|
| 161 |
+
"mean_token_accuracy": 0.9083712756633758,
|
| 162 |
+
"num_tokens": 1390498.0,
|
| 163 |
+
"step": 85
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2647058823529412,
|
| 167 |
+
"grad_norm": 0.2923528254032135,
|
| 168 |
+
"learning_rate": 0.0002,
|
| 169 |
+
"loss": 0.3114,
|
| 170 |
+
"mean_token_accuracy": 0.9138196527957916,
|
| 171 |
+
"num_tokens": 1472418.0,
|
| 172 |
+
"step": 90
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.27941176470588236,
|
| 176 |
+
"grad_norm": 0.2987738847732544,
|
| 177 |
+
"learning_rate": 0.0002,
|
| 178 |
+
"loss": 0.3226,
|
| 179 |
+
"mean_token_accuracy": 0.9115102827548981,
|
| 180 |
+
"num_tokens": 1554338.0,
|
| 181 |
+
"step": 95
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.29411764705882354,
|
| 185 |
+
"grad_norm": 0.3070703446865082,
|
| 186 |
+
"learning_rate": 0.0002,
|
| 187 |
+
"loss": 0.334,
|
| 188 |
+
"mean_token_accuracy": 0.9086510419845581,
|
| 189 |
+
"num_tokens": 1636258.0,
|
| 190 |
+
"step": 100
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.3088235294117647,
|
| 194 |
+
"grad_norm": 0.2919357419013977,
|
| 195 |
+
"learning_rate": 0.0002,
|
| 196 |
+
"loss": 0.322,
|
| 197 |
+
"mean_token_accuracy": 0.9099951267242432,
|
| 198 |
+
"num_tokens": 1718178.0,
|
| 199 |
+
"step": 105
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.3235294117647059,
|
| 203 |
+
"grad_norm": 0.3079027235507965,
|
| 204 |
+
"learning_rate": 0.0002,
|
| 205 |
+
"loss": 0.3195,
|
| 206 |
+
"mean_token_accuracy": 0.9123972117900848,
|
| 207 |
+
"num_tokens": 1799262.0,
|
| 208 |
+
"step": 110
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.3382352941176471,
|
| 212 |
+
"grad_norm": 0.32008472084999084,
|
| 213 |
+
"learning_rate": 0.0002,
|
| 214 |
+
"loss": 0.3211,
|
| 215 |
+
"mean_token_accuracy": 0.9098729312419891,
|
| 216 |
+
"num_tokens": 1881182.0,
|
| 217 |
+
"step": 115
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.35294117647058826,
|
| 221 |
+
"grad_norm": 0.33167868852615356,
|
| 222 |
+
"learning_rate": 0.0002,
|
| 223 |
+
"loss": 0.3165,
|
| 224 |
+
"mean_token_accuracy": 0.9122434020042419,
|
| 225 |
+
"num_tokens": 1963102.0,
|
| 226 |
+
"step": 120
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.36764705882352944,
|
| 230 |
+
"grad_norm": 0.26130759716033936,
|
| 231 |
+
"learning_rate": 0.0002,
|
| 232 |
+
"loss": 0.3101,
|
| 233 |
+
"mean_token_accuracy": 0.9149560272693634,
|
| 234 |
+
"num_tokens": 2045022.0,
|
| 235 |
+
"step": 125
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.38235294117647056,
|
| 239 |
+
"grad_norm": 0.3016408681869507,
|
| 240 |
+
"learning_rate": 0.0002,
|
| 241 |
+
"loss": 0.3149,
|
| 242 |
+
"mean_token_accuracy": 0.9118534028530121,
|
| 243 |
+
"num_tokens": 2126154.0,
|
| 244 |
+
"step": 130
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.39705882352941174,
|
| 248 |
+
"grad_norm": 0.3000870645046234,
|
| 249 |
+
"learning_rate": 0.0002,
|
| 250 |
+
"loss": 0.3157,
|
| 251 |
+
"mean_token_accuracy": 0.9116202533245087,
|
| 252 |
+
"num_tokens": 2208074.0,
|
| 253 |
+
"step": 135
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.4117647058823529,
|
| 257 |
+
"grad_norm": 0.2947154939174652,
|
| 258 |
+
"learning_rate": 0.0002,
|
| 259 |
+
"loss": 0.2991,
|
| 260 |
+
"mean_token_accuracy": 0.916422301530838,
|
| 261 |
+
"num_tokens": 2289994.0,
|
| 262 |
+
"step": 140
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4264705882352941,
|
| 266 |
+
"grad_norm": 0.29345065355300903,
|
| 267 |
+
"learning_rate": 0.0002,
|
| 268 |
+
"loss": 0.3192,
|
| 269 |
+
"mean_token_accuracy": 0.9102272808551788,
|
| 270 |
+
"num_tokens": 2371914.0,
|
| 271 |
+
"step": 145
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.4411764705882353,
|
| 275 |
+
"grad_norm": 0.2984428107738495,
|
| 276 |
+
"learning_rate": 0.0002,
|
| 277 |
+
"loss": 0.298,
|
| 278 |
+
"mean_token_accuracy": 0.9163951098918914,
|
| 279 |
+
"num_tokens": 2453143.0,
|
| 280 |
+
"step": 150
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.45588235294117646,
|
| 284 |
+
"grad_norm": 0.2700878977775574,
|
| 285 |
+
"learning_rate": 0.0002,
|
| 286 |
+
"loss": 0.291,
|
| 287 |
+
"mean_token_accuracy": 0.9183040201663971,
|
| 288 |
+
"num_tokens": 2535063.0,
|
| 289 |
+
"step": 155
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.47058823529411764,
|
| 293 |
+
"grad_norm": 0.30076536536216736,
|
| 294 |
+
"learning_rate": 0.0002,
|
| 295 |
+
"loss": 0.3097,
|
| 296 |
+
"mean_token_accuracy": 0.9130865216255188,
|
| 297 |
+
"num_tokens": 2616983.0,
|
| 298 |
+
"step": 160
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.4852941176470588,
|
| 302 |
+
"grad_norm": 0.30549952387809753,
|
| 303 |
+
"learning_rate": 0.0002,
|
| 304 |
+
"loss": 0.3136,
|
| 305 |
+
"mean_token_accuracy": 0.9121212244033814,
|
| 306 |
+
"num_tokens": 2698903.0,
|
| 307 |
+
"step": 165
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.5,
|
| 311 |
+
"grad_norm": 0.2821143865585327,
|
| 312 |
+
"learning_rate": 0.0002,
|
| 313 |
+
"loss": 0.3006,
|
| 314 |
+
"mean_token_accuracy": 0.9160520434379578,
|
| 315 |
+
"num_tokens": 2780150.0,
|
| 316 |
+
"step": 170
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.5147058823529411,
|
| 320 |
+
"grad_norm": 0.2865024507045746,
|
| 321 |
+
"learning_rate": 0.0002,
|
| 322 |
+
"loss": 0.3109,
|
| 323 |
+
"mean_token_accuracy": 0.9121701002120972,
|
| 324 |
+
"num_tokens": 2862070.0,
|
| 325 |
+
"step": 175
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.5294117647058824,
|
| 329 |
+
"grad_norm": 0.299447238445282,
|
| 330 |
+
"learning_rate": 0.0002,
|
| 331 |
+
"loss": 0.3045,
|
| 332 |
+
"mean_token_accuracy": 0.914674985408783,
|
| 333 |
+
"num_tokens": 2943990.0,
|
| 334 |
+
"step": 180
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.5441176470588235,
|
| 338 |
+
"grad_norm": 0.28584349155426025,
|
| 339 |
+
"learning_rate": 0.0002,
|
| 340 |
+
"loss": 0.296,
|
| 341 |
+
"mean_token_accuracy": 0.9169232726097107,
|
| 342 |
+
"num_tokens": 3025910.0,
|
| 343 |
+
"step": 185
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 0.5588235294117647,
|
| 347 |
+
"grad_norm": 0.28912603855133057,
|
| 348 |
+
"learning_rate": 0.0002,
|
| 349 |
+
"loss": 0.2828,
|
| 350 |
+
"mean_token_accuracy": 0.9202346205711365,
|
| 351 |
+
"num_tokens": 3107830.0,
|
| 352 |
+
"step": 190
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5735294117647058,
|
| 356 |
+
"grad_norm": 0.2780699133872986,
|
| 357 |
+
"learning_rate": 0.0002,
|
| 358 |
+
"loss": 0.2943,
|
| 359 |
+
"mean_token_accuracy": 0.917925238609314,
|
| 360 |
+
"num_tokens": 3189750.0,
|
| 361 |
+
"step": 195
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.5882352941176471,
|
| 365 |
+
"grad_norm": 0.2849072813987732,
|
| 366 |
+
"learning_rate": 0.0002,
|
| 367 |
+
"loss": 0.2909,
|
| 368 |
+
"mean_token_accuracy": 0.9186461567878723,
|
| 369 |
+
"num_tokens": 3271670.0,
|
| 370 |
+
"step": 200
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.6029411764705882,
|
| 374 |
+
"grad_norm": 0.287589967250824,
|
| 375 |
+
"learning_rate": 0.0002,
|
| 376 |
+
"loss": 0.3006,
|
| 377 |
+
"mean_token_accuracy": 0.9150293409824372,
|
| 378 |
+
"num_tokens": 3353590.0,
|
| 379 |
+
"step": 205
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 0.6176470588235294,
|
| 383 |
+
"grad_norm": 0.3039202392101288,
|
| 384 |
+
"learning_rate": 0.0002,
|
| 385 |
+
"loss": 0.3017,
|
| 386 |
+
"mean_token_accuracy": 0.9141373574733734,
|
| 387 |
+
"num_tokens": 3435510.0,
|
| 388 |
+
"step": 210
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6323529411764706,
|
| 392 |
+
"grad_norm": 0.29136523604393005,
|
| 393 |
+
"learning_rate": 0.0002,
|
| 394 |
+
"loss": 0.2937,
|
| 395 |
+
"mean_token_accuracy": 0.9157746970653534,
|
| 396 |
+
"num_tokens": 3517430.0,
|
| 397 |
+
"step": 215
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.6470588235294118,
|
| 401 |
+
"grad_norm": 0.28994059562683105,
|
| 402 |
+
"learning_rate": 0.0002,
|
| 403 |
+
"loss": 0.2948,
|
| 404 |
+
"mean_token_accuracy": 0.9153592526912689,
|
| 405 |
+
"num_tokens": 3599350.0,
|
| 406 |
+
"step": 220
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.6617647058823529,
|
| 410 |
+
"grad_norm": 0.3030713200569153,
|
| 411 |
+
"learning_rate": 0.0002,
|
| 412 |
+
"loss": 0.3021,
|
| 413 |
+
"mean_token_accuracy": 0.9139174222946167,
|
| 414 |
+
"num_tokens": 3681270.0,
|
| 415 |
+
"step": 225
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6764705882352942,
|
| 419 |
+
"grad_norm": 0.2715919017791748,
|
| 420 |
+
"learning_rate": 0.0002,
|
| 421 |
+
"loss": 0.2973,
|
| 422 |
+
"mean_token_accuracy": 0.9151881873607636,
|
| 423 |
+
"num_tokens": 3763190.0,
|
| 424 |
+
"step": 230
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.6911764705882353,
|
| 428 |
+
"grad_norm": 0.29798802733421326,
|
| 429 |
+
"learning_rate": 0.0002,
|
| 430 |
+
"loss": 0.3004,
|
| 431 |
+
"mean_token_accuracy": 0.9151026546955109,
|
| 432 |
+
"num_tokens": 3845110.0,
|
| 433 |
+
"step": 235
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.7058823529411765,
|
| 437 |
+
"grad_norm": 0.31128421425819397,
|
| 438 |
+
"learning_rate": 0.0002,
|
| 439 |
+
"loss": 0.3049,
|
| 440 |
+
"mean_token_accuracy": 0.9125122249126434,
|
| 441 |
+
"num_tokens": 3927030.0,
|
| 442 |
+
"step": 240
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.7205882352941176,
|
| 446 |
+
"grad_norm": 0.282503604888916,
|
| 447 |
+
"learning_rate": 0.0002,
|
| 448 |
+
"loss": 0.2808,
|
| 449 |
+
"mean_token_accuracy": 0.919000506401062,
|
| 450 |
+
"num_tokens": 4008950.0,
|
| 451 |
+
"step": 245
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.7352941176470589,
|
| 455 |
+
"grad_norm": 0.2817753255367279,
|
| 456 |
+
"learning_rate": 0.0002,
|
| 457 |
+
"loss": 0.2879,
|
| 458 |
+
"mean_token_accuracy": 0.9177908301353455,
|
| 459 |
+
"num_tokens": 4090870.0,
|
| 460 |
+
"step": 250
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.75,
|
| 464 |
+
"grad_norm": 0.29370447993278503,
|
| 465 |
+
"learning_rate": 0.0002,
|
| 466 |
+
"loss": 0.2798,
|
| 467 |
+
"mean_token_accuracy": 0.9193670749664307,
|
| 468 |
+
"num_tokens": 4172790.0,
|
| 469 |
+
"step": 255
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.7647058823529411,
|
| 473 |
+
"grad_norm": 0.2587876617908478,
|
| 474 |
+
"learning_rate": 0.0002,
|
| 475 |
+
"loss": 0.2799,
|
| 476 |
+
"mean_token_accuracy": 0.920650064945221,
|
| 477 |
+
"num_tokens": 4254710.0,
|
| 478 |
+
"step": 260
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.7794117647058824,
|
| 482 |
+
"grad_norm": 0.26823118329048157,
|
| 483 |
+
"learning_rate": 0.0002,
|
| 484 |
+
"loss": 0.2896,
|
| 485 |
+
"mean_token_accuracy": 0.9174364805221558,
|
| 486 |
+
"num_tokens": 4336630.0,
|
| 487 |
+
"step": 265
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.7941176470588235,
|
| 491 |
+
"grad_norm": 0.2886073589324951,
|
| 492 |
+
"learning_rate": 0.0002,
|
| 493 |
+
"loss": 0.2807,
|
| 494 |
+
"mean_token_accuracy": 0.9185728430747986,
|
| 495 |
+
"num_tokens": 4418550.0,
|
| 496 |
+
"step": 270
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.8088235294117647,
|
| 500 |
+
"grad_norm": 0.2849334478378296,
|
| 501 |
+
"learning_rate": 0.0002,
|
| 502 |
+
"loss": 0.29,
|
| 503 |
+
"mean_token_accuracy": 0.9182918071746826,
|
| 504 |
+
"num_tokens": 4500470.0,
|
| 505 |
+
"step": 275
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.8235294117647058,
|
| 509 |
+
"grad_norm": 0.3190767467021942,
|
| 510 |
+
"learning_rate": 0.0002,
|
| 511 |
+
"loss": 0.2815,
|
| 512 |
+
"mean_token_accuracy": 0.9185608327388763,
|
| 513 |
+
"num_tokens": 4582187.0,
|
| 514 |
+
"step": 280
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.8382352941176471,
|
| 518 |
+
"grad_norm": 0.28610959649086,
|
| 519 |
+
"learning_rate": 0.0002,
|
| 520 |
+
"loss": 0.2932,
|
| 521 |
+
"mean_token_accuracy": 0.9168866276741028,
|
| 522 |
+
"num_tokens": 4664107.0,
|
| 523 |
+
"step": 285
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.8529411764705882,
|
| 527 |
+
"grad_norm": 0.282124787569046,
|
| 528 |
+
"learning_rate": 0.0002,
|
| 529 |
+
"loss": 0.2833,
|
| 530 |
+
"mean_token_accuracy": 0.9193059802055359,
|
| 531 |
+
"num_tokens": 4746027.0,
|
| 532 |
+
"step": 290
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 0.8676470588235294,
|
| 536 |
+
"grad_norm": 0.27180016040802,
|
| 537 |
+
"learning_rate": 0.0002,
|
| 538 |
+
"loss": 0.2743,
|
| 539 |
+
"mean_token_accuracy": 0.9207478165626526,
|
| 540 |
+
"num_tokens": 4827947.0,
|
| 541 |
+
"step": 295
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.8823529411764706,
|
| 545 |
+
"grad_norm": 0.2949499785900116,
|
| 546 |
+
"learning_rate": 0.0002,
|
| 547 |
+
"loss": 0.2809,
|
| 548 |
+
"mean_token_accuracy": 0.9198436141014099,
|
| 549 |
+
"num_tokens": 4909867.0,
|
| 550 |
+
"step": 300
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.8970588235294118,
|
| 554 |
+
"grad_norm": 0.29020780324935913,
|
| 555 |
+
"learning_rate": 0.0002,
|
| 556 |
+
"loss": 0.2749,
|
| 557 |
+
"mean_token_accuracy": 0.9195137023925781,
|
| 558 |
+
"num_tokens": 4991787.0,
|
| 559 |
+
"step": 305
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.9117647058823529,
|
| 563 |
+
"grad_norm": 0.28802114725112915,
|
| 564 |
+
"learning_rate": 0.0002,
|
| 565 |
+
"loss": 0.2692,
|
| 566 |
+
"mean_token_accuracy": 0.9228883624076843,
|
| 567 |
+
"num_tokens": 5073398.0,
|
| 568 |
+
"step": 310
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.9264705882352942,
|
| 572 |
+
"grad_norm": 0.2924538850784302,
|
| 573 |
+
"learning_rate": 0.0002,
|
| 574 |
+
"loss": 0.2765,
|
| 575 |
+
"mean_token_accuracy": 0.919696980714798,
|
| 576 |
+
"num_tokens": 5155318.0,
|
| 577 |
+
"step": 315
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.9411764705882353,
|
| 581 |
+
"grad_norm": 0.29523536562919617,
|
| 582 |
+
"learning_rate": 0.0002,
|
| 583 |
+
"loss": 0.2729,
|
| 584 |
+
"mean_token_accuracy": 0.920906662940979,
|
| 585 |
+
"num_tokens": 5237238.0,
|
| 586 |
+
"step": 320
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.9558823529411765,
|
| 590 |
+
"grad_norm": 0.2890452444553375,
|
| 591 |
+
"learning_rate": 0.0002,
|
| 592 |
+
"loss": 0.2734,
|
| 593 |
+
"mean_token_accuracy": 0.9217497706413269,
|
| 594 |
+
"num_tokens": 5319158.0,
|
| 595 |
+
"step": 325
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.9705882352941176,
|
| 599 |
+
"grad_norm": 0.2990953326225281,
|
| 600 |
+
"learning_rate": 0.0002,
|
| 601 |
+
"loss": 0.2701,
|
| 602 |
+
"mean_token_accuracy": 0.922544002532959,
|
| 603 |
+
"num_tokens": 5401078.0,
|
| 604 |
+
"step": 330
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.9852941176470589,
|
| 608 |
+
"grad_norm": 0.27057918906211853,
|
| 609 |
+
"learning_rate": 0.0002,
|
| 610 |
+
"loss": 0.2841,
|
| 611 |
+
"mean_token_accuracy": 0.9188294410705566,
|
| 612 |
+
"num_tokens": 5482998.0,
|
| 613 |
+
"step": 335
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.0,
|
| 617 |
+
"grad_norm": 0.29458126425743103,
|
| 618 |
+
"learning_rate": 0.0002,
|
| 619 |
+
"loss": 0.2791,
|
| 620 |
+
"mean_token_accuracy": 0.9199340343475342,
|
| 621 |
+
"num_tokens": 5561846.0,
|
| 622 |
+
"step": 340
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.0,
|
| 626 |
+
"step": 340,
|
| 627 |
+
"total_flos": 2.024814536766259e+16,
|
| 628 |
+
"train_loss": 0.3264346291037167,
|
| 629 |
+
"train_runtime": 814.8752,
|
| 630 |
+
"train_samples_per_second": 6.672,
|
| 631 |
+
"train_steps_per_second": 0.417
|
| 632 |
+
}
|
| 633 |
+
],
|
| 634 |
+
"logging_steps": 5,
|
| 635 |
+
"max_steps": 340,
|
| 636 |
+
"num_input_tokens_seen": 0,
|
| 637 |
+
"num_train_epochs": 1,
|
| 638 |
+
"save_steps": 500,
|
| 639 |
+
"stateful_callbacks": {
|
| 640 |
+
"TrainerControl": {
|
| 641 |
+
"args": {
|
| 642 |
+
"should_epoch_stop": false,
|
| 643 |
+
"should_evaluate": false,
|
| 644 |
+
"should_log": false,
|
| 645 |
+
"should_save": true,
|
| 646 |
+
"should_training_stop": true
|
| 647 |
+
},
|
| 648 |
+
"attributes": {}
|
| 649 |
+
}
|
| 650 |
+
},
|
| 651 |
+
"total_flos": 2.024814536766259e+16,
|
| 652 |
+
"train_batch_size": 8,
|
| 653 |
+
"trial_name": null,
|
| 654 |
+
"trial_params": null
|
| 655 |
+
}
|